EEVblog Electronics Community Forum
Electronics => Manufacturing & Assembly => Topic started by: ServoKit on March 05, 2016, 02:05:24 pm
-
Creating a new thread specific to this particular machine. TVM802B is mostly identical but has more feeders.
Update on TVM802B: The machine comes with Juki / Juki-compatible nozzles. TVM802A has screw-in nozzles like in Neoden TM220/240 and similar.
Cost of machine as purchased: $3,400 including shipping from China to Germany.
Manufacturer: http://www.qihesmt.com/sell-2403372-pick-and-place-machine-vision-mounter-qihe-tvm802a-qihekj-qihe-smt-machine.html (http://www.qihesmt.com/sell-2403372-pick-and-place-machine-vision-mounter-qihe-tvm802a-qihekj-qihe-smt-machine.html)
Blog post re. purchase on Ali Express, importing & customs: http://servokit.com/blog/2016/03/03/pick-place-tvm-802-a-arrived-today/ (http://servokit.com/blog/2016/03/03/pick-place-tvm-802-a-arrived-today/)
Blog post with unboxing (plenty of hi-res images): http://servokit.com/blog/2016/03/04/unboxing-the-pick-and-place-tvm802a/ (http://servokit.com/blog/2016/03/04/unboxing-the-pick-and-place-tvm802a/)
Blog post with setting up the machine: http://servokit.com/blog/2016/03/05/tvm802a-pick-and-place-setup/ (http://servokit.com/blog/2016/03/05/tvm802a-pick-and-place-setup/)
Will post more information and experiences as we go along.
Regards, Axel
-
That table doesn't look very solid - I'd expect it to wobble a bit with the machine running....
-
Yep, it does, a bit. Once we've settled for the layout of the production it will be bolted to the wall.
Regards, Axel
-
Hi
Just to keep this fairly self contained, it might be a good idea to post the current delivered pricing in this thread. All of these machines are "made to sell at a price point". We quickly spiral off comparing $250,000 machines to $3,900 machines and wondering why the $3,900 gizmo "does not measure up". The most basic answer is that you didn't spend $250,000.
====
Bolt the table to a wall. Better yet, bolt it into a corner between two walls.
Bob
-
@ServoKit: Did the machine come with any kind of CE conformity documentation for customs clearance?
Regards,
PG
-
@uncle_bob: Good point, I've added it to the intro post.
@PG: No certificate or declaration of conformity. However, customs didn't ask, they just wanted their pound of flesh...
Regards, Axel
-
Hi Servokit,
Just wanted to congratulate you on the quality of your blog posts.
The guys over in the NeoDen4 forum would do well to take this style and use it as a reference standard, because there are a fair few issues being reported, but little on the steps and procedures that do function, but don't have proper documentation.
Well done and please keep them coming.
-
Hi
So now (at least to me) some of the most fundamental questions:
1) How well does it do on 0.5 mm pitch leaded parts?
2) How well does it do on 0.5 to 1.0 mm pitch BGA's?
You may or may not place a dozen of either one on a board. Most modern boards (other than Blinky Light stuff) is rapidly going to fine(r) pitch parts.
Bob
-
Axel thanks for creating this post and sharing with all of us in such detail.
I'm very excited to see how well this machine works for you. I have high hopes for it based on all the videos I have seen of similar systems.
I'm considering the CharmHigh version with vision + LCD that's selling for $5,000 based on how well this works out for you.
Keep the great review coming along.
Can you share with us what kind of boards you plan on running and how fine pitch the parts you plan on placing?
I would need the machine to place 0.5mm spaced no lead parts accurately.
-
Hi
One thing, at least in my case ... I could care less (to some limit) how long it takes to place a 484 ball FPGA. If it wants to take three pictures of both top and bottom of the part and six pictures of the board, that's fine. It can run in the most awful dead slow mode it wants to while getting the job done. For speed, I care about the other 578 parts on the board, not the single big IC. Doing the test with all the settings set to "be careful" is fine.
Bob
-
@thommo: Thanks!
@Bob, rwb:
We won't do any fine pitch or BGA stuff, at least for now. I see this as an entry level machine to produce the first generation of our controller boards. They have around 30 parts, mostly 0805, some TSSOPs and SOT23. The boards will be panelized with a typical run being 20 panels with 6-8 PCBs each. We wanted to have the boards made but the online calculators we tried were like $10 per board for manufacturing alone. With a little luck the machine will pay for itself after a few runs.
The machine comes with a sample PCB (33 x 100mm) that has some fine pitch components on it (.5 or .65, not sure). Once I get my hands on some cheap chips of the right size I will try this out. BGAs are most certainly not an option with this machine, the cameras are simply not good enough wrt resolution and optics. No leads, maybe.
Regards, Axel
-
When do you plan on running your boards or doing some initial testing?
I'm very eager to see how the software programming and first run goes of for you.
-
Will post a video later today; we hope to finalize the PCB design in two weeks and have the first boards in late March / early April.
Regards, Axel
-
Ok, video of the very first test run is here:
https://www.youtube.com/watch?v=bfJv7-eNLGg (https://www.youtube.com/watch?v=bfJv7-eNLGg)
Blog post: http://servokit.com/blog/2016/03/06/tvm802a-first-test-run/ (http://servokit.com/blog/2016/03/06/tvm802a-first-test-run/)
Regards, Axel
-
For test runs,use thin film doublesided sticky tape to stop placed parts moving
-
That table doesn't look very solid - I'd expect it to wobble a bit with the machine running....
Would the table-wobbling affect the placement? If the machine is internally sturdy enough to not warping while the entire unit moves then it should be all good - or?
-
@matseng: Not the actual placement. The machine itself does not warp or anything, it's rock solid. But when the head moves fast and then stops you feel the table shaking ever so slightly. Parts sticking to solder paste wouldn't move, I think, but chips just lying there do.
@mike: Thanks for the tip.
Regards, Axel
-
If it shakes,parts may jump out of the tape.
-
make a concrete slab to bolt it too.
Problem solved.
-
Concrete maybe a bit Ott, though that,s what Essemtech do for their machines
-
Video #2:
https://www.youtube.com/watch?v=6czoy3XIVsA (https://www.youtube.com/watch?v=6czoy3XIVsA)
I found some empty smd breakout boards in the cave which make for a more realistic test. Also experimented with the speeds and contrast settings for the cameras a bit. I also found that the "Quick" setting for the vision system is more than sufficient if the part is already properly aligned, i.e. not skewed in the tray.
The software has very useful feature, don't know if this is common for that kind of machine: If you have no pick place data you can position the camera over the PCB, move it where the center of a part is and add this coordinate to the part list, kind of a teach-in.
If you look closely, you see that part #3 moves a bit after released by the nozzle. I think this wouldn't happen with paste.
Regards, Axel
-
Excellent high quality videos Axel!
The machine looks to do great in the 2nd video.
The vision makes all the difference with this line of machines.
Now lets see it place 1000 x 0603's with zero errors? :-+
If I end up buying the Charmhigh version I promise to do just as good of a review as you are doing.
-
@thommo: Thanks!
@Bob, rwb:
We won't do any fine pitch or BGA stuff, at least for now. I see this as an entry level machine to produce the first generation of our controller boards. They have around 30 parts, mostly 0805, some TSSOPs and SOT23. The boards will be panelized with a typical run being 20 panels with 6-8 PCBs each. We wanted to have the boards made but the online calculators we tried were like $10 per board for manufacturing alone. With a little luck the machine will pay for itself after a few runs.
The machine comes with a sample PCB (33 x 100mm) that has some fine pitch components on it (.5 or .65, not sure). Once I get my hands on some cheap chips of the right size I will try this out. BGAs are most certainly not an option with this machine, the cameras are simply not good enough wrt resolution and optics. No leads, maybe.
Regards, Axel
Hi
I certainly do not disagree with your evaluation that this is an entry level machine and one can easily ask to much of it. My problem is that when I take a look at modern MCU's and FPGA's it is rare to find a non-BGA part. When you do, 0.5 mm pitch is what they offer. It may be leaded, it may be one of the no lead packages. Either way you are stuck with a beast of a part to place. If I'm looking at that situation on new designs today, it's only going to get worse over a (hopefully) multi-year life of the machine.
======
You might consider putting the machine on the floor until you get your table tied down. I'd also recommend haunting the local sales lists for a real assembly table. Often you can find them cheap (like haul it away) and they are a lot more sturdy than anything you can buy at a big box store.
Bob
-
@Bob: For the second video we tied the table to another one which eliminated almost all shaking (note that we place the same parts in roughly half the time and have better results). The camera is on a tripod isolated from the table and all apparent movement of the image is from focus pumping. For the time being perfectly acceptable.
@rwb: After three days with the machine I certainly wouldn't even consider a system without vision. I think you will want to have at least a camera for the fiducials, everything else is just guessing. The software lets you define multiple fiducials for panelized PCBs and will optionally center each of them before placing parts on the individual PCBs. Very useful and a good compensation for not having encoders.
Regards, Axel
-
Ok, video of the very first test run is here:
https://www.youtube.com/watch?v=bfJv7-eNLGg (https://www.youtube.com/watch?v=bfJv7-eNLGg)
Blog post: http://servokit.com/blog/2016/03/06/tvm802a-first-test-run/ (http://servokit.com/blog/2016/03/06/tvm802a-first-test-run/)
Regards, Axel
That vision is taking forever...
-
Great progress Axel.
Tell me please, is progression of the 'tape' dependant on the 'peeler' effectively pulling it forward, or is the another mechanism to advance the tape which I'm not seeing?
Eg - is there some kind of drag pin positioned in the head unit?
-
I watched these videos and Youtube recommeded this one:
https://youtu.be/IvUU8joBb1Q
We need this guy to build pick and place machines using that novell concept to feed and drop parts
8)
-
make a concrete slab to bolt it too.
Problem solved.
Or buy a nice block of granite. It's plane, clean, heavy, resistant, stable over time and over temperature changes.
-
make a concrete slab to bolt it too.
Problem solved.
Or buy a nice block of granite. It's plane, clean, heavy, resistant, stable over time and over temperature changes.
Hi
If you are actually going to spend real cash on a bench top:
http://www.grizzly.com/products/Maple-Butcher-Block-60-x-24-x-2-1-4-/T21247 (http://www.grizzly.com/products/Maple-Butcher-Block-60-x-24-x-2-1-4-/T21247)
Is a pretty nice way to go. The stuff looks good. You can drill holes in it to run wires through. It's incredibly sturdy. There are a *lot* of people who sell the stuff, so shop around. The thing to watch is the thickness. You probably don't *need* 2.25" thick, but why not ...
Various people will also sell you "two thousand pound" industrial leg sets to go with a top like that. It then becomes a DIY table kit. With some help on the heavy lifting, you can get it together quite quickly. I prefer fixed height legs over the adjustable ones. You always need a back brace to go with the legs.
Bob
-
@thommo: There's a prick in the head that comes down on the arc shaped spring to release the tension on the tape. As far as I can see (without having actual reels) the tape advancement comes solely from pulling the cover tape. Interestingly my machine has two motors for this, one on each side of the reel rack (the CHMT48VA has only one). I hope to have parts on reels later this week and will make a close-up video of the mechanism.
Addendum after closer inspection: The prick comes down a little in front of the spring, presumably into a hole of the tape, the head is then moved away from the feeder effectively pulling the tape, at the same time the motors for pulling the cover tape engage. Then the prick comes up again and the nozzle moves over the tape to pick the part. Video will follow.
Regards, Axel
-
@ServoKit Sigh, these are again carefully choreographed demonstration videos that I would expect to see from a manufacturer. It would really be interesting to see how the machine handles components that are not perfectly centered in the tray (as is the case in your video) and not pushed against an edge but at odd angles in the tray. Let's see this machine properly make an angle adjustment and place the component correctly. Please?
-
@sedelman: The second video was a test of the "Quick" vision algorithm; this will only work if the parts are already properly aligned, i.e. not skewed in the tray. Also, the very small SOIC-8 parts (which would usually be on a tape) have to be properly located in the respective tray pocket or the nozzle can't pick them up.
Regards, Axel
-
Clarifying on my previous post re. the trays.
The tray pockets are roughly 20x20mm. For each tray you specify the location where the nozzle should come down. One option is to use the center of a pocket at x = 10mm, y = 10mm, and place the parts accordingly, like so:
(http://servokit.com/images/tray-1.jpg)
Ideally this should let the nozzle reliably pick up everything that's put in the proper center of the pocket. However, it's easy to misplace parts this way, especially smaller ones. Only 2 mm to the right or so and the pick can fail.
Better use a point near the top of the pocket, like x = 10mm and y = 16mm. Then you can simply shove the parts towards the edge, like so:
(http://servokit.com/images/tray-2.jpg)
As a bonus this placement will mostly eliminate the need for rotating the part for alignment; only left is adjusting for x,y deviation which is much faster.
Of course, eventually we will use custom trays with proper cut-outs of the correct size (the position, layout and dimensions of up to 10 trays can be specified in the software).
Regards, Axel
-
Clarifying on my previous post re. the trays.
The tray pockets are roughly 20x20mm. For each tray you specify the location where the nozzle should come down. One option is to use the center of a pocket at x = 10mm, y = 10mm, and place the parts accordingly, like so:
(http://servokit.com/images/tray-1.jpg)
Ideally this should let the nozzle reliably pick up everything that's put in the proper center of the pocket. However, it's easy to misplace parts this way, especially smaller ones. Only 2 mm to the right or so and the pick can fail.
Better use a point near the top of the pocket, like x = 10mm and y = 16mm. Then you can simply shove the parts towards the edge, like so:
(http://servokit.com/images/tray-2.jpg)
As a bonus this placement will mostly eliminate the need for rotating the part for alignment; only left is adjusting for x,y deviation which is much faster.
Of course, eventually we will use custom trays with proper cut-outs of the correct size (the position, layout and dimensions of up to 10 trays can be specified in the software).
Regards, Axel
Hi
One alternative to fully custom trays are drop in inserts for the trays you have or something very much like them. More or less: A rectangle that fits the opening with one corner chomped out for the part. The stock trays may or may not be suitable for this sort of insert. Unless you have a very set mix of parts, it is a lot easier to organize inserts than to order up fully machined multi cavity trays for each job. The main downside of the insert approach is pulling the setup off the machine and storing it ... not so much.
Bob
-
Those trays are pretty big -can you redefine the tray locations to use a more densely packed tray or strip of taoe? Can you define arbitary tray/strip locations in the placement area?
-
@Mike: Yes, you can define up to ten trays with arbitrary layout (10x1, 1x12, 5x3 etc.) and size within the work area. When the software is running it keeps track of a "tray counter", takes chip from designated tray for PCB #1 from first pocket, chip for PCB #2 from second pocket etc. Strips should work too if you can keep them straight and flat. - As a test, I will 3D print a tray with pockets for the TSSOP-28 we are going to use later this week. And here's something where the mfg. could easily improve the machine: add a grid of threaded holes to the work area to fix up stuff.
@Bob: Inserts are another option, I could problably cut them out of heavy stock with my Cameo.
Regards, Axel
-
@ServoKit It would be nice to see how the machine behaves when parts are aligned non-optimally. I realize the trays provide a form of mechanical centering/alignment when you place the ICs against a corner or edge, etc. but a NeoDen TM220A without vision can do the same. Can you show us how fast the vision works (and if it works) when ICs in the tray are rotated 45DEG in the tray? Many thanks for your consideration.
-
@sedelman: here you go, unedited, except for reencoding:
http://www.servokit.com/images/Produce.mp4 (http://www.servokit.com/images/Produce.mp4)
Vision at accurate, effective speed 30%.
BTW, if you think about it: 45 degrees or more won't work in this configuration (looking from bottom). The system tries to rotate the chip to fit a rectangle which sides are parallel to the X/Y axes of the machine. If the chip is at 45° the machine can't tell what side goes where (Likewise the software doesn't know where Pin 1 of the chip is. In effect the chip must be in the tray already very close to the desired orientation.)
Addendum: In the part list you can specify a rotation angle but that is applied after the chip has been aligned.
Addenum II: It seems to me that ideally you wouldn't have to use vision at all, save for locating the fiducials (something the TM220/240 can't). Every trip to the camera costs time; a bespoke tray would eliminate this need. Vision would be a last resort, like when you only make a few PCBs where a custom tray can't be justified.
Regards, Axel
-
Question:
If the vision system always orients the part to 0 degrees is it then able to place it accurately on the PCB at an aritrary angle?
The obvious answer is yes of course but here's what I'm thinking about:
The part is picked up "off center", ie the nozzle isn't centered on the part. The part is moved to the camera where it gets centered up and oriented to 0 degrees. But now, when it's rotated to the arbitrary angle it's supposed to be placed at on the PCB the parts center "wobbles" since it isn't concentric with the nozzle.
Does the vision system and control software compensate for this wobble or am I overthinking this?
-
Question:
If the vision system always orients the part to 0 degrees is it then able to place it accurately on the PCB at an aritrary angle?
The obvious answer is yes of course but here's what I'm thinking about:
The part is picked up "off center", ie the nozzle isn't centered on the part. The part is moved to the camera where it gets centered up and oriented to 0 degrees. But now, when it's rotated to the arbitrary angle it's supposed to be placed at on the PCB the parts center "wobbles" since it isn't concentric with the nozzle.
Does the vision system and control software compensate for this wobble or am I overthinking this?
Hi
Unless you rotate the part over the vision system and then do a "second look" at it, you will get a skew from an off center pick followed by a random rotation. Simple answer: don't lay out the board so giant / precise parts are rotated 45 degrees.
Bob
-
@H.O: Let's assume the part is oriented properly, i.e. unskewed, but shifted by some millimeters in the tray. The job of the vision system then is to find the offset of the part's center to the center of the nozzle. When this offset is known, the software can a.) correct the path to the part's position on the board by this offset and b.) adjust the center of rotation accordingly if the part has to be rotated before placement.
From a programmer's standpoint this would probably come down to simply concatenating two matrices, one translation matrix for the offset and one rotation matrix that contains the part's desired angle.
Regards, Axel
-
Some notes about adjusting vision contrast and max. usable part height:
http://servokit.com/blog/2016/03/07/notes-on-vision-contrast-fiducials-and-max-part-height-with-the-tvm802a/ (http://servokit.com/blog/2016/03/07/notes-on-vision-contrast-fiducials-and-max-part-height-with-the-tvm802a/)
Regards, Axel
-
Thanks, I understand that it's possible to compensate (it's "just math" right), question is does this machine do it?
As for not laying out boards with parts at 45 degree angle....I think it's fairly common to do exactly that with larger parts (TQFPs etc) since it can make the fan-out easier. Besides, it doesn't have to be 45 degrees or some arbitrary angle. If the vision system always aligns the part to 0 degrees then you'll get the possible COG shift if it's rotatated to 90 (or 180 or 270) degrees as well, right?
But yes, surely the software must compensate for this.
-
I just tried this out and here's what the machine does:
It picks the part, rotates nozzle to apply the desired rotation, then moves to the camera and does the aligning, then places the part on the PCB. Even if the part is shifted several mm in the tray this results in correct placement.
A 45° rotated part would have to be placed either close to square into the tray and have the rotation applied or could be put in a custom diamond shaped tray pocket with no further rotation.
Regards, Axel
-
Some notes about adjusting vision contrast and max. usable part height:
http://servokit.com/blog/2016/03/07/notes-on-vision-contrast-fiducials-and-max-part-height-with-the-tvm802a/ (http://servokit.com/blog/2016/03/07/notes-on-vision-contrast-fiducials-and-max-part-height-with-the-tvm802a/)
Regards, Axel
re. the red resist - fiducials should have a resist-free area around them ( to avoid a misaligned resist overlap ofsetting the position) so shouldn't be an issue if your fids are done properly and the machine's vision is sensible
-
Yes, besides the one hole being to close to the fid, the res-free circle around both fids is very small and the resist layer is misaligned by at least 2/10mm so that the fids don't stand free properly.
Regards, Axel
-
@ServoKit thanks for providing the video showing the vision system in action.
I am sure that you looked at the CHMT48VA as well as the SmallSMT offerings. What motivated you to select the TVM802A over the other choices?
-
@sedelman:
CHMT48VA
1. It has more of the large feeders (16 & 24). That you can define your own trays on the TVM802A makes partly up for that.
2. It has encoders on the steppers. However, I reckoned that for the small boards we make plus having fiducials on each PCB losing steps wouldn't be much of an issue (Let alone that encoders on the steppers are NOT a true closed loop, you can still have slippage with the belts etc. What you really want are optical encoders on the rails, only then does the software know where the head is at any time; AFAIK none of the sub-10K machines has this.)
3. It runs stand-alone while controlled with the touchscreen. Not sure if this is really practical, would love to hear from someone who has used this.
SmallSMT
Judging from the photos the machines look sound. Didnt't look too much into that as I was confused about the offering, which is also apparent in the Neoden 4 thread: Sometimes the machines are called SmallSMT, then Yushengtech, then AutoSMT, then VisionPlacer. Who exactly is the manufacturer? Who do I talk to if I have a problem? Why the flip is the software dongle protected? Confusion is not good when buying big item tickets.
Regards, Axel
-
@servokit I have been thinking about the Charmhigh machine vs your machine which also has vision.
From the looks of it you don't really need the built in LCD screen if you can do everything on the PC.
The Charmhigh does have encoders on the steppers but I'm not sure how beneficial that is with the belt driven machine. I do have a belt driven 3x4 foot laser cutter machine and its very precise using belt's which have been on the machine for years now with zero maintenance. I'm sure the encoders can't hurt though.
Other than that it looks like they machines are almost identical other than a few larger feeders.
For the price Axel paid I have to say it looks to be a bargin now that the machine has vision built in.
Only time will tell how good and reliable it is once it Axel puts it to work building tons of boards.
I'm just curious how capibable these machines are at placing .5mm pitch 32 pin TSSOP chips cause I have a few of those that would need to be placed.
-
Yes, we have two 3D printers in the office, both belt driven, no problems. Losing steps is more of an issue with machines where the head exerts physical force, like with a router or mill. But then again, you would use leadscrews rather than belts on that kind of machine.
Regards, Axel
-
Yea, I have a 3x4 foot cnc machine and it's using a screw drive.
-
Ok, guys, got a nice shipment from Digikey today and took the opportunity to place some resistors:
https://youtu.be/RjAU077Dsmc (https://youtu.be/RjAU077Dsmc)
First half with machine speed at 50%, second at 100% (watch table shake...). No vision, obviously.
Had to use double sided tape or the little critters would just fly off. Resistors are 0805, the smallest part we will be using for our current project but I think that at least 0603 would be ok too.
Regards, Axel
-
Nice, but the 50% speed seemed so slow that I almost think I can do it by hand and faster :-)
Can you repeat the run at 100% speed and have it use both heads?
-
I could but I only have one No. 1 nozzle (the smallest size), the larger nozzles don't work with 0805. 100% is actually not the highest speed, it goes up to 120%...
Regards, Axel
-
Like the Neoden, looks like the speed setting is dumb and applies to all movements - there's no reason to slow down when there is no part on the nozzle (though probably useful to have the option).
-
PS look for clearer double-sided tape, so you can see the pad positions better. I also found that the white stuff I had was a total nightmare to remove.
-
Ok, I found that I can use the No. 2 size nozzle with 0805 if I switch off the pressure sensor (probably wouldn't want to do this in a production setting).
Here you go, two nozzles:
https://youtu.be/6SSU50hn_zI (https://youtu.be/6SSU50hn_zI)
Regards, Axel
-
Looks like it does 3600 CPH utilizing both heads, no vision and the shortest possible distance between the pcb and the reel.
I wonder how the setup must look like to get the 7000 CPH they claim in the specs? Crank it up overdrive at 120% will not take you all the way.
http://www.zjyingxing.com/e_products/TVM802A-126.html (http://www.zjyingxing.com/e_products/TVM802A-126.html)
-
I think it's more like 5000 cph, it does the fiducial recognition before starting to place parts which only has to be done once and after that we get to 1.5parts/sec. or thereabouts. Also, the parts are rotated, that takes time too.
There are a few more arcane settings in the software I didn't explore yet but I think it's safe to say that for all practical purposes the video shows the max. achievable speed. More than fast enough for our purposes, really, but YMMV.
Regards, Axel
-
Looks like it does 3600 CPH utilizing both heads, no vision and the shortest possible distance between the pcb and the reel.
I wonder how the setup must look like to get the 7000 CPH they claim in the specs? Crank it up overdrive at 120% will not take you all the way.
http://www.zjyingxing.com/e_products/TVM802A-126.html (http://www.zjyingxing.com/e_products/TVM802A-126.html)
Maybe parts that happen to be placed in pairs spaced apart by the nozzle distance so it doesn't need to move between placements, and a very small PCB right next to the feeder
-
Good point; I'm curious how the parts are exported from the EDA software. Are close together parts exported as pairs? Maybe an Altium or Eagle user can clarify this. The machine's software does no optimization as far as I can see, the parts are picked in order.
Regards, Axel
-
Looks like it starts the placement @5 seconds and is done @21 so 16 seconds to place 16 parts giving aboutish 3600 cph.
I'd be happy with that speed if I get a PnP some day - then it would only take 20 minutes to populate each my memory pcbs for my (so far only on the drawingboard (and some small real-part tests)) computer made of of discrete transistors and diodes..... Need 128 of them so it will take a while.
-
Hi Axel,
Looks interested machine.
Have you try with LQFP Package Devices ?
and if yes, how is the result ?
Thanks !
-
Not tried. I will try 28 pin TSSOPs with 0.65mm pitch next and report back. That's the highest density which we use in the current project. I did already try SOICs with 1.27mm pitch and they work fine.
Regards, Axel
-
Here is a video if this same machine I came across while on Taobao.com
http://cloud.video.taobao.com/play/u/2381895619/e/1/t/3/p/2/32102185.swf (http://cloud.video.taobao.com/play/u/2381895619/e/1/t/3/p/2/32102185.swf)
http://world.taobao.com/item/521022603339.htm?spm=a312a.7700714.0.0.BsG7LM#detail (http://world.taobao.com/item/521022603339.htm?spm=a312a.7700714.0.0.BsG7LM#detail)
The sale page has this machine listed for $2600 and it says it can place .5mm pitch IC parts.
-
That seems to be the manufacturer's page, QiHe. I'm still in the market for a printer and a larger oven, can anyone comment on those offers:
Printer ($260): http://world.taobao.com/item/525291626153.htm (http://world.taobao.com/item/525291626153.htm)
Oven ($1050): http://world.taobao.com/item/525259351947.htm (http://world.taobao.com/item/525259351947.htm)
Addendum: Aha, the devil is in the details. They quote free shipping but only to mainland China. For overseas destinations it's 90yuan/kg (at least for the printer and oven, for the TVM802A it says contact the seller; at 60kg and this rate this would come to Y5,400 = $830)
Regards, Axel
-
I'm about to buy one of those Stencil Printers, and as for the Oven, I know that the Neoden guys have discontinued carry it. Can't help with better alternative but i suspect it's very much a matter of getting what you pay for re the oven.
I'm considering a small conveyor from Torch, but more expensive.
That seems to be the manufacturer's page, QiHe. I'm still in the market for a printer and a larger oven, can anyone comment on those offers:
Printer ($260): http://world.taobao.com/item/525291626153.htm (http://world.taobao.com/item/525291626153.htm)
Oven ($1050): http://world.taobao.com/item/525259351947.htm (http://world.taobao.com/item/525259351947.htm)
Regards, Axel
-
i suspect it's very much a matter of getting what you pay for re the oven.
That's what I figured. The printer looks solid to me, with shipping that would come out at $600 or so.
Regards, Axel
-
Ok, here's a high speed test with optimized part list:
https://youtu.be/4HvR1EBtv18 (https://youtu.be/4HvR1EBtv18)
Speed 100%, hand-made part list where the space in x between parts is exactly the distance between the nozzles. Parts are resistors 0805. Space in y between parts 3mm. No vision, no rotation, no pressure sensor, no fiducial recognition.
Places 50 parts in a little under 50secs which comes out at somewhere obove 3600 cph. With the 120% speed setting we would get to around 4300-4400 cph.
Interesting observation: The nozzle #2 (right) has clearly more consistent placement than #1, it's the bigger size (No. 2). I wonder if that's because it has better grip on the part.
Regards, Axel
-
Nice little test.
Swap nozzles and see if the effect is the same with the nozzle on the right instead of left.
-
I will. - Apparently the software does some sort of optimization; for the video I had hard coded which nozzle to use; when I leave it to the software (specify nozzle "1/2" rather than "1" or "2"), the machine places the parts column wise, left column first, right column second. This takes a little longer (3-4 secs.).
Regards, Axel
-
So can you mill out your own custom part trays and have the machine pick parts from those custom trays?
I'm wondering if you can use half your work area for custom part trays if you needed that many parts or if you just run out of feeders.
-
Yes, tried that already and it works as expected. I'll post more about it later today.
You can basically use up the whole space that's not occupied by PCBs; obviously there are some constraints from the rods that guide the PCB holder.
Regards, Axel
-
Here's a pic of a sample 3D printed tray.
It has 8 pockets for TSSOP-28 (0.65mm pitch). The little dimples indicate Pin 1 (eventually they will be inlaid with white paint or so).
Note the hole in the center of pocket #1. Using the down camera I can pinpoint the location of the tray as detailed in this blog post:
http://servokit.com/blog/2016/03/09/tvm802a-coordinate-systems/ (http://servokit.com/blog/2016/03/09/tvm802a-coordinate-systems/).
Of course eventually any trays will have to be bolted down, for now it's double-sided tape.
Regards, Axel
-
Sample video using custom tray at 35% effective speed.
https://youtu.be/c6dBbCXtRzo (https://youtu.be/c6dBbCXtRzo)
0.65mm pitch, perfect placement, no vision required.
Regards, Axel
-
That's a nice looking 3D printed IC holder :-+
Nice to see it placing so accurately.
The more I see about this machine the more I like it.
The only thing pushing me towards the Charmhigh version is the better nozzles + encoders + built in LCD which is not needed but is cool. But for the 3K price tag on your machine it's really hard to beat. Needing lots of feeders is what's pushing me towards the Neoden 4 but we have to wait and see how this whole test goes with Ichan.
-
Yes, I think I made the right decision with this machine. - Re. the Charmhigh: I like how the tapes go into the machine to keep the work area clean but where do they come out? Can't tell from the available images.
Regards, Axel
-
@ServoKit Can you comment on the nozzles that the machine is using? Are these industry standard (e.g. Samsung or Juki) or proprietary? Also, how easy is it to change the nozzles? Do you need to use a tool to change it?
-
@ServoKit Can you comment on the nozzles that the machine is using? Are these industry standard (e.g. Samsung or Juki) or proprietary? Also, how easy is it to change the nozzles? Do you need to use a tool to change it?
Hi
What ... no carousel auto changer ??? :)
Bob
-
I don't think any of the current crop of Chinese machines have auto tool-change, which is a shame, as it shouldn't really cost that much - would be nice to have even if an optional extra.
-
I don't think any of the current crop of Chinese machines have auto tool-change, which is a shame, as it shouldn't really cost that much - would be nice to have even if an optional extra.
Hi
The auto change stuff looks nice, but it's easy to get it wrong. I would gladly trade off features like that for a machine that would just chug along placing 0.5 mm pitch IC's at a reasonable rate. If it will do that, then all of the passives I use should be no problem at all. The sub 0402 parts are not that big a part of my mix.
Bob
-
Will this place 0402's?
-
Will this place 0402's?
Hi
If it will place 0.5 mm pitch IC's then 0402 parts should be a pice of cake. They will self align over a much greater range than a big fat IC. The only gotcha might be having a nozzle that is small enough to grab them.
Bob
-
@Robot: I'm fairly certain that 0603 won't be a problem, 0402 don't know. Same with .5 pitch. As tested, .65 is not a problem.
@sedelman: The nozzles are screwed in and locked with two wrenches, replacements are $10-15 or so. I found that I can handle all the parts we use with the No.2 (for resistors and the like) and No.3 nozzle (everything else; this one will even pick and hold an ESP8266-12F reliably, a fairly large component.)
Regards, Axel
-
Here's a more detailed write-up on how to setup custom trays:
http://servokit.com/blog/2016/03/11/setting-up-a-custom-tray-for-the-tvm802a-pick-and-place-machine/ (http://servokit.com/blog/2016/03/11/setting-up-a-custom-tray-for-the-tvm802a-pick-and-place-machine/)
Regards, Axel
-
@ServoKit Can you comment on the nozzles that the machine is using? Are these industry standard (e.g. Samsung or Juki) or proprietary? Also, how easy is it to change the nozzles? Do you need to use a tool to change it?
The nozzles looks like the ones used on the earlier Neoden machines (TM240). Here's a video (https://www.youtube.com/watch?v=3oLVAVeUssU) showing them (Neoden) changing swapping them and if you go to AliExpress or EBAY and search for Neoden nozzles you'll find them. The new Neoden 4 machine looks to be using another type.
-
Unsure about the thread but it looks like they are identical.
Regards, Axel
-
I don't think any of the current crop of Chinese machines have auto tool-change, which is a shame, as it shouldn't really cost that much - would be nice to have even if an optional extra.
Hi
The auto change stuff looks nice, but it's easy to get it wrong. I would gladly trade off features like that for a machine that would just chug along placing 0.5 mm pitch IC's at a reasonable rate. If it will do that, then all of the passives I use should be no problem at all. The sub 0402 parts are not that big a part of my mix.
Bob
That does depend on your part mix - if you're doing a lot of panels that have both small and large parts, swapping nozzles for every panel will be a major pain, especially if they screw on (more so if you need to turn the machine off, as shown in the TM240 video).
Nozzles aren't carrying any load so a push-fit or magnetic retainer would be fine, and more amenable to auto-change. Screw-fit is probably the worst possible choice as it's fiddly, and has the scope to vibrate loose.
-
I don't think any of the current crop of Chinese machines have auto tool-change, which is a shame, as it shouldn't really cost that much - would be nice to have even if an optional extra.
Hi
The auto change stuff looks nice, but it's easy to get it wrong. I would gladly trade off features like that for a machine that would just chug along placing 0.5 mm pitch IC's at a reasonable rate. If it will do that, then all of the passives I use should be no problem at all. The sub 0402 parts are not that big a part of my mix.
Bob
That does depend on your part mik - if you're doing a lot of panels that have both small and large parts, swapping nozzles for every panel will be a major pain, especially if they screw on (more so if you need to turn the machine off, as shown in the TM240 video).
Nozzles aren't carrying any load so a push-fit or magnetic retainer would be fine, and more amenable to auto-change. Screw-fit is probably the worst possible choice as it's fiddly, and has the scope to vibrate loose.
Hi
Given the work area of the machine, you can do a lot of placing on reasonably dense boards before you need to do a nozzle change. With two nozzles, you might / might not get away without a swap.
More or less --- this is why they also sell 4 head machines.
Bob
-
@ServoKit I think it is another plus for the Charmhigh CMHT48VA that it uses industry standard Juki nozzles with slotted/magnetic attachment. You don't need a tool to change them and you can purchase them from anywhere.
You asked the question once about where the empty tape goes on the CMHT48VA. I have the same question, but I think it just rolls up inside the machine. I've not seen the underside, but I expect it is not coming out of the bottom so it must roll up inside. I'm not sure how much it can handle in there and how you retrieve it after a run.
-
@ServoKit I think it is another plus for the Charmhigh CMHT48VA that it uses industry standard Juki nozzles with slotted/magnetic attachment. You don't need a tool to change them and you can purchase them from anywhere.
You asked the question once about where the empty tape goes on the CMHT48VA. I have the same question, but I think it just rolls up inside the machine. I've not seen the underside, but I expect it is not coming out of the bottom so it must roll up inside. I'm not sure how much it can handle in there and how you retrieve it after a run.
Hi
Ok, so how about a thread on the CMHT48VA all by its self? At least as I see it (and I easily could be wrong) there is a price boost between the CMHT and the TVM.
Bob
-
@bob unless someone has a CHMT48VA who can provide facts, I'm not sure if it makes sense to start a new thread. @ServoKit is doing an excellent job reviewing the TVM802A machine. I am actually re-thinking whether it is a good thing to have embedded Linux on the CMHT48VA because it will complicate the firmware update process (for the end-user) and therefore software updates (bug fixes, feature addition, etc.) are likely less forthcoming from the manufacturer than if it was a Windows based solution.
Also, as others have pointed out, the encoders on the stepper motors are useful to detect missteps, but it would have been better if the machine utilized linear optical encoders so that the machine actually knows the actual position at all times. The encoders on the steppers add marginal value.
It looks like Charmhigh came out with a new unit that utilizes pneumatic feeders. The model is CHMT530P and appears to be the same price as the CMHT48VA. These look like Yamaha feeders.
http://www.smtpickandplacemachines.com/sale-7565529-high-speed-smt-pick-and-place-machine-with-pneumatic-feeder-vision-system.html (http://www.smtpickandplacemachines.com/sale-7565529-high-speed-smt-pick-and-place-machine-with-pneumatic-feeder-vision-system.html)
I just confirmed with Kimi that it indeed uses Yamaha feeders. Here is the current pricing:
Unit price: $5160 ;
Shipping to Canada door to door: $760;
Total price: $ 5180+$790 = $5920
Feeder is standard Yamaha feeder, please refer pictures as attached. Feeder
price is as follows:
8mm-- $65
12MM-- $95
16MM-- $95
24MM-- $260
CHMT530P Working video:
https://www.youtube.com/watch?v=fHuPCrryCG0 (https://www.youtube.com/watch?v=fHuPCrryCG0)
-
@sedelman: Juki nozzles are obviously more common and easier to source. I've removed the cover of the head and found that the receptacle for the nozzle is directly attached to a stepper (with hollow shaft for vacuum). It shouldn't be too difficult to either buy or build a Juki compatible receptacle provided you stay within the Z limits of the machine (Z distance from nozzle to rest of machine can be adjusted in the software)
Regards, Axel
-
Hmmm,
Independent Yamaha [pneumatic] feeders are a game changer in my mind.
Looks like the revolution has finally arrived - and at just over $5K.
Now we need Neoden's conveyor/rail system and BAM!
Servokit - Axel, loving your work in this forum!
-
Thanks, thommo.
Look at the specs for the machine. It needs .5MPa for pressure, that's 50bar. That's a serious requirement. The compressor will cost more than the machine...
Regards, Axel
-
Thanks, thommo.
Look at the specs for the machine. It needs .5MPa for pressure, that's 50bar. That's a serious requirement. The compressor will cost more than the machine...
Regards, Axel
0.5 MPa = 5 Bar, not 50
-
Yup, you are right. I used Google for conversion and if I enter 0,5 (german decimal separator) it gives "50"...
Regards, Axel
-
Yup, you are right. I used Google for conversion and if I enter 0,5 (german decimal separator) it gives "50"...
Regards, Axel
just tried, actually google does not care if it is a dot or comma.
-
Don't want to derail this thread but it does here.
Regards, Axel
-
Don't want to derail this thread but it does here.
Ah, if you type into the search "0,5 Mpa = Bar", it auto corrects it to the dot in that box. But if you manually replace it with a comma in that box, then it shows a wrong result.
-
0.5 MPa is only a minimal pressure requirement of under 10 psi. See the attached picture that Kimi sent indicating the size of the compressor that they are using.
I have been using California Air Tools compressor which is very quiet and I can operate it in my lab and still hear myself think.
https://www.youtube.com/watch?v=ZZryRKHrCKc (https://www.youtube.com/watch?v=ZZryRKHrCKc)
-
A few more details about the CMHT530P and then I'll stop hijacking this thread. The 530P has the same mechanical construction as the CMHT48VA and therefore it has encoders on all steppers. The maximum Z-axis travel is 20mm, so the tallest component it can place is under 10mm. If you're placing 10mm tall parts, you need to place them last and from right to left so that the head with part don't bump into the parts already placed. The other difference from the CMHT48VA is the fact that the maximum board size is 350mm x 250mm (as oppossed to 350mm x 350mm for the CMHT48VA). This is still acceptable to me, because I only typically use letter size panels (A4 would work also).
-
@bob unless someone has a CHMT48VA who can provide facts, I'm not sure if it makes sense to start a new thread. @ServoKit is doing an excellent job reviewing the TVM802A machine. I am actually re-thinking whether it is a good thing to have embedded Linux on the CMHT48VA because it will complicate the firmware update process (for the end-user) and therefore software updates (bug fixes, feature addition, etc.) are likely less forthcoming from the manufacturer than if it was a Windows based solution.
Also, as others have pointed out, the encoders on the stepper motors are useful to detect missteps, but it would have been better if the machine utilized linear optical encoders so that the machine actually knows the actual position at all times. The encoders on the steppers add marginal value.
It looks like Charmhigh came out with a new unit that utilizes pneumatic feeders. The model is CHMT530P and appears to be the same price as the CMHT48VA. These look like Yamaha feeders.
http://www.smtpickandplacemachines.com/sale-7565529-high-speed-smt-pick-and-place-machine-with-pneumatic-feeder-vision-system.html (http://www.smtpickandplacemachines.com/sale-7565529-high-speed-smt-pick-and-place-machine-with-pneumatic-feeder-vision-system.html)
I just confirmed with Kimi that it indeed uses Yamaha feeders. Here is the current pricing:
Unit price: $5160 ;
Shipping to Canada door to door: $760;
Total price: $ 5180+$790 = $5920
Feeder is standard Yamaha feeder, please refer pictures as attached. Feeder
price is as follows:
8mm-- $65
12MM-- $95
16MM-- $95
24MM-- $260
CHMT530P Working video:
https://www.youtube.com/watch?v=fHuPCrryCG0 (https://www.youtube.com/watch?v=fHuPCrryCG0)
Hi
My only real issue is that a lot of very useful info on this or that machine gets buried (and tangled up) in other threads. Even without buying one, you obviously have done a lot of research on their machines. Sometimes even a "this came out and then that came out and now they make this other one" sort of trail can be really useful. Zigs and zags adding this or that feature don't bother me a lot. Going to an approach that likely makes the old firmware obsolete .... that worries me. Often a decision gets made (hopefully years from now) to drop support for this or that "old feature". Yes, that's just an observation, but it does apply to the feeder change they just made.
Bob
-
@sedelman: Follow up on the nozzles, here's an adapter for Juki nozzles:
http://www.robotdigg.com/product/517/Nozzle+w/+Adapter+to+5mm+Hollow+Shaft+Stepper (http://www.robotdigg.com/product/517/Nozzle+w/+Adapter+to+5mm+Hollow+Shaft+Stepper)
(from a post on the Neoden thread)
Not sure if this works without tools (thumbscrew). Also no clue if it's really airtight. Given the shape of the nozzle's head I think you would need at least an O-ring or so.
Regards, Axel
-
@sedelman: Follow up on the nozzles, here's an adapter for Juki nozzles:
http://www.robotdigg.com/product/517/Nozzle+w/+Adapter+to+5mm+Hollow+Shaft+Stepper (http://www.robotdigg.com/product/517/Nozzle+w/+Adapter+to+5mm+Hollow+Shaft+Stepper)
(from a post on the Neoden thread)
Not sure if this works without tools (thumbscrew). Also no clue if it's really airtight. Given the shape of the nozzle's head I think you would need at least an O-ring or so.
Regards, Axel
Hi
One of the reasons auto-change nozzles work in the first place is that they really don't need to be airtight. You design the tubing and the pump size to accommodate the level of leaks you expect to get. Having fancy seals makes the change process much less reliable (the seals hang up and wear out). Bumping the tubing up one or two sizes is cheap. Routing the bigger tubes is a bit of a hassle that you would rather avoid, but you do it. The bigger pump (and bigger valves and ...) do cost real money. Most quick change setups make the same sort of decision for the same basic reasons.
Bob
-
Presuming that the big-boy machines also have pressure sensors to detect missing / failed parts, how do they compensate for a leaky nozzle? At least with my machine the sensor is so sensitive that it will usually signal a leak when I have a larger nozzle with a larger part, probably because neither the surface of the nozzle nor the surface of the part are completely flat.
Regards, Axel
-
Presuming that the big-boy machines also have pressure sensors to detect missing / failed parts, how do they compensate for a leaky nozzle? At least with my machine the sensor is so sensitive that it will usually signal a leak when I have a larger nozzle with a larger part, probably because neither the surface of the nozzle nor the surface of the part are completely flat.
Regards, Axel
Hi
Again, it's part of the design. Rather than using a super sensitive sensor with no calibration, they use something they can actually measure changes on. When the thing is in the "up and empty" situation they can (but may not always) calibrate the sensor. When they change heads from giant to next to no diameter at all, they can calibrate for that head. They might store that value or measure it each change. Storing it lest them detect the "oops missing head / wrong head" problem. Yes it's a bit more firmware. Yes it works better than the alternative. There are likely as many variations on how to do it as people doing it.
It still is not perfect, but no solution ever is. The smallest head does still set a limit on the leaks. You can't have more air through the leak than through the head. A screw down gasket fitting has a *much* lower leak rate than the flow through that small head. You can have a lot more than a 1x10^-8 cc/atm/sec leak and still detect a missing part just fine.
Bob
-
Ah, ok, thanks. Perhaps a bit too much to ask from my $4k machine ;-)
Regards, Axel
-
That seems to be the manufacturer's page, QiHe. I'm still in the market for a printer and a larger oven, can anyone comment on those offers:
Printer ($260): http://world.taobao.com/item/525291626153.htm (http://world.taobao.com/item/525291626153.htm)
Oven ($1050): http://world.taobao.com/item/525259351947.htm (http://world.taobao.com/item/525259351947.htm)
Addendum: Aha, the devil is in the details. They quote free shipping but only to mainland China. For overseas destinations it's 90yuan/kg (at least for the printer and oven, for the TVM802A it says contact the seller; at 60kg and this rate this would come to Y5,400 = $830)
Regards, Axel
Good day Axel,
I purchased one of those ovens (T-960) and overall it seems to be a pretty solid unit. Given its price my original plan was to install my own control hardware and simply reuse the rest... However, I did some basic tests and it actually works pretty good. That being said, I will be qualifying the oven just to be sure and can report back if you or anyone is interested.
Cheers,
Sam
-
Sam, very interested in this. Please keep us updated.
Regards, Axel
-
I added a few high-res photos of the p&p head with the cover removed:
http://servokit.com/blog/2016/03/14/a-closer-look-at-the-pick-and-place-head-of-the-tvm802a/ (http://servokit.com/blog/2016/03/14/a-closer-look-at-the-pick-and-place-head-of-the-tvm802a/)
Regards, Axel
-
I added a few high-res photos of the p&p head with the cover removed:
http://servokit.com/blog/2016/03/14/a-closer-look-at-the-pick-and-place-head-of-the-tvm802a/ (http://servokit.com/blog/2016/03/14/a-closer-look-at-the-pick-and-place-head-of-the-tvm802a/)
Regards, Axel
Hi
The pressure pump either is a blow off / clean pump (low pressure) or it's a bit higher pressure to run an optional dispensing head. You see applications where adhesive is dispensed to hold down big parts in addition to the ever (un) popular solder dispensers.
Bob
-
No dispensing head, the pressure is far too low, must be blow-off. So far each part came of voluntarily...
Regards, Axel
-
No dispensing head, the pressure is far too low, must be blow-off. So far each part came of voluntarily...
Regards, Axel
Hi
One of the "features" you need with a blow off pump is a "scrap cup". Often this gets overlooked in the first pass design. Blowing stuck parts off randomly all over the machine is *not* a good idea. Since you can't reliably release the part, placing it in the proper location is highly unlikely. The right approach is to move the head over the cup and "spray" the part into it. Designing the cup so the part does not bounce out is a bit interesting. (Hint: think of a milk jug as a good design model and a dinner plate as a really rotten one).
Bob
-
The machine has one of those; it's at the end of the tray and when the pressure sensor detects a dangling (or missing part) the head travels there and drops the part. Could be that the machine actually blows at this moment to make sure the part does indeed come off, difficult to tell because it happens really fast. Re. the design, the cup is formed like a ramp, apparently some thought went into this.
Regards, Axel
-
The machine has one of those; it's at the end of the tray and when the pressure sensor detects a dangling (or missing part) the head travels there and drops the part. Could be that the machine actually blows at this moment to make sure the part does indeed come off, difficult to tell because it happens really fast. Re. the design, the cup is formed like a ramp, apparently some thought went into this.
Regards, Axel
Hi
The normal process when it heads over the cup is to give the thing a blast of reverse pressure. It does not take long at all.
Bob
-
More tests and observations re. max. usable part height:
http://servokit.com/blog/2016/03/16/more-about-the-tvm802a-maximum-part-height/ (http://servokit.com/blog/2016/03/16/more-about-the-tvm802a-maximum-part-height/)
Also, we've finalized and ordered our PCBs; I hope to have the first prototypes sometime next week for some more realistic tests.
If you read german, we have a thread about this machine over at:
https://www.mikrocontroller.net/topic/371594 (https://www.mikrocontroller.net/topic/371594)
Regards, Axel
-
So it can place 10mm tall parts if you use the part trays + your other components that are on the PCB are no taller than 4mm?
Plus a 10mm part has to be raised up 15mm from the base of the machine to get picked up properly correct.
Good info to know.
-
So it can place 10mm tall parts if you use the part trays + your other components that are on the PCB are no taller than 4mm?
Plus a 10mm part has to be raised up 15mm from the base of the machine to get picked up properly correct.
Good info to know.
Hi
One issue with large parts is likely to be weight. In some cases bigger is also heavier. At some point a given size nozzle simply runs out of "pull" to reliably lift the part and control it while it is moved. On a machine with only two nozzles, there may not be much use for the giant nozzle on other parts. With no fast change capability, it may be faster to hand place a single large part and let the nozzles each handle a range of things that covers everything else.
Bob
-
With stock s/w it CAN'T place anything higher than 5mm.
With custom s/w it should be possible to place parts up to 10mm from tape and tray.
I too own TVM802B.
I used it to assemble around 160 boards with total component count around 15000.
I'm not very happy.
It has multiple issues in h/w and s/w. I intend to fix all of those. I building new feeder, peeling mechanism. I want to replace the vacuum pump (not powerful enough), replace pressure sensors (it is probably mechanical and not very reliable). Possibly steppers drivers in future. I also want to build a nozzle changer.
We already started taking apart their network protocol. After just 3 days we can control almost all binary functions (like turn on and off pump etc) and actual movement is a matter of couple more days. To get to the point of having minimally usable s/w will probably be another month or two.
-
AR_Systems do you have any video you can share of your machine running?
-
Yes, I do... I just need to upload it.
Btw, to those that bash belt driven machines... vs closed loop with encoders or with lead-screw drive. From my experience the belt system even on this machine is sufficiently reliable. Judging by the lack of any kind of drift of pick up locations, I can say there is essentially no missed step and therefore no loss of accuracy - from the movement.
-
ServoKit ... Very nice write up on your blog.
ar__systems can you comment on the reel feeder system and how it worked over 15K part placements? Do you have more details or even a Github repository with the progress you have been making on the network protocol?
I have one of these machines heading my way. Was also wondering if a vibration feeder can be added... (Not sure how the software would deal with the reel position)
-
ar__systems can you comment on the reel feeder system and how it worked over 15K part placements? Do you have more details or even a Github repository with the progress you have been making on the network protocol?
I have one of these machines heading my way. Was also wondering if a vibration feeder can be added... (Not sure how the software would deal with the reel position)
vibration feeder would be easy to add even with stock s/w. you just define a tray 'stack' with size of one, and it will pick up parts from the same location every time.
Reel feeders on this machine SUCK. The main problem is the peeling mechanism. It just can't be made to work reliably. It constantly changes from pulling on the cover tape too much or too little. I mean, if you have only one or two parts it is fine, you can set it up. I used almost all of the 8mm feeders. It does not happen on all of them at the same time, but there were several that just would not work consistently. If it pulls too much it moves the carrier tape! It needs constant supervision. Other problems is the feeder block itself. the metal part that holds the tape in place is 1mm steel laser cut into shape. It scrapes some black plastic tapes and the shavings get in the way of the pick up. I would unscrew it and clean it up, but all the screws (30 of them!) are locktited and would not move. Also, it does not like clear plastic tapes - I had several of those with bigger ceramic caps. Because the tape is only held at the ends of the feeder channels and those plastic tapes are not very rigid, sometimes instead of pulling the tape through it just mangle it out of shape, bending it into upside down U. Naturally it creates a lot of mess.
The cover film from the feeder is spooled onto the smaller reels. it is routed over a rod. Works ok for 8mm tapes, but for one of my 16mm tapes the cover tape was kind of sticky and was sticking to the rod. We tried to twist it but then the head catches onto it as it moves. Oh well, we ended up wrapping a small roll of paper around the rod :) fixed the problem.
So there are dozens of small oversights and the result is that machine requires basically constant supervision to fix mispicks, peeling issues, and what not.
Also, don't run the s/w on W10. the docs says XP is not supported but we actually ran it on XP and it worked fine. On W10 it has some connectivity issues resulting in some commands not executed properly. I first attributed it to trying the machine over wifi, but no. I ended up with a broken nozzle and bent pick up needle as a result of trying w10.
-
AR_Systems Excellent info!
Please post video and pictures of what ever you can about this machine. I was on the fence about buying one.
-
Also, don't run the s/w on W10. the docs says XP is not supported but we actually ran it on XP and it worked fine. On W10 it has some connectivity issues resulting in some commands not executed properly. I first attributed it to trying the machine over wifi, but no. I ended up with a broken nozzle and bent pick up needle as a result of trying w10.
Thanks for the tips... I was heading the W10 way. :phew:
How are you finding the nozzles? I see there are adapters for Juki nozzles that are much easier to find...
How are you finding the Vision system?
Sorry for all the questions.... :-+
-
Regarding the nozzles, mine actually came already equipped with Juki nozzles (well... probably not original :) ). I'm surprised to see here photos of the machines with some yellow metal nozzles. What brand are those?
I recommend buying some spares from the start.
Vision system is a must for PNP. Although s/w overall looks like it was written by a student over a weekend, the visual system actually works consistently reliably. I don't remember a single issue with it. Ok there is one - looking for fiducials it can get confused by the presence of other things in the field of view, like silk screen or traces or holes. So it is better to keep area around fiducials free of anything else.
-
Regarding the nozzles, mine actually came already equipped with Juki nozzles (well... probably not original :) ). I'm surprised to see here photos of the machines with some yellow metal nozzles. What brand are those?
The metal nozzles are the same as what Neoden uses. Its hard to tell where things originated when products are being developed side by side... ^-^ They are fitted on the TVM802A before the B was developed. Adapters are available but might be fun to press-fit.
-
I am very interested in the network protocol. :-BROKE
It was one of the main reasons I picked this machine. Possibility of writing custom software for it opens lots of doors even if its just used for special movements.
AR_Systems, do you have any thoughts on solutions for the feeder issue?
-
Again, it's part of the design. Rather than using a super sensitive sensor with no calibration, they use something they can actually measure changes on. When the thing is in the "up and empty" situation they can (but may not always) calibrate the sensor. When they change heads from giant to next to no diameter at all, they can calibrate for that head. They might store that value or measure it each change. Storing it lest them detect the "oops missing head / wrong head" problem. Yes it's a bit more firmware. Yes it works better than the alternative. There are likely as many variations on how to do it as people doing it.
It still is not perfect, but no solution ever is. The smallest head does still set a limit on the leaks. You can't have more air through the leak than through the head. A screw down gasket fitting has a *much* lower leak rate than the flow through that small head. You can have a lot more than a 1x10^-8 cc/atm/sec leak and still detect a missing part just fine.
Bob
Nice theory and I wish things were done that way, but the reality is that for TVM802A/TVM802B the pressure sensor is a dumb mechanical pressure switch that can't be adjusted (I mean not without disassembly). The pressure status transmitted to the PC is ***ONE BIT***. It works accordingly poorly. Often times it does not feel the part is on the nozzle and throws it out, even though it looks absolutely fine. It especially hates ceramic caps. But also some huge parts that were sitting absolutely solidly on the nozzle. So I'm guessing the switch is not very consistent. I had to disable pressure for several components because it was just discarding parts over and over and over.
-
Again, it's part of the design. Rather than using a super sensitive sensor with no calibration, they use something they can actually measure changes on. When the thing is in the "up and empty" situation they can (but may not always) calibrate the sensor. When they change heads from giant to next to no diameter at all, they can calibrate for that head. They might store that value or measure it each change. Storing it lest them detect the "oops missing head / wrong head" problem. Yes it's a bit more firmware. Yes it works better than the alternative. There are likely as many variations on how to do it as people doing it.
It still is not perfect, but no solution ever is. The smallest head does still set a limit on the leaks. You can't have more air through the leak than through the head. A screw down gasket fitting has a *much* lower leak rate than the flow through that small head. You can have a lot more than a 1x10^-8 cc/atm/sec leak and still detect a missing part just fine.
Bob
Nice theory and I wish things were done that way, but the reality is that for TVM802A/TVM802B the pressure sensor is a dumb mechanical pressure switch that can't be adjusted (I mean not without disassembly). The pressure status transmitted to the PC is ***ONE BIT***. It works accordingly poorly. Often times it does not feel the part is on the nozzle and throws it out, even though it looks absolutely fine. It especially hates ceramic caps. But also some huge parts that were sitting absolutely solidly on the nozzle. So I'm guessing the switch is not very consistent. I had to disable pressure for several components because it was just discarding parts over and over and over.
Hi
That's the whole point, If you do a more exotic machine, a *lot* of things change. It's not just a "put this on and you are done" mod. The whole thing has to work together. The price goes up in big steps *not* just because of the accessory you add. It goes up at least as much for all the other stuff you have to mess with to get it to work right.
Bob
-
Not sure what you mean.... I've read entire Neoden-4 thread over the last few days. Many comments are like "oh you bought cheap machine, what do you expect". Here is what I think.
When I buy cheap car, I don't expect it to be on par with 200K jaguar. It might not be as fast, it might not have 4WD, it might have crappy interior. But I sure as hell expect NOT to have to get out of the car every 100 meters to do some tweaking under the hood.
Yes we buy cheap machine and we already compromise on a lot of things. It is slow, it does not have conveyor, it does not have nozzle changer, it can't mount 0201 etc etc. But I'm kind of expecting it to be more or less reliable. it is not.
Well ok, at least the amount of money I paid for mine is about how much a DYI kit to build a PNP would've cost me.
And yes on this machine several things can be changed that will make significant improvement even with stock s/w.
-
do you have any thoughts on solutions for the feeder issue?
I don't have any ideas for quick and easy fix :( I'm planning completely replacing feeders together with the peeling mechanism. I plan to use one dc motor per reel to peel off the cover tape. I don't believe the existing single shaft system can be made to work. Although I spoke to an engineer today and he said that if we replace rubber rings in those spools with bronze it might be made to produce more consistent friction. But I'm not holding my breath.
-
In the end it all needs to play nice.
Not sure if this is possible with the TVM802X’s s/w but at work I have written proxy app’s that basically sits between hardware and 3rd party software. The proxy app would just listen on the machine port and forward the data to the 3rd party software.
This opens the door to injecting messages to and from the machine and even transforming them.
An example would be to add better pressure sensors and get the proxy app to spit out the appropriate comms to the 3rd party s/w to place or dump a part.
-
do you have any thoughts on solutions for the feeder issue?
I don't have any ideas for quick and easy fix :( I'm planning completely replacing feeders together with the peeling mechanism. I plan to use one dc motor per reel to peel off the cover tape. I don't believe the existing single shaft system can be made to work. Although I spoke to an engineer today and he said that if we replace rubber rings in those spools with bronze it might be made to produce more consistent friction. But I'm not holding my breath.
How about commercial reel dispensers? Like the Juki, Panasonic and so on…. I know they will be much more expensive but in my case I could add a few at a time. That would increase productivity over time.
-
Not sure if this is possible with the TVM802X’s s/w but at work I have written proxy app’s that basically sits between hardware and 3rd party software. The proxy app would just listen on the machine port and forward the data to the 3rd party software.
Should be possible. You can direct the S/W to talk to a specific port on specific IP, which can operated by your application. Not sure if it is a good approach. You need to build in a lot of intelligence into this application. It needs to know what the original s/w is intending to do, just from observing it's actions. I think it is just easier to write an entire application and scrap the original one altogether.
-
How about commercial reel dispensers? Like the Juki, Panasonic and so on…. I know they will be much more expensive but in my case I could add a few at a time. That would increase productivity over time.
I did not look into those. Might be impossible to fit mechanically. The pickup location on the TVM802A/B is about 10-15cm inside of the perimeter of the machine, and the vertical clearance is only about 20mm. I only briefly looked at various big machine feeders and I don't think they would fit without major modifications to the TVM.
-
@rwb, Bob:
I think the 10mm capability is there but really just theoretically as long as you can't tell the software to navigate around obstacles. I'm mostly interested in caps of the 5.8mm variety and those shouldn't be a problem. Re. the weight of parts: The #3 nozzle does lift reliably an ESP8266-12F. That's a fairly large part (24x16mm, 2-3grams).
Re. W10: I'm running the machine on Windows 7 Home, 64 Bit with 8GB. So far no issues. This is a dedicated computer that does nothing else.
Re. nozzles: It seems that the 802B is different in this regard, photos on Ali show them with Jukis but they don't seem have the quick change. Maybe ar_systems can clarify this.
Vision system is a must for PNP. Although s/w overall looks like it was written by a student over a weekend, the visual system actually works consistently reliably. I don't remember a single issue with it. Ok there is one - looking for fiducials it can get confused by the presence of other things in the field of view, like silk screen or traces or holes. So it is better to keep area around fiducials free of anything else.
My exact sentiment and experience.
I'm looking forward to our PCBs so that I can do some real tests.
Regards, Axel
-
Nozzles are easy to change just by pulling on them down. They are only held in place by the friction in o-ring between the nozzle and the yellow tube they are insterted into.
-
Not sure what you mean.... I've read entire Neoden-4 thread over the last few days. Many comments are like "oh you bought cheap machine, what do you expect". Here is what I think.
When I buy cheap car, I don't expect it to be on par with 200K jaguar. It might not be as fast, it might not have 4WD, it might have crappy interior. But I sure as hell expect NOT to have to get out of the car every 100 meters to do some tweaking under the hood.
Yes we buy cheap machine and we already compromise on a lot of things. It is slow, it does not have conveyor, it does not have nozzle changer, it can't mount 0201 etc etc. But I'm kind of expecting it to be more or less reliable. it is not.
Well ok, at least the amount of money I paid for mine is about how much a DYI kit to build a PNP would've cost me.
And yes on this machine several things can be changed that will make significant improvement even with stock s/w.
Hi
If you go back through the entire sub-thread that I was commenting on, the discussion was related to an auto changer for multiple nozzles for this machine. That generally involves a wide range of nozzles as well as the changer it's self. The comments were related to the various other parts of the system's hardware (and software) that might be impacted by simply bolting on a changer. The original comments started on a "that should only be a couple hundred dollars" sort of theme.
Bob
-
@rwb, Bob:
I think the 10mm capability is there but really just theoretically as long as you can't tell the software to navigate around obstacles. I'm mostly interested in caps of the 5.8mm variety and those shouldn't be a problem. Re. the weight of parts: The #3 nozzle does lift reliably an ESP8266-12F. That's a fairly large part (24x16mm, 2-3grams).
Hi
The other item that gets big are power inductors / transformers. Those tend to get heavy when they get big.
Bob
-
With a big nozzle it has no problem holding inductors or anything else.
-
With a big nozzle it has no problem holding inductors or anything else.
Hi
... then I guess all those parts I've dropped over the years must have been from another issue...
Bob
-
I tried it with the largest nozzle (#4) and little pieces of brass I have lying around. Pictured is the highest weight that will be picked reliably, it's about 4grams (1/8oz.). The larger one in the pic drops off. The Juki nozzles (even knock-offs) are perhaps better, i.e. air tighter.
Regards, Axel
-
I tried it with the largest nozzle (#4) and little pieces of brass I have lying around. Pictured is the highest weight that will be picked reliably, it's about 4grams (1/8oz.). The larger one in the pic drops off. The Juki nozzles (even knock-offs) are perhaps better, i.e. air tighter.
Regards, Axel
Hi
Which is not terribly surprising. You only have a finite suction area and a max of 15 psi over that area. Once the weight of the part exceeds the area times pressure ... the part falls off. It's just physics. The standard drill here is to always use at least one nozzle bigger than you think you need (go up one from the smallest one that will pick up and hold the part).
Bob
-
Larger Juki nozzles have rubber tips so they will hold even better.
Unfortunately due to a poor pump in the TVM we do not have 15 PSI differential. I measured the pressure of 560mmHg, and that's with no direct path for leaking.... This is why I want to replace the vacuum pump with a much bigger one. Not only this one creates only a weak vacuum, it takes a couple of seconds for it to reach even this level. But in a couple of seconds the part is already delivered to the destination so it is kind of late....
-
Servo, keep in mind just lifting is not enough. It needs to withstand lateral acceleration without shifting or being dropped.
-
Yup, I tried with our heaviest and largest part, 16x24mm, 1.45g precisely. No problems.
Regards, Axel
-
How does the Vacuum pump on this unit look?
-
I've opened the machine (will post pics later); from a quick glance there are two pumps marked "Hargraves", this seems to be the mfg: https://www.hargravesfluidics.com/ (https://www.hargravesfluidics.com/)
Regards, Axel
-
Haha... I was wondering about adding a second boost pump. But I take it the one is vacuum and the other is blowing.
This might be an option for a boost pump or replacement... https://www.sparkfun.com/products/10398 (https://www.sparkfun.com/products/10398)
-
Ok, the first pic from the far side of the moon.
From left to right:
USB I/F board (outside frame to the left)
LAN I/F & controller board, white ribbon cable goes to front panel of machine
Driver board (attached to underside of work plate)
Blower pump (black plastic cuboid)
Two vacuum pumps (blue, mfg. Hargraves)
Power supply (outside frame to the right)
The wiring looks reasonably clean, plenty of air in the box.
Addendum: Those pumps seem to be the Hargraves BTC-IIs double head; not cheap, price starts at $240 a piece.
Addendum 2: The driver board is double-sided, the actual driver ICs sit on top, using the work plate as a heat sink. The board has seven drivers: 1 x, 1 y, 2 feeder, 2 nozzle rotation, 1 nozzle up/down. Apparently no drivers in the head. The drivers seem fairly beefy, definitely not A4988-kind of stuff but not M542-level either. The board also supplies the pumps.
Addendum 3: Both cameras are connected directly to the USB I/F board, apparently the controller only switches the camera illumination.
Regards, Axel
-
The controller.
MCU is STM32, Cortex ARM 3, smaller chip is TI DP83848, Ethernet IF.
Regards, Axel
-
Main point of the peek inside the machine was to establish a safe area where I can drill mounting holes to fix up stuff. The power supply, the two vacuum pumps, the blow pump, and the motor driver board are attached to the underside of the work plate. That leaves the reddish area in the photo. Trays outside this area will have to be fixed with double-sided tape etc.
Regards, Axel
-
More pics.
#1: Vacuum pumps, seen from the right side of the machine; not sure if it's one pump per nozzle or the two are in series (or whatever that's called with pumps), will have to examine head more closely.
#2: USB I/F and controller at rear of machine; white ribbon cable goes to buttons on front panel, the left white connector is camera illumination and solenoids, the right white connector is limit switches, colored ribbon cable goes to driver, not visible is Ethernet jack behind white ribbon cable.
#3: Another view of the pumps, yellow/green is power for pumps
Regards, Axel
-
This only reinforces my initial impression of relatively solid build. I mean they use quality parts and components, but somehow the result is still underwhelming...
The boost pump from sparkfun is even weaker than the one they use.
I intend to attach something like this:
http://www.globalindustrial.ca/p/hvac/service/chemicals/mastercool-90066-2v-110-6-cfm-vacuum-pump-two-stage?infoParam.campaignId=T9F&gclid=CIy86MapyssCFQwPaQodPCkH-w (http://www.globalindustrial.ca/p/hvac/service/chemicals/mastercool-90066-2v-110-6-cfm-vacuum-pump-two-stage?infoParam.campaignId=T9F&gclid=CIy86MapyssCFQwPaQodPCkH-w)
-
I intend to attach something like this:
http://www.globalindustrial.ca/p/hvac/service/chemicals/mastercool-90066-2v-110-6-cfm-vacuum-pump-two-stage?infoParam.campaignId=T9F&gclid=CIy86MapyssCFQwPaQodPCkH-w (http://www.globalindustrial.ca/p/hvac/service/chemicals/mastercool-90066-2v-110-6-cfm-vacuum-pump-two-stage?infoParam.campaignId=T9F&gclid=CIy86MapyssCFQwPaQodPCkH-w)
Correct me if I am wrong but those are very loud to run? (Same as a normal compressor)
How about having an inline air-tank but for negative pressure...
-
Yes, kind of loud, but nowhere near as loud as my run-of-the-mill compressor. Still I keep it in another room. I run a small tube to the room where I work.
In-line tank - not sure. Might work too.
-
Update on the driver board: Someone at mikrocontroller.net remarked that it looks a lot like the one in the Neoden TM220/240.
Regards, Axel
-
I've added mounting holes to the base plate for attaching trays and the like.
Step-by-step: http://servokit.com/blog/2016/03/19/adding-mounting-holes-to-the-tvm802a-base-plate/ (http://servokit.com/blog/2016/03/19/adding-mounting-holes-to-the-tvm802a-base-plate/)
Regards, Axel
-
Axel, do you have some photos of the reel feeders and how they work?
You also mentioned that the controller board looked the same as Neoden's, what model was that? Wondering because the lower priced models seem to not have vision and ethernet.
-
@protoneer: Motor driver board is apparently like TM220/240, not controller board. Also, the controller on the TVM doesn't handle the vision, the cameras are directly connected to the PC. Re. feeder, it's really very simple: all the motors do is peeling the cover tape, description of process is on my blog post about the head:
http://servokit.com/blog/2016/03/14/a-closer-look-at-the-pick-and-place-head-of-the-tvm802a/ (http://servokit.com/blog/2016/03/14/a-closer-look-at-the-pick-and-place-head-of-the-tvm802a/)
Regards, Axel
-
Thanks.. I have studied all your blog posts.. :)
But I am not sure about the motor that turns the plastic cover collector wheels. Does it have some clutch system for each reel? Tension on the wheels seem to be hard to control...
-
There are actually two motors turning the rod with the wheels that wind up the cover tape. The wheels themselves rotate independently from each other but they don't rotate freely, you have to apply some force to do it, it's a bit like an overload clutch. Yesterday, while playing with manual mode, I accidentally engaged the stripper motors without advancing the tape and the respective wheel didn't move. How well this works when more than one reel is attached remains to be seen. Have to wait for my PCBs.
If there are problems like ar_systems mentioned they might come from the fact that the load / tension on the rod is distributed asymmetrical, e.g. when only using eight feeders on the lower end. It might help to only use every other feeder position or so in such a case, to distribute the load across the rod.
Regards, Axel
-
Today I finally run a test with an external vac pump. I'd like to say the result is better, however with many factors at play it is difficult to say with certainty that it was due to the better vacuum.
I first ran a base line test with the built in pump populating 320 components at 80% speed and then another 320 at 100%. 100% actually came out more decent. In the 80% test 3 were placed on the sides and a couple were rotated almost 90 degrees. 100% did not have any such defects.
So far I run only 160 parts with external vac at 80 speed. All parts were placed ok.
Now, I noticed something else. There is a systemic error in component placement. I mean, consistency is quite good, definitely good enough to place 0603 without help of a camera, but the coordinates are messed up.
I ran a total of 5 test boards, each with 160 components split into 8 sections. The attached picture shows section 1 and 2 of the board I placed today. As you can see section 1 is all shifted down. And it is like that on all 5 boards. But not all of the components are shifted down. Section 7 (not shown) is shifted up. And other sections all shifted a bit in different directions. Looks like entire board was slightly rotated.
So, just like with Neoden, looks like the software is not capable of doing basic math correctly :(
-
Could be a resonance effect. I have this at 30%, the machine shakes a bit and the motors make funny noises, at 40% it's gone. The table also plays a role here.
Best way to test coordinates: Create an array of components (right click parts list for context menu, it's in there somewhere) and place parts 10mm apart in x, with y constant. Then put .5mm divided ruler (or printout) along x and let it run.
Also: Does it make a difference which feeder bank you use?
Regards, Axel
-
No. Original tests with built in pump were done with the left banks, and today I used north banks. The shift is the same.
I have the table bolted to a wall, so it is quite stable.
Your test is only good to gauge consistency, and I already know it is quite good. It screws up coordinates when matching fiducials locations to the placement data.
-
Created a 3D printed rail to attach tapes to the work area and have the nozzles pick parts directly from the tape. Preferable to taking parts out of the tape and putting them into a tray, I think.
Write-up: http://servokit.com/blog/2016/03/22/adding-a-tape-holder-to-the-tvm802a/ (http://servokit.com/blog/2016/03/22/adding-a-tape-holder-to-the-tvm802a/)
Regards, Axel
-
That's pretty cool Axel.
I had been thinking about how long it would take to load a custom milled CNC tray but your way allows you to pre-program the part spacing and then just pull load the part strip and pull back the clear tape which has to save a good amount of time vs hand placing parts.
I think this is a excellent way to go.
You just have to make sure the tapes are loaded properly spacing wise so the head picks the parts at the right spot if the parts are small. No so much a problem with larger parts since you have vison.
Can't wait to see how your first runs go and if any of the issues AR_Systems has experienced also happen to you.
-
Hi Axel,
No need to lathe some brass for the reel locating pin, just use 1.5mm ferrule - mine is blue in color.
(http://static.rapidonline.com/catalogueimages/Module/M063941P01WL.jpg)
You can not use only two pin on the edge as the tape actually shrink or stretched in use, you will need several locating pins along the tape.
Also with this kind of feeder, carefully set the pick height of the nozzle - if the nozzle push the part too hard then many other neighboring parts will jumps over their pocket, like fleas :D.
-ichan
-
Thanks, ichan, good idea - I like to work on the lathe, though...
Regards, Axel
-
@rwb: Good point re. the spacing. The parts that are candidates for the tape fixture, like the MOSFETs in the pic and our TSSOPs, are compartmentalized in their tapes. The tapes have little grooves on the underside between the compartments which can be used as a guide, just have to add a counterpart to the rail.
Regards, Axel
-
We are almost done with the protocol. At this point we can control everything, the only thing still left to figure out is setting the speed of the movements.
Looks like the stepper driver in the machine is the most basic one - no microstepping, unfortunately. That's the reason for horrible noises the machine produces at slow speeds - it basically comes to full stop in between steps. I really wish they at least support acceleration control. Will know next week.
-
The motors must run with microstepping: With the belt drive (diameter of pulley c. 16mm = circumference 50mm) and the standard 200 steps/revolution we would only have a resolution of 0.25mm...
Regards, Axel
-
How well does these machines deal with Micro's that are angled 45 degrees?
I take it the vision will ensure that the micro is straight first and then it will be turned into the 45 degree angle....
Also how well are these machines at placing 0.5mm pitch micro's? I am looking at using it for 100pin 0.5mm pitch micro's if possible...
(My machine is close but still stuck with customs :)
-
The rotation happens directly after picking the part, before alignment, I've written about it earlier in the thread. 0.65mm definitely works. With larger parts keep in mind that the vision system has a limit of 20x20mm.
Regards, Axel
-
The rotation happens directly after picking the part, before alignment, I've written about it earlier in the thread. 0.65mm definitely works. With larger parts keep in mind that the vision system has a limit of 20x20mm.
Thanks Axel, the micro's I am looking at are 14x14mm and at 45 degrees its just under 20mm(Opposing corners).
-
Once you get your machine, please report the software version so that we can keep track of it (no download section on the mfg. web page). I'm on V2.1, date 26.11.2015.
Regards, Axel
-
Will do...
-
Happy to report that I did the first run with our prototype PCBs today. So far only resistors and stuff because I haven't finished the tray setup yet. No issues with the feeders, perfect locating of fids (I've used Lady Ada's fids for Eagle, one in bottom/left, other in upper/right corner). Will post video tomorrow or so, once I'm satisfied with the tray layout.
Regards, Axel
-
@Axel I receivedthe TVM802A yesterday and the software version in the CD are "English-20150928-V2.08" and "English-20151126-V2.10(Trial Version)", have you tried both? Any noticeable diference?
My machine is still in the box as I'm moving it to a new location...
Cheers,
Paul
-
Same here, so far I didn't notice any difference.
Regards, Axel
-
Follow-up on the microstepping: To move the crosshair across a fiducial (dia. 1mm) with the cursor keys in "Low speed" requires 32 clicks. With .25mm movement per step that comes out at 8x microstepping.
Regards, Axel
-
Ah, you are correct. Well ok, then there is hope :)
-
Hey folks. Sergei from Fotokite here. Just got to finally play with our TVM802B.
Background: We got it to do small proto builds of the Fotokite Phi and various new Pro products (anything delivered is professionally assembled, but for internal testing this just takes too long). Big thanks to Axel/Servokit for his in-depth blog review of the machine that convinced us to take the jump. I've done lots of hand soldering, stenciling, and oven baking before for prototypes, and have visited CMs in various places including Shenzhen, but never used a pick&place myself. New experience! :-)
Ordering/shipping was *fast*. Paid with paypal, machine shipped next day, arrived in 6 days (ordered Thursday, received Wed). Built quality looks good, only complaint is a) the CD and b) holy crap, spelling mistakes.
Same software versions as electrfunch.
I'm still going through the docs but from initial runs I think something is weird with the Y coordinate system: when going through "measure" and "goto stacks" it goes to the right places, but when running the example program it runs into the Y- limit at full speed. Perhaps there is a coordinate offset I haven't discovered yet. Also weird: it can detect the small nozzle in the calibration but doesn't seem to find the big one -- if I play with thresholds it'll find the large one but not the small one. Not sure here :(.
There is a pressure sensor (nice touch), but by default the PnP program doesn't use it. Not clear what its purpose is.
It's loud as hell, both in movement and when the pump is on.
Boards / stencils arriving next week along with reels, hope to test it in action soon..
-
Sergeil Thanks for sharing your experinces with this machine also.
Can you tell us or show us the kind of boards your going to be making on the machine?
I'm interested in how well this machine can place .5mm pitch IC parts. Do you plan on running any of these?
-
We've got lqfp64 (.5mm pitch), lots of 0402s, and I think smaller leadless, including several parts with hidden pins and and a USB-C connector with hidden pins. We've been able to solder everything by hand with a good hot air rework station, microscope, patience and some tears |O . Around 300 components per board, most on one side, we plan to place around 270 with the PnP, rest (singles/easy ones) by hand. Got a stencil setup + IR oven from China as well - not worried about oven but stencil setup needs improvement. Anyway hopefully a smoother process than tweezers + microscope + tears :)
Would love to get hints regarding the Y axis from the "veterans", though maybe it's a matter of clicking around the software...
-
There is a pressure sensor (nice touch), but by default the PnP program doesn't use it. Not clear what its purpose is.
It should be used to detect a mis-pick, so it can retry
-
Sweet, can't wait to see how well the machine handles those smaller and fine pitch parts.
How do you plan on placing 270 parts when the machine only holds 30 or so reels + the trays?
I also have designs with more than 100 parts so it's ruff that these machines only hold 30% of the required parts to get the whole board completed.
How soon do you plan on running the first board?
Dude your drone looks sweet!
(https://www.eevblog.com/forum/reviews/pick-and-place-machine-tvm802a-tvm802b/?action=dlattach;attach=211973;image)
-
Thanks. 300 actual physical components, but around 60 types of parts. The TVM802B has 46 reels + whatever trays. Too late to look at BoM now but if I recall correctly doing PnP on the top-occurring 25 or so gives us 270 components placed.
That's actually the Fotokite Pro, this will be used for the next-gen Fotokite Phi build.. Don't wanna turn this (very useful) thread into a Fotokite discussion, but if you want to learn more check the (now long finished) campaign page: http://igg.me/at/fotokite (http://igg.me/at/fotokite)
Actual first build expected Tue/Wed. If anyone wants to give me some hints or tell me what the heck is up with the example program/Y axis... would be most grateful. P.S. we have 2 Ultimaker 3D printers and that's my plan for both securing our (non-rectangular, non-panelized) PCB in the machine and for making trays.
-
Sweet! Looking forward to seeing how the build goes.
How many boards do you have to run?
-
Hi sergeil, welcome to the club!
Pressure sensor is for dangling / missing parts. You have to enable it in the Machine setup (last tab) and for the individual part in the parts list. If the sensor then detects a "leak" (because part is dangling or not there) it drives to the dump at the right of the tray and drops the part. It then tries 5x to get the part right, after that an alert comes up. Sensor is very sensitive, would only work reliably for smaller parts on smaller nozzles.
Re. nozzle calibration: Make sure to switch of ceiling lights etc., the up-looking camera is very sensitive. Possibly a 802B thing because the camera is at a different position. Also could be the Juki nozzles.
Re. y-axis. Not sure about that. If it's only with the sample data it's probably a setup issue (look at machine setup and make sure that the feeder positions are at least sensible; look at part list while placing: It highlights done parts so you can narrow it down) or again specific to 802B.
Re. noise: Movement noise comes with some combinations of overall speed and individual part's speed, especially a lower speeds. Let it run faster. Also the round alu bar on which the reels sit tends to vibrate.
Ultimaker here, too. Good man.
Regards, Axel
-
I also have designs with more than 100 parts so it's ruff that these machines only hold 30% of the required parts to get the whole board completed.
..and is there really no scope to do some engineering to significantly reduce that BOM?
-
First real production run with the TVM:
https://youtu.be/gqUc85uAubs
We don't have panels yet, instead I'm using four separate PCBs. Fiducials in lower right corner of each PCB (one fid detection takes a little longer because I accidentally smeared glue over it...).
For some reason the two caps (at lower edge of board) are off by the same amount on all four PCBs. Not sure what the reason is, could be the feeder setup or an error in the exporter ULP. - Update: The tape was a little off in the feeder, half a sprocket diameter. Dang.
Took me four weeks to get to this point but now I'm a happy camper.
Regards, Axel
-
I also have designs with more than 100 parts so it's ruff that these machines only hold 30% of the required parts to get the whole board completed.
..and is there really no scope to do some engineering to significantly reduce that BOM?
I just closely followed the EVM designs in hopes of keeping the same performance I get when testing the EVM boards. At first I thought nothing of the part numbers per board until I started looking at the possibility of buying one of these smaller PNP machines.
-
Axel looks good.
I'm surprised you not using Vision on the IC parts. Looks like its doing a great job.
Is that all the parts you need to place on your board or is this just a quick test.
-
With the bespoke trays, vision is no longer needed (except for the fids). Huge speed-up. Missing at the moment are just three small parts (SOT size, 8mm tapes) which we don't have on reels yet and the TTH stuff, obviously. The machine table is now bolted to the (brick-)wall, so even at 100% nothing moves or shakes.
As for your boards, you might have to daisy chain machines; even the larger, still affordable, Mechatronikas only go to 64 feeders or so.
Regards, Axel
-
sergeil please, where have you bought yours ? You can reply in private if you wish.
Thanks for the videos and updates guys.
-
Whoa looks great. Can confirm, shake is an issue: ours is on what we thought is a sturdy table, but still moves all over the place.
Update on Y-axis: i did rough manual alignment of the fiducials, resulting in a Y offset of ~300. Once this was done I was able to run the example program (including machine vision fiducial alignment). *However* turns out my CSV file has all sorts of nasty characters (unicode encoding from Chinese?) which I had to fix -- otherwise program crashes. In some ways it's the perfect setup: documentation is there, but broken, but good enough to point in the right direction and force you to understand exactly what's going on :-)
@Axel, any chance you can share top-level specs for the trays (hole pattern, pickup height?)? Did you take any special precautions for ESD for the trays? Have you been able to get successful reliable pickups from cut tape in your custom cut tape trays?
I got the machine via Aliexpress, http://www.aliexpress.com/item/Small-SMT-Machines-Pcb-Assembly-Small-SMT-Pick-and-Place-Machine-With-Camera-TVM802B/32499850720.html (http://www.aliexpress.com/item/Small-SMT-Machines-Pcb-Assembly-Small-SMT-Pick-and-Place-Machine-With-Camera-TVM802B/32499850720.html) (i'm in no way affiliated with the store and don't benefit from referrals).
-
Nice video Axel, i wonder why all of this kind machine manufacturer had no will to make video like this - the video of their own machine making real things.
I think this machine will fill the gap between Neoden TM2xx and their N4 machine.
Have you try the machine with components with plastic tape? Like thick smd capacitor or SOT23 in 8mm plastic reel?
-ichan
-
Hi ichan, we will have thicker tapes (SOT23 & tactile switch) later this month or in April, will then make tests. As long as we are under 5mm, I don't think there will be problems. I think that TM2xx or any other machine without vision don't make much sense anymore, given that they are not much cheaper; you need vision at least for fiducials.
@sergeil: Will post drawings of trays and everything during the next days.
Regard, Axel
-
Nice video Axel, i wonder why all of this kind machine manufacturer had no will to make video like this - the video of their own machine making real things.
I also find this puzzling. The vast majority of the videos merely place 2 pin passives, one of the most forgiving components. :-//
Anyway, yes thank you very much for the videos.
-
Yes - and only one component type/size/config.
None show the machine placing a wide variety of components across the entire work area either - it's always the minimum travel distance.
I'm not saying these machines are not capable, but it would do wonders for buyer confidence to show this in good quality videos - that's for sure.
Nice video Axel, i wonder why all of this kind machine manufacturer had no will to make video like this - the video of their own machine making real things.
I also find this puzzling. The vast majority of the videos merely place 2 pin passives, one of the most forgiving components. :-//
Anyway, yes thank you very much for the videos.
-
Axel, you say vision is not longer needed on these boards due to your custom trays, but have you considered what happens when the board profile is slightly incorrect?
The 2 two processes are independent of each in manufacturing - board profiling and pad construction.
What is the finest pitch component your loading on this board? Would be great to see a close up image.
Do you always only ever place 1 fiducial on your designs?
Congratulations on the successful build - it's great to see your success after that week of patience practising!
With the bespoke trays, vision is no longer needed (except for the fids). Huge speed-up. Missing at the moment are just three small parts (SOT size, 8mm tapes) which we don't have on reels yet and the TTH stuff, obviously. The machine table is now bolted to the (brick-)wall, so even at 100% nothing moves or shakes.
As for your boards, you might have to daisy chain machines; even the larger, still affordable, Mechatronikas only go to 64 feeders or so.
Regards, Axel
-
Here's a close-up from this morning's run. This time with solder paste and all smd components. Soldering went fine except for the 8266 and MOSFET; I think the solder paste was still too cold and didn't reflow properly. Could also be the surface finish, maybe the soldering profile wasn't correct.
@thommo: The PCBs have two fids in opposite corners but are so small that one is sufficient. Don't understand the question re. board profile. All the vision does is rectify and center the the chip, the tray takes care of that. Parts can move and skew much more in the tape than in the tray.
Regards, Axel
-
Thanks for running this thread, Its helping me make a choice on which machine to buy in Early April. ITs a new year on 1 April for us. I've been able to take my list from 3 potential's to 2 today, The SmallSMT machines just dropped off the list for "non technical" reasons.
-
What oven did you use? I have a few T962's and had the same issue... I solved it by doing a single empty warm up cycle before each run of boards.
Updating the firmware on these ovens also made a it run much better with very little to no cold spots.
-
[quote ]
@thommo: The PCBs have two fids in opposite corners but are so small that one is sufficient. Don't understand the question re. board profile. All the vision does is rectify and center the the chip, the tray takes care of that. Parts can move and skew much more in the tape than in the tray.
Regards, Axel
[/quote]
How many Fids can be specified per sub panel? Would one need more than 2?
-
Nice video Axel, i wonder why all of this kind machine manufacturer had no will to make video like this - the video of their own machine making real things.
And other weird things.. Its the first "Chinese" company I've had any dealings with that would make me to go Germany to see their machine. It doestn make any sense at all.
-
@protoneer: You can specify two fids per PCB, don't know if more would be of any use. At least with our smallish boards using the opposite fid wouldn't improve placement. Obviously with large panels (like ichan's) YMMV. What I DID notice was that ENIG boards (like the one in the attachment) have faster fid recognition than HASL. The HASL fids are shiny and a bit uneven, the vision algo doesn't like that.
Re. soldering, it was the paste, 2nd run was perfect (I'm using a pizza oven with the beta layout controller).
My gut feeling says .5mm pitch would work with the machine, I would expect problems coming from other areas like stenceling.
Regards, Axel
-
with the machine, I would expect problems coming from other areas like stenceling.
Regards, Axel
Never under estimate the importance of the solder paste, the solder stencils, getting your apatures right.. its the root of all evil.
-
Nice video. What speed was that running? Would be nice to be able to use two heads for all those passives then do a second job for just the modules, but changing the nozzle is too tedious?
Would really like to see how this does with qfn 0.5mm and such. Seriously considering one.
-
2 fids are sufficient for alignment. Most subcontractors usually specify 3, the third being used to verify that the panel has been inserted the right way round.
You sometimes see local fids around large fine-pitch parts, but not sure how common it is for these to be used
-
@sparkswillfly: 80% overall, the parts from the tapes are @100%, from trays @60%. In a real production environment you would place the 8266 by hand (the pitch is 2mm, don't even need tweezers for that)and use both nozzles for smaller parts but I was curious to see how the suction cup works.
Regards, Axel
-
@sparkswillfly: 80% overall, the parts from the tapes are @100%, from trays @60%. In a real production environment you would place the 8266 by hand (the pitch is 2mm, don't even need tweezers for that)and use both nozzles for smaller parts but I was curious to see how the suction cup works.
Regards, Axel
OK thanks. My ocd would want to use machine placement for the modules anyway. |O
-
with the machine, I would expect problems coming from other areas like stenceling.
Regards, Axel
Never under estimate the importance of the solder paste, the solder stencils, getting your apatures right.. its the root of all evil.
Hi
I totally agree with that.
So ... just to hijack this poor thread once more ...
What is a "good enough" stencil printer to run 0.5 mm pitch parts? Consider that it will get used with the TVM802x so printers costing more than the pick and place are out of consideration. Same issue also applies to the size of the stencils and type of stencils. There is no need for 1 meter square steel frame stencils in this case. The 0.5 mm point seems to be about as far as this machine will get (which is fine by me). There is no need for a printer that does 0.35 mm just fine, if the rest of "the production line" will not support that.
Bob
-
Regarding fiducials: the software supports one or two per PCB. They should be circles or squares, and should have good contrast. It has support for panelized PCB's but it's not exactly clear how that's supposed to work, since you can't specify which component belongs to which PCB (perhaps it uses the fiducials to set new offsets and repeats?).
Fun discovery: while the TVM802B officially supports 46 reels + 10 tray components, the software actually has the "left" and "back" stacks both defined up to #30. So you can define up to 30+30+10=70 places to pick up parts, though the left/back stacks will always go through a prick phase to try to feed tape.
Got to test out the camera alignment feature for simple SMT parts, seems to work well. That and pressure sensor makes a lot of sense, esp for poorly aligned 0402 parts.
Figured out a way to make a cut tape adapter work (it's basically a 3d printed slider with a mylar cover and an opening to pick up parts).. Working to modularize that to be able to quickly feed single parts.
Beginning to really like this machine. Nice touches like a cutout in the base plate for spent tape to go into. And it's pretty robust, so far.
-
What is a "good enough" stencil printer to run 0.5 mm pitch parts?
My last run (manual placement) I did with a frameless 150u stencil from Elecrow, not even using a paste printer, just some leftover PCBs taped to a wooden panel to form a frame for the PCB. Squeegee was a putty knife.
Paste print was just perfect (checked under the microscope).
To make a long story short: The printer IMHO does not make much difference, provided the alignment between stencil and PCB is good.
-
What is a "good enough" stencil printer to run 0.5 mm pitch parts?
My last run (manual placement) I did with a frameless 150u stencil from Elecrow, not even using a paste printer, just some leftover PCBs taped to a wooden panel to form a frame for the PCB. Squeegee was a putty knife.
Paste print was just perfect (checked under the microscope).
To make a long story short: The printer IMHO does not make much difference, provided the alignment between stencil and PCB is good.
I think the issue here is having a printer that gets reliable results every time with minimal fiddling.
Paste print quality is the biggest factor in getting good yield, so having a reliable pasting process is important.
-
Fun discovery: while the TVM802B officially supports 46 reels + 10 tray components, the software actually has the "left" and "back" stacks both defined up to #30. So you can define up to 30+30+10=70 places to pick up parts, though the left/back stacks will always go through a prick phase to try to feed tape.
Got to test out the camera alignment feature for simple SMT parts, seems to work well. That and pressure sensor makes a lot of sense, esp for poorly aligned 0402 parts.
Can you clarify on exactly how one would use up to 70 parts on this machine?
I see this machine can hold up to 46 reels and I know you can make custom part trays like Axel did but Where are you getting the 30 + 30 + 10 parts locations from? The ad below shows the machine with 46 feeders on the left and back, which I assume is 23 parts on the left and 23 on the back.
http://www.aliexpress.com/item/Small-SMT-Machines-Pcb-Assembly-Small-SMT-Pick-and-Place-Machine-With-Camera-TVM802B/32499850720.html (http://www.aliexpress.com/item/Small-SMT-Machines-Pcb-Assembly-Small-SMT-Pick-and-Place-Machine-With-Camera-TVM802B/32499850720.html)
70 feeders sounds really good!
-
I tried alignment pins the other day and that worked really well.
Basically I have a panel consisting out of 6 sub boards. The panel and SS stencil has alignment holes and I bought some alignment pins from Ali that I fixed to a flat plastic sheet. The pins were aligned and glued in place.
I do lots of small resistor arrays, this way of doing yielded a consistant and accurate result.
-
I tried alignment pins the other day and that worked really well.
Basically I have a panel consisting out of 6 sub boards. The panel and SS stencil has alignment holes and I bought some alignment pins from Ali that I fixed to a flat plastic sheet. The pins were aligned and glued in place.
I do lots of small resistor arrays, this way of doing yielded a consistant and accurate result.
Sounds nice. Do you have any pictures you can share of this setup?
-
Sounds nice. Do you have any pictures you can share of this setup?
Attached are a few photo's. Next version will have shorter pins.
-
The issue with the extra feeder locations is that the pickup location is always the same, other with the trays where the location changes with each pick, depending on the tray layout. I'm not sure if that's really useful unless you can make the prick mechanism work for you.
Re. PCB layout: For multi PCBs the SW assumes that all PCBs get the same parts. This is what happens in the video: R1 goes to PCB1, then PCB2, then PCB3 etc. Then the same for R2. Only way around this would be to concatenate multiple part lists into one but that would require to offset the part coordinates accordingly because the SW expects all coordinates relative to the PCB origin.
Re. the stencil alignment pins, http://www.eurocircuits.com/ec-registration-system (http://www.eurocircuits.com/ec-registration-system)
Regards, Axel
-
Re. the stencil alignment pins, http://www.eurocircuits.com/ec-registration-system (http://www.eurocircuits.com/ec-registration-system)
Two alignment pins should work better.
Have you considered laser cutting trays instead of 3D printing?
-
I would have, if I had a laser cutter... The only concern re. 3D printed trays is ESD. The material (PLA) is not conductive. Eventually I'll probably mill the trays from aluminium on my CNC.
Regards, Axel
-
As the feeder price of my LSV60 machine are sky high then the machine now has a full of custom tray feeders ;D
The easiest way to make it is just to drill a series of 1.5 mm holes on a thick plastic sheet and use some 1.5mm ferrule to hold the tape on position. What i use is 10 mm acrylic, and the holes is made by laser cutting machine.
-ichan
-
Nice! It really helps if you have a well-equipped shop (and DIY mentality) when you run a business like this.
Regards, Axel
-
As the feeder price of my LSV60 machine are sky high then the machine now has a full of custom tray feeders ;D
The easiest way to make it is just to drill a series of 1.5 mm holes on a thick plastic sheet and use some 1.5mm ferrule to hold the tape on position. What i use is 10 mm acrylic, and the holes is made by laser cutting machine.
-ichan
I've made countless trays for the Juki, using Laser Cut Acylic.. Its cheap, easy to use, and probably a really bad choice ESD wise, but touch wood so far it has'nt given me too much hassle!
-
About the ESD thing: Most commercial trays I see seem to be made of anodized alu which isn't conductive either which makes me wonder if it's really an issue. But if you watch the video where Dave visits his PCB manufacturer:
https://www.youtube.com/watch?v=bTij7Juj5qE (https://www.youtube.com/watch?v=bTij7Juj5qE))
you see theat they went through great lengths to prevent any ESD troubles, the ladies operating the machines even wear anti static vests.
Regards, Axel
-
Lucky you MrP!
ESD is one of those crazy things that creeps up on you. Often not visible, but certainly in there doing its level best to disrupt things as best it can.
Hope those fingers stay crossed a while longer. ;)
As the feeder price of my LSV60 machine are sky high then the machine now has a full of custom tray feeders ;D
The easiest way to make it is just to drill a series of 1.5 mm holes on a thick plastic sheet and use some 1.5mm ferrule to hold the tape on position. What i use is 10 mm acrylic, and the holes is made by laser cutting machine.
-ichan
I've made countless trays for the Juki, using Laser Cut Acylic.. Its cheap, easy to use, and probably a really bad choice ESD wise, but touch wood so far it has'nt given me too much hassle!
-
Re. the stencil alignment pins, http://www.eurocircuits.com/ec-registration-system (http://www.eurocircuits.com/ec-registration-system)
I use the EC printer, and the pin system works very well with stainless stencils, even on large panels.
My main complaint is the EC printer is very over-engineered, so much more expensive than it could be.
-
Re: 30+30+10 parts - please see screenshot below. You get 30 choices for "Left stack" and "Back stack" each. The pickup location of each choice is specified in the "Sys config" menu, see below. Practically, you use the left nozzle (nozzle one) to figure out exactly where to pick up the part and the software automatically sets both the needle ("prick") movement and potential nozzle2 pickup location from that.
As Axel mentioned, the "stack" locations always pick up in the same place, but you can make your own simple advancement mechanisms by using the needle action (think simple rail). YMMV.
New mystery: the software has a self-calibration process hidden under Home|Measure. I see that it measures rotation offsets for the nozzles (nice touch, really impressed!), but after that it goes to what appears to be LeftStack2 and gets stuck there. Can't seem to find a reference to it in the "vedio course" (sic, i love it)
-
Yeah, the measure button, no clue what it does. Also, there's a nozzle calibration in sys config > nozzles > test. That also seems to test for nozzle runout. As for the prick, one could also think of a rotational tray / carousel.
Currently working on a swivel arm to automatically load / unload PCBs. Will post pics when I have something to show.
Regards, Axel
-
Ok, got the beast configured with a few 0402/0603 tape&reels today. Back stack is pretty much full. Impressions:
Good:
- overall the process works and once things are configured, the tiny components are pick&placed correctly.
- nice touches like detection if the solenoid pin did not retract correctly, the pressure sensor, and the option to use machine vision to fine-align components all make for a workable experience.
The bad:
- the mylar tape collection system is shit. It uses rollers on set screws attached to a single shaft, which is advanced each time a part is picked up from either TR stack. The idea is that if there is slack/tape to be peeled, the set-screw friction takes care of it. In reality many of the set screws have to be tweaked: too much friction and the mylar tape thing will move the actual tape, screwing up the pickup. Too little friction and the tape doesn't get picked up, also messing up the pickup. In the end I got it to work but it took a while.
- back stack doesn't have a place for excess tape to go, like the left stack, so it must be managed.
- the whole rigging process is a pain
- reels have a tendency to not sit in the exactly right spot until you hand-fix them. Only a problem whenever the reel axle is picked up though (e.g. add/remove/swap reel)
BTW, each stack actually is configured for 22 physical reels: 18 8-mm, 3 12-mm and 1 16-mm. So it's either 54 parts (22+22 reels +10 tray software locations) or 30+30+10 (software limitation).
-
Thanks for the feedback.
So are you running any production boards or are you just playing around for now?
Share some pictures if you can. I'm interested in this machine and I'd be impressed to see it placing .5mm pitch IC's and 0402 parts correctly.
-
There is some ESD-proofing spray paint, I can't remember what its called but maybe you could spray your trays with it
-
The bad:
- the mylar tape collection system is shit. It uses rollers on set screws attached to a single shaft, which is advanced each time a part is picked up from either TR stack. The idea is that if there is slack/tape to be peeled, the set-screw friction takes care of it. In reality many of the set screws have to be tweaked: too much friction and the mylar tape thing will move the actual tape, screwing up the pickup. Too little friction and the tape doesn't get picked up, also messing up the pickup. In the end I got it to work but it took a while.
- back stack doesn't have a place for excess tape to go, like the left stack, so it must be managed.
- the whole rigging process is a pain
- reels have a tendency to not sit in the exactly right spot until you hand-fix them. Only a problem whenever the reel axle is picked up though (e.g. add/remove/swap reel)
Hi
There are several hundred years worth of designs for friction clutches out there. Using a loose set screw *might* be one of them. It certainly is not in the top 1,000 all time greatest ways to do it.
Sounds like a good project for a 3D printer.
Bob
-
There are several hundred years worth of designs for friction clutches out there. Using a loose set screw *might* be one of them. It certainly is not in the top 1,000 all time greatest ways to do it.
How would you do it?
-
There are several hundred years worth of designs for friction clutches out there. Using a loose set screw *might* be one of them. It certainly is not in the top 1,000 all time greatest ways to do it.
How would you do it?
Hi
Most reasonable designs use a fairly large contact surface (think round plates), a friction material (think felt), and a way to adjust the friction (think in terms of compressing the plates). The plates each slide on the felt. The harder you compress the plates the more fiction you get. Everything is smooth so you get no abrupt start / stop hangups. It can rotate for a *long* time without changing the properties of the materials involved.
One of many examples:
Clamp a smooth sided take up wheel between two drive washers. Felt or cloth goes between the washers and the take up. The washers engage the drive shaft. The wheels just float on the shaft. Big spring on the end of the shaft clamps it all together. Spin the shaft and everything tries to rotate. The wheels that are "at tension" slip.
That's hardly an optimum design, but it's not terribly expensive either. The only real trick is getting the washers to engage with a groove in the shaft. That's what makes it a 3D print job rather than just a trip to the hardware store.
Bob
-
This is a very interesting topic becuase its also how the Yungsheng / SmallSMT machines work.. And i see this being a real down point of these machines..
Its certiainly why the pro machines dont do their take up this way..
-
Best current solution then appears to be independent 'peeler' control - offered by both NeoDen 4, and Charmhigh with their Yamaha pneumatic feeder [both about the same $ with feeders I believe].
Charmhigh are working on a 64 and 128 feeder version at present I'm told.
-
one of the simplest methods was what one of the early low cost machiens did. They put a fishing weight on a string.. It was simple and it worked!
-
If you look at this video you'll see the YS machine, look at 1:10 and you'll see the little device on the left htat is advancing the tape in the feeder. each time it is picked. That gave me an idea.. in combination with a fence wire puller. :-)
along side the carrier tape being pulled, you could have another cam based system that grabed the plastic tape and pulled it back. a clamping mechaism with a spring would hold it while its not being pulled, The cam could be run back and forward across the reels, just like this device is pulling the tape. Build that the right way, and its only going to be a single device that does both jobs at the same time.
I rpobably need to do some modeling of this, but in my head i think its possible.
https://www.youtube.com/watch?time_continue=77&v=-hweqX_Nbh8 (https://www.youtube.com/watch?time_continue=77&v=-hweqX_Nbh8)
-
I wonder if it makes a difference that the motors are positioned differently on the 802B. With the 802A the two motors sit on both ends of the stack while with the 802B they sit on one end. I would expect that to cause uneven pull on the tapes. Also it seems that all four motors on the B engage even when parts are picked from only one stack. Finally, if the 802B has the same controller than the A, the driver that drives two motors on the A now drives 4 motors on the B, likely with less Amps.
Regards, Axel
-
Hmmm - good luck with that one :)
Sounds like another 'project' coming on
If you look at this video you'll see the YS machine, look at 1:10 and you'll see the little device on the left htat is advancing the tape in the feeder. each time it is picked. That gave me an idea.. in combination with a fence wire puller. :-)
along side the carrier tape being pulled, you could have another cam based system that grabed the plastic tape and pulled it back. a clamping mechaism with a spring would hold it while its not being pulled, The cam could be run back and forward across the reels, just like this device is pulling the tape. Build that the right way, and its only going to be a single device that does both jobs at the same time.
I rpobably need to do some modeling of this, but in my head i think its possible.
https://www.youtube.com/watch?time_continue=77&v=-hweqX_Nbh8 (https://www.youtube.com/watch?time_continue=77&v=-hweqX_Nbh8)
-
Best current solution then appears to be independent 'peeler' control - offered by both NeoDen 4, and Charmhigh with their Yamaha pneumatic feeder [both about the same $ with feeders I believe].
Charmhigh are working on a 64 and 128 feeder version at present I'm told.
We need a new thread about CHMT530P,BOREY T15-F30-T4, TVM920
Also, what is the difference between TVM802B and TVM802D
-
About the ESD thing: Most commercial trays I see seem to be made of anodized alu which isn't conductive either which makes me wonder if it's really an issue.
Many P&P machines have ionisers built into them, to discharge static build up around the entire machine. The P&P head may have its own small ioniser attached. Some of the companies which make ionisers for work benches also make ionisers specifically to fit in and around board assembly machines.
-
Would be nice - but I suspect it's wishful thinking at this level
About the ESD thing: Most commercial trays I see seem to be made of anodized alu which isn't conductive either which makes me wonder if it's really an issue.
Many P&P machines have ionisers built into them, to discharge static build up around the entire machine. The P&P head may have its own small ioniser attached. Some of the companies which make ionisers for work benches also make ionisers specifically to fit in and around board assembly machines.
-
Report from the trenches:
Populated 6 "half-PCBs" today (just one side, hence half). 130 components populated per board, majority of them 0402. Learnings:
When the process is down, it works quite well. Last 2 PCB's had only 1-2 rejected pickups and the last PCB took 8 mins to populate, including a few hand-placed parts. That could probably be even faster running at full speed (I'm at 70) and actually using both nozzles (machine comes with only one "0402-compatible" nozzle, so it's mostly a single nozzle show for now).
Notes:
- up-facing vision system is useless for 0402 components, at least for caps. In fact, it actually messes things up, badly, due to incorrect angle detections. Once I turned it off for 0402 components they were placed like a charm (once alignment was good, see below)
- nozzle alignment is absolutely key for each and every reel. Even for bigger components, misalignment means occasional needle jams. For 0402, it's the difference between perfect repeatable placement and components placed on their side or 50m across the lake from the board
- digikey digireel is shit and has a very good chance of breaking the machine - the connection point between the cut tape and the "header tape" often jams, and with thick enough tape can probably bend/break the needle.
- I'm mystified by how the tray counting algorithm works but it works
- CSV import from Altium: lots of magic. I ended up creating a spreadsheet to translate CSV directly into machine-compatable CSV, specifying nozzles, speeds, and so on. I imagine everyone out there will end up making some version of this for themselves
- occasional 90d rotations - maybe because things get juggled at pickup? Not quite sure but a quick inspection before baking helps detect and correct these (happened to 3 components out of the lot)
- the fucking tape collection system - it's the problem everytime, after alignment. Basically if you did the alignment but things are still wrong, it's 90% a problem of tension in the tape collector.. or the used tape not running smoothly out. Eventually some magic tiny turns to certain friction reels seem to have resulted in a well-working system.. for now.
- fiducials with no solder mask around are undetectable by the down-camera. Thought it was a good practice but now ended up permanent-marker-filling in the space around the marker metal.
My stenciling needs improvement. Using T3 paste right now and it's really either all over the place or barely on the pad. Feels like it's too thick. Ordered T5 paste, hope that improves things.
Overall once everything is setup it's a wonderful thing to watch. Small-leaded components are no problem so far. Watch out for those digi-reels. Rigging takes forever.
Hope this is useful. Pictures of the setup once I regain consciousness.
-
- fiducials with no solder mask around are undetectable by the down-camera. Thought it was a good practice but now ended up permanent-marker-filling in the space around the marker metal.
Thanks for the nice update. With the Fid's ,what kind of plating did you have on your board?
-
Report from the trenches:
Populated 6 "half-PCBs" today (just one side, hence half). 130 components populated per board, majority of them 0402. Learnings:
When the process is down, it works quite well. Last 2 PCB's had only 1-2 rejected pickups and the last PCB took 8 mins to populate, including a few hand-placed parts. That could probably be even faster running at full speed (I'm at 70) and actually using both nozzles (machine comes with only one "0402-compatible" nozzle, so it's mostly a single nozzle show for now).
Notes:
- up-facing vision system is useless for 0402 components, at least for caps. In fact, it actually messes things up, badly, due to incorrect angle detections. Once I turned it off for 0402 components they were placed like a charm (once alignment was good, see below)
- nozzle alignment is absolutely key for each and every reel. Even for bigger components, misalignment means occasional needle jams. For 0402, it's the difference between perfect repeatable placement and components placed on their side or 50m across the lake from the board
- digikey digireel is shit and has a very good chance of breaking the machine - the connection point between the cut tape and the "header tape" often jams, and with thick enough tape can probably bend/break the needle.
- I'm mystified by how the tray counting algorithm works but it works
- CSV import from Altium: lots of magic. I ended up creating a spreadsheet to translate CSV directly into machine-compatable CSV, specifying nozzles, speeds, and so on. I imagine everyone out there will end up making some version of this for themselves
- occasional 90d rotations - maybe because things get juggled at pickup? Not quite sure but a quick inspection before baking helps detect and correct these (happened to 3 components out of the lot)
- the fucking tape collection system - it's the problem everytime, after alignment. Basically if you did the alignment but things are still wrong, it's 90% a problem of tension in the tape collector.. or the used tape not running smoothly out. Eventually some magic tiny turns to certain friction reels seem to have resulted in a well-working system.. for now.
- fiducials with no solder mask around are undetectable by the down-camera. Thought it was a good practice but now ended up permanent-marker-filling in the space around the marker metal.
My stenciling needs improvement. Using T3 paste right now and it's really either all over the place or barely on the pad. Feels like it's too thick. Ordered T5 paste, hope that improves things.
Overall once everything is setup it's a wonderful thing to watch. Small-leaded components are no problem so far. Watch out for those digi-reels. Rigging takes forever.
Hope this is useful. Pictures of the setup once I regain consciousness.
Your fiducial detection problem is maybe caused by your ambient light. Don't place a light above the machine and test under different light conditions.
-
Hello, I'm time-constrained and ordering a TVM802A would mean starting production tests earlier than assembling one, learning openpnp and vision stuff. So, I'd like to ask the owners : if you had to redo it, would you buy the TVM802 again ? Thank you !
-
Hello, I'm time-constrained and ordering a TVM802A would mean starting production tests earlier than assembling one, learning openpnp and vision stuff. So, I'd like to ask the owners : if you had to redo it, would you buy the TVM802 again ? Thank you !
Any new machine or process is going to involve significant time to master the process. The only difference between cheap Chinese stuff and the traditional players is the latter will usually do on-site installation and training.
Of course time spent working around badly designed and/or poorly documented software doesn't help but that's going to be the same for all the Chinese machines.
-
@Koen: Yes.
Regards, Axel
-
Yup, it's not perfect, but it does a great job for the price. The only thing that would change that is if it went up in flames tomorrow because of shoddy parts/assembly, but from the parts I've seen it's pretty solid.
-
My machine finally arrived this afternoon and here are a few photos of the winders that drive the reels. The Clutch system now makes more sense...
-
A brief update on our protocol reverse engineering effort.
The controller turns out to be not so primitive after all, at least not as primitive as the software :)
It allows independent control of acceleration, deceleration, and max speed on 6 axis. (X, Y, Z, R1, R2, and strip motors.) So yes, 18 independent parameters. Which is kind of nice. I think acel/decel values are even more important than speed. And we can change them at every step. I.e. move fast accel fast when moving empty and slower when carry part. Original s/w links all three together into a single control.
Also for Z we can accel fast moving down but need to decel slower to avoid losing the part. And moving up, vice versa, we need to accel slowly to keep the part on, but can decel fast.
I don't have time to write up the entire protocol, so I think we'll just publish the source code for our test program. We'll just clean up and organize it a bit. Should be able to post it next week. There are still bits and pieces in the protocol that we don't know what they do (seemingly nothing) but what we do know allows full control of everything that is required to start working on the actual placement software.
corrected.
-
My machine finally arrived this afternoon and here are a few photos of the winders that drive the reels. The Clutch system now makes more sense...
Lucky you have those plastic inserts. Mine came with rubber inserts that provide way, way too much friction. I had to take them apart and leave maybe quarter of the circle to make it more or less reasonable.
-
There are still bits and pieces in the protocol that we don't know what they do (seemingly nothing) but what we do know allows full control of everything that is required to start working on the actual placement software.
Awesome work!
-
I have the same inserts as protoneer. Seems to be another difference between A and B.
Regards, Axel
-
I don't have time to write up the entire protocol, so I think we'll just publish the source code for our test program. We'll just clean up and organize it a bit. Should be able to post it next week. There are still bits and pieces in the protocol that we don't know what they do (seemingly nothing) but what we do know allows full control of everything that is required to start working on the actual placement software.
Looking forward to seeing it. If you put it on Github we can start helping with the documentation part...
-
My machine finally arrived this afternoon and here are a few photos of the winders that drive the reels. The Clutch system now makes more sense...
Lucky you have those plastic inserts. Mine came with rubber inserts that provide way, way too much friction. I had to take them apart and leave maybe quarter of the circle to make it more or less reasonable.
Hi
Consider some sort of lubrication if you have to much friction. The tough part is finding something that does not mess up the rubber or wind up all over everything in sight.
Bob
-
Consider some sort of lubrication if you have to much friction. The tough part is finding something that does not mess up the rubber or wind up all over everything in sight.
Tried, does not help much. I'm in the process of completely replacing cover tape collector.
-
- fiducials with no solder mask around are undetectable by the down-camera. Thought it was a good practice but now ended up permanent-marker-filling in the space around the marker metal.
That is strange, for me unmasked fiducials actually works the best. Have tried adjusting threshold in the PCB setup? Are you using FR-4 boards or something exotic? Masked fiducials is not a good idea since the camera can't distinguish between circle of the fiducial and circle of the mask. It will lose some accuracy.
-
Did my first placement last night and I am pretty happy with the results.
https://www.youtube.com/watch?v=-uwkm41XAew (https://www.youtube.com/watch?v=-uwkm41XAew)
Results of placement attached.
I have placed all of my production boards by hand up to now. My intention with this machine is increase volume and to free up some time. It has raised an interesting situation for me.
When manually P&P'ing a proportion of time is spent on making sure parts are placed correctly. With these machines focus have been shifted and I can see a need for extra quality control.
This has pointed me to Automated Optical Inspection for:
- Checking solder paste coverage
- Missing/misplaced components before reflow
- Misplaced components after reflow
- Checking obvious shorts after wave soldering
Found a really nice blog on the topic by Bob Baddeley
http://bobbaddeley.com/2015/12/creating-an-automated-optical-inspector-for-50/ (http://bobbaddeley.com/2015/12/creating-an-automated-optical-inspector-for-50/)
The python script works with OpenCV and brings up a life video stream that magically focuses on issue parts.
(Needs a small amount of training but works amazingly well)
-
On machines of this class is changing reels painful and requires any re-alignment or is it easy (when building two sided boards) ?
-
Regarding the pressure sensor in the machine. I initially wrote that it has a mechanical pressure switch, but I was wrong. It actually has an electronic pressure switch. Then I'm wondering why is working so unreliably... Maybe the threshold is set incorrectly. I wish they transmitted digital pressure value to the PC, but alas, they only send the result of comparison with the setpoint.
-
On machines of this class is changing reels painful and requires any re-alignment or is it easy (when building two sided boards) ?
I have only had a few hours play and doing double sided boards should not be an issue. On the board side alignment is done with fiducials and it works really well. (Jobs can be stopped half way if a nozzle change needs to be made and the machine will just do a fiducial check on restarting.)
As for changing reels. Once you understand its working it should not be to hard to change. With my first few runs I noticed that the nozzles kept picking up resistors off center. The placement as seen above is okay for 0805 but could be better.
Yesterday I played with a reel-less off cut of Sot-23's. Had the same issue with the component not being picked up in the centre. With SOT23's misplacement was a bigger issue and not acceptable. Played around and the problem ended up being the offset between the nozzle and the needle that advances the reels.
The files that the machine saves includes all of the config and offsets settings. As I play with , it is starting to get more and more accurate.
I use Kicad and getting the component positions out is pretty easy. Things to consider is the component orientation in the tray.
Micro's are also not that bad to work with. I played with an ATMEGA328P-AU (0.8mm pitch) at an 45degree angle. The camera system aligns well but as mentioned before the system is very sensitive to light sources. I am thinking of making shades for the bottom camera as soon as the machine has been moved to a permanent spot.
-
The files that the machine saves includes all of the config and offsets settings. As I play with , it is starting to get more and more accurate.
A bit of correction: the placement file only contains locations of the fiducials. (I mean, obviously, aside from actual components). All of the offsets and other data, such as camera-to-nozzle, nozzle-to-needle, location of the feeders and everything else, the s/w stores in the n/v memory of the machine, which has space to store at least 400*4 bytes of data. Maybe more, but the s/w uses about that many.
-
I have attached an example file that the machine software exported/saved. (For example it includes the nozzle-to-needle offset at 1.60. So it looks like these settings get sent to the machine when the files gets loaded)
Actually makes sense cause things like exposure settings could change for board colors and should be saved as part of the config.
I am starting to like this little machine... :-/O
-
I have attached an example file that the machine software exported/saved. (For example it includes the nozzle-to-needle offset at 1.60. So it looks like these settings get sent to the machine when the files gets loaded)
Actually makes sense cause things like exposure settings could change for board colors and should be saved as part of the config.
I am starting to like this little machine... :-/O
Nozzle to needle offset is two coodinates, not one :) 1.60 is a magic number that nobody really knows what it is. It is also 1.60 in the files that I generate and I copied it from mathias's blog :)
You can try and change numbers in the System Config, where actual nozzle-to-needle is located. See if it will change in the file you save :)
You can change parameters in the sys config and not save any file at all and yet they will be restorered next time you open the program.
-
Haha... Lots to learn and discover.
Tonight I will try some old boards without fids. Hopefully I can use a pad om some kind to align it.
-
Guys, how is the focus on your down cameras? I just noticed that the focus get significantly better if PCB is lowered a bit.
-
random thought but 1.6 is a pretty common pcb thickness...?
-
random thought but 1.6 is a pretty common pcb thickness...?
Now that sounds like a totally sensible guess! You are probably correct.
-
random thought but 1.6 is a pretty common pcb thickness...?
Looking at it again last night, PCB settings gets saved to the export file but not the machines settings as ar__systems pointed out. 1.6mm is the PCB thickness and gets saved,yes.
Focus on both my camera's seems to be good. The cameras seems to have adjustable lenses so that might be something to play with. Exposure set at 7 works well for red boards on the down camera.
Last night's test went well and I was able to align my old boards without fid's. The software has the option to manually align boards with the down camera or to run it as it is. (Used TH pins to align the 6 boards on my panel)
The reel feeders are the trickiest bit to setup. "Nozzle 2 Cam" , "Nozzle 2 Advancement Needle" needs to be calibrated properly.
At the moment I have two small issues.
- The nozzle does not pick up the part in the center
- The advancement needle keeps getting stuck with one particular reel.
-
Focus on my down camera is very good, up camera seems to be a little out of focus. At least on my machine the focus is not adjustable, lenses are glued / loctited.
The off-center picking happened on one of my reels, too (0805 caps in the tape).
Regards, Axel
-
Mine were glued too, but I was able to remove the glue from the down camera. Now I can adjust focus.
Initially I was testing down camera only with gold plated boards. And Fiducial detection worked well. But now I tried with HASL boards, green mask - and none of the "exposure" settings are working. Either the the fiducial and unmasked PCB around it are underexposed, or overexposed.
I wonder why they decided to use that stupid red backlight?
-
With HASL boards, reflection of incident light seems to be the biggest problem, darken the room or put a cardboard shade around the camera. The ENIG boards with a matte finish from PCB Pool worked best in my tests (too bad they cost so much more that the Seeed boards, at least for small volume.)
Regards, Axel
-
Alpha version of automatic loading and unloading of PCBs into the machine
Write-up about progress so far:
http://servokit.com/blog/2016/04/08/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-i/ (http://servokit.com/blog/2016/04/08/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-i/)
Video follows.
Regards, Axel
-
Axel, the Chinese manufacturers do ENIG as well, although the cost-up is a bit steep.
My latest boards were from MultiCB, at first look same quality as PBB-Pool, but about1/3 cheaper.
Max
-
Alpha version of automatic loading and unloading of PCBs into the machine
Write-up about progress so far:
http://servokit.com/blog/2016/04/08/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-i/ (http://servokit.com/blog/2016/04/08/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-i/)
Video follows.
Regards, Axel
:-+ :-+ :-+ I did 40 similar size boards,hand feeding it. The boards did not have Fids and I only had one that I stopped after one part to realign.
I am very happy with this little machine. I load the CSV configuration,insert the board and hit the play button on the machine. Once the board is done, I load another empty board and press play button to start the next run.
Next set of boards will be panelized and will have 6 sub boards and a 6 in one SS stencil. I once again found that solder paste done properly helps a lot. If the paste is misaligned it makes it even harder to see misaligned parts.
With the 250+ parts I placed I had 4 stops where the advancement needle got stuck. All on the same reel with the first pick. Luckily these machines are easy to diagnose.
The needle only gets stuck for two alignment reasons.
- Y direction, the reel position needs adjustment(Easy app setting).
- X direction, its the cover stripping that is moving the tape instead of the advancement needle. To fix it put some slack in the cover tape collector and if need be adjust the screws on it to make it pull with less force.
-
Yeah, I got about the same results, but I don't find them quite satisfactory. I've had boards that have 220 parts per one board, and that would mean 4 stops every board. Not quite good enough for me.
-
Another user of the 802A in Germany has placed .5mm QFN successfully:
https://www.mikrocontroller.net/topic/371594#4539279 (https://www.mikrocontroller.net/topic/371594#4539279)
Regards, Axel
-
Some open heart surgery... - I soldered wires to the Run/Stop LED and the start/stop button so that I can see the machine state and control the machine from an Arduino.
Regards, Axel
-
Another user of the 802A in Germany has placed .5mm QFN successfully:
https://www.mikrocontroller.net/topic/371594#4539279 (https://www.mikrocontroller.net/topic/371594#4539279)
Regards, Axel
yes, thats me.
I purchased the TVM802B last week.
My boards contain mainly 0603 parts and QFN pitch 0.5mm microcontrollers.
It took a few days and many hours to setup the machine correctly. The threshold of Cam-1 was very important for accuracy.
After some trial and error things worked well and it was time to test the first QFN, an ATXmega128A4U-MU with 44 pins and a thermal pad.
Vision set to Accuracy and speed to 50%.
The QFN was positioned very precise, I could not see any significant offset with the microscope. It is much better compared to my previous manual placement, so I do not expect any soldering problems.
Currently I am working on an easy integration into Eagle, that is almost done. Then I will CNC a few trays for various ICs. I hope to do the first real boards this week and will post microscope-photos when done.
Regards, Harry
-
script for eagle. This is modified script I took from mathias' blog. parts named as FIDx will be included in fiducials.
-
Another user of the 802A in Germany has placed .5mm QFN successfully:
https://www.mikrocontroller.net/topic/371594#4539279 (https://www.mikrocontroller.net/topic/371594#4539279)
Regards, Axel
yes, thats me.
I purchased the TVM802B last week.
My boards contain mainly 0603 parts and QFN pitch 0.5mm microcontrollers.
It took a few days and many hours to setup the machine correctly. The threshold of Cam-1 was very important for accuracy.
After some trial and error things worked well and it was time to test the first QFN, an ATXmega128A4U-MU with 44 pins and a thermal pad.
Vision set to Accuracy and speed to 50%.
The QFN was positioned very precise, I could not see any significant offset with the microscope. It is much better compared to my previous manual placement, so I do not expect any soldering problems.
Currently I am working on an easy integration into Eagle, that is almost done. Then I will CNC a few trays for various ICs. I hope to do the first real boards this week and will post microscope-photos when done.
Regards, Harry
Hi
There are a number of people *very* much interested in how the machine does with 0.5 mm parts. In my case, I could care less how slow it needs to go (50% is fine) when placing fine pitch. I also could care less how much one time calibration is required when setting up the machine (up to a point). If the machine takes four hours to re-calibrate every day .... that's probably worth knowing about.
Please keep us all informed about your progress !!!
Bob
-
here is my first picture:
I put a tiny piece of double sided adhesive tape on the PCB and then placed this microcontroller QFN 0.5mm pitch.
As we can see the precision is very well.
It is extremely important to set the fiducials as precise as possible. Just a little mistake with the measurements and the QFN cannot be placed accurately.
The TVM802 works best with little round fiducials. Rectangles do not work as well.
Harry
-
here is my first picture:
I put a tiny piece of double sided adhesive tape on the PCB and then placed this microcontroller QFN 0.5mm pitch.
As we can see the precision is very well.
It is extremely important to set the fiducials as precise as possible. Just a little mistake with the measurements and the QFN cannot be placed accurately.
The TVM802 works best with little round fiducials. Rectangles do not work as well.
Harry
Hi
That looks quite good. The PC board it is on is out by roughly as much as the placement. That has always been one of the arguments for down facing cameras that "spot" the pads on the board. The counter arguments have generally revolved around the (very) low speed of that process and the difficulty of recognizing pads in the clutter of a board layout plus solder paste.
Bob
-
here is my first picture:
I put a tiny piece of double sided adhesive tape on the PCB and then placed this microcontroller QFN 0.5mm pitch.
As we can see the precision is very well.
It is extremely important to set the fiducials as precise as possible. Just a little mistake with the measurements and the QFN cannot be placed accurately.
The TVM802 works best with little round fiducials. Rectangles do not work as well.
Harry
Hi
That looks quite good. The PC board it is on is out by roughly as much as the placement. That has always been one of the arguments for down facing cameras that "spot" the pads on the board. The counter arguments have generally revolved around the (very) low speed of that process and the difficulty of recognizing pads in the clutter of a board layout plus solder paste.
Bob
Downward vision to locate pads is hopeless due to paste.
If you need higher local accuracy you use local fiducials. But you really want the basic accuracy of the machine to be enough to only need corner fids. A machine that can't do 0.5mm pitch without local help is pretty poor.
-
Hi all, purchased the same TVM802B, now I am in the process of calibration, there is a small problem with the position of the components on the board, reference points exhibit normal. What it is necessary to slightly shift the position. maybe change Mark PCB position few degrees, or Real mark? or something else?
-
If all components are off, check the nozzle calibration (Sys Config > Nozzle > Test buttons). If just one is off, might be feeder issue, see if the needle comes down at the right place, correct if necessary in Sys Config > Left Stack / Back Stack.
Regards, Axel
-
If all components are off, check the nozzle calibration (Sys Config > Nozzle > Test buttons). If just one is off, might be feeder issue, see if the needle comes down at the right place, correct if necessary in Sys Config > Left Stack / Back Stack.
Regards, Axel
All component are off, nozzles test parameters attached, camera define nozzles like on video from CD, before that had brought down the settings to stand again(left stack,back stack, prick, camera position)
-
The values look ok to me, much like on my machine. Some things you might want to check:
1. Make sure, the fiducial coordinates are correct relative to the origin of the board. The edges of the different layers might not be exactly aligned (see Neoden thread for an example: https://www.eevblog.com/forum/reviews/neoden-4-pick-and-place/msg897138/#msg897138 (https://www.eevblog.com/forum/reviews/neoden-4-pick-and-place/msg897138/#msg897138))
2. Make sure PCB is fixed and can not move (stupid but happened to me).
3. Nozzle size can be an issue, look closely if parts are drifting or rotating when the nozzle comes down. Also there's a "place delay" in the config, I found "0" here working best.
Perhaps you could post a photo of a placement error.
Regards, Axel
-
my first board is finished, here are some pictures.
The QFN was positioned very well.
The 0603 Rs and Cs are not so precise because it does not use the camera.
This has to do with the pick-up process, I'm not sure if nozzle 1 or 2 is the best choice.
The glass diodes did not work at all, I corrected them manually. For any reason the nozzle cannot hold them strong enough and they move around on the way from the stack to the PCB.
I will not use glass diodes any more. Other diodes in SOD case are no problem.
The FPC connector was placed manually because I currently don't know how to define it's offset.
The glass diodes are the only parts which I corrected manually before soldering in the oven.
Even if some R and Cs looked not so good, they corrected themself during the soldering process.
Harry
-
All component are off, nozzles test parameters attached, camera define nozzles like on video from CD, before that had brought down the settings to stand again(left stack,back stack, prick, camera position)
after measuring the ficudials the board is synchronized with the upper camera. This works quite well.
If all components are off then I would play with the Camera2 - Nozzle1-Offset settings which should move all components.
I will test that tomorrow.
Harry
-
my first board is finished, here are some pictures.
The QFN was positioned very well.
The 0603 Rs and Cs are not so precise because it does not use the camera.
This has to do with the pick-up process, I'm not sure if nozzle 1 or 2 is the best choice.
The glass diodes did not work at all, I corrected them manually. For any reason the nozzle cannot hold them strong enough and they move around on the way from the stack to the PCB.
I will not use glass diodes any more. Other diodes in SOD case are no problem.
The FPC connector was placed manually because I currently don't know how to define it's offset.
The glass diodes are the only parts which I corrected manually before soldering in the oven.
Even if some R and Cs looked not so good, they corrected themself during the soldering process.
Harry
Hi
That all looks quite good.
Even with gear much fancier than a TVM802, we have given up on round body parts. MELF's, diodes, and all the rest have been off our "products we use" list for a very long time. I can't think of any time I've been unable to do a design because of that restriction.
Bob
-
It's nice to see this machine accuratley placing .5mm pitch QFN parts :clap:
Keep them pictures and feedback coming guys!
-
after measuring the ficudials the board is synchronized with the upper camera. This works quite well.
If all components are off then I would play with the Camera2 - Nozzle1-Offset settings which should move all components.
...
tested it today, and it works.
I adjusted the Camera2 - Nozzle1-Offset values and now it is perfekt.
Today I placed a lot of SOD-523 diodes. This is a really small case, but Nozzle Nr.1 (Juki 503) picks it up well.
Placement was not so good, so I activated the cam "accurate mode", now it is fine. Of course, that takes a long time, but I'm not in a hurry.
Harry
-
First video of the automated PCB loading and unloading:
https://youtu.be/mseT093Ib80
Write-up follows.
Regards, Axel
-
anyone ever try something like a bga package ?
The ones I use are generally low ball count and I place them by hand now. (actually find it easier than qfn as I think it aligns better if it is pretty close to position).
A part like:
http://www2.st.com/content/ccc/resource/technical/document/datasheet/group0/cb/91/ba/7d/0b/c1/4d/f6/CD00147128/files/CD00147128.pdf/jcr:content/translations/en.CD00147128.pdf (http://www2.st.com/content/ccc/resource/technical/document/datasheet/group0/cb/91/ba/7d/0b/c1/4d/f6/CD00147128/files/CD00147128.pdf/jcr:content/translations/en.CD00147128.pdf)
WLCP (CS) package
All my boards are prototype runs and it is too much of an hassle and expense to outsource that.
Pretty much all 0402 / 0603 parts and a few ic's. 0.5 mm pitch qfn is normal.
I actually have the liteplacer and it can work quite well. Setup for new parts is super easy. But it needs quite a lot more refinement to be really useful for me.
The neoden 4 looks quite good but when you get all the goodies you want it is a 10k pricetag ex tax. Which is doable ... but it is a Chinese machine.
If it was running open pnp or something like than it would be another story...
So even if this 3k machine just does my qfn's and 0402 somewhat reliable I would be happy, but is that the case?
edit:
found many video's including 0402 placement without vision:
http://www.soku.com/search_video/q_TVM802A?f=1&kb=040200000000000__TVM802A&_rp=1460819713986hCZIRb&_rp=1460819713986hCZIRb (http://www.soku.com/search_video/q_TVM802A?f=1&kb=040200000000000__TVM802A&_rp=1460819713986hCZIRb&_rp=1460819713986hCZIRb)
looks good
-
Great Job Axel!
Can't wait to see how this progresses :popcorn:
-
Thanks! Printing is next.
Regards, Axel
-
anyone ever try something like a bga package ?
The ones I use are generally low ball count and I place them by hand now. (actually find it easier than qfn as I think it aligns better if it is pretty close to position).
A part like:
http://www2.st.com/content/ccc/resource/technical/document/datasheet/group0/cb/91/ba/7d/0b/c1/4d/f6/CD00147128/files/CD00147128.pdf/jcr:content/translations/en.CD00147128.pdf (http://www2.st.com/content/ccc/resource/technical/document/datasheet/group0/cb/91/ba/7d/0b/c1/4d/f6/CD00147128/files/CD00147128.pdf/jcr:content/translations/en.CD00147128.pdf)
WLCP (CS) package
All my boards are prototype runs and it is too much of an hassle and expense to outsource that.
Pretty much all 0402 / 0603 parts and a few ic's. 0.5 mm pitch qfn is normal.
I actually have the liteplacer and it can work quite well. Setup for new parts is super easy. But it needs quite a lot more refinement to be really useful for me.
The neoden 4 looks quite good but when you get all the goodies you want it is a 10k pricetag ex tax. Which is doable ... but it is a Chinese machine.
If it was running open pnp or something like than it would be another story...
So even if this 3k machine just does my qfn's and 0402 somewhat reliable I would be happy, but is that the case?
edit:
found many video's including 0402 placement without vision:
http://www.soku.com/search_video/q_TVM802A?f=1&kb=040200000000000__TVM802A&_rp=1460819713986hCZIRb&_rp=1460819713986hCZIRb (http://www.soku.com/search_video/q_TVM802A?f=1&kb=040200000000000__TVM802A&_rp=1460819713986hCZIRb&_rp=1460819713986hCZIRb)
looks good
Hi
So far the consensus seems to be that, with enough fine tuning, the machine will do 0.5 mm pitch parts. The only thing that the $250K machines could do on a BGA is optically pick up the balls on the bottom of the part. That's not to say that they all *do* pick them up that way. They also are a bit smoother in doing motion. The motion stuff applies equally to a QFN and a BGA so that sort of drops out in this case.
From what I can see here, we may have an "open source" TVM802 driver at some point. That might allow you to enhance all sorts of things ....
Bob
-
what i did with my liteplacer was to let the machine pick it up at the center (via vision) and just let it place according to the pnp location.
That works fine even for 2x2mm dc/dc converters and so. For bga angle and center correction (like you do with a qfn or so) is more than enough.
Since in my opnion some qfn's are more difficult than bga's. The ones I use are 0.5mm pitch anyway.
Any opinions regarding getting the larger B version?
It is about 900 euro more but also has 20 more feeders or so.
-
what i did with my liteplacer was to let the machine pick it up at the center (via vision) and just let it place according to the pnp location.
That works fine even for 2x2mm dc/dc converters and so. For bga angle and center correction (like you do with a qfn or so) is more than enough.
Since in my opnion some qfn's are more difficult than bga's. The ones I use are 0.5mm pitch anyway.
Any opinions regarding getting the larger B version?
It is about 900 euro more but also has 20 more feeders or so.
Hi
If you *need* the extra feeders, it's a no-brainer. If your usage pattern is such that you do a switch out job once a month on the A model ...I'd say that's not a big need. Others would say it is. The board size difference between the two is not significant in terms of anything I would do on a machine like this. For the guy with a lot of 400 mm long boards .... it's a big deal.
It's not clear if both machines got a nozzle upgrade or if the A still is screw ins and only the B gets "Juki" nozzles.
Bob
-
It's not clear if both machines got a nozzle upgrade or if the A still is screw ins and only the B gets "Juki" nozzles.
"A" model still comes with screw nozzles(Received mine a few weeks ago). Not that that is a problem, will upgrade them to Juki when I need replacement nozzles. (I bought plenty of spares before seeing the Juki upgrade option)
I do like the idea that the "B" model's camera is at the back.It opens up more space at the front for feeders and if the bottom camera is not used a lot it is just in the way.
Upgrading the bottom camera is a good idea but might not be compatible with the current software.
-
I tried 0.4mm pitch and the results are pretty good!!! :-+ :-+ :-+
The MCU in the photo's is a NRF51XX 48-pin QFN with a 0.4mm pitch.
I did it a 40% speed and with the bottom camera set to accurate placement.
Used a number 1 nozzle from the front stack of the machine.
-
The little pick and place engine that could!
Re. the camera, I think the SW talks to the cameras on the machine, see the serial numbers in the sys config panel. Replacing is likely not an option.
Regards, Axel
-
From the pictures, the placement looks as good, if not better than the landing pads for the QFN. Maybe just the optics on the pictures, but looks really good!
Kudos.
I tried 0.4mm pitch and the results are pretty good!!! :-+ :-+ :-+
The MCU in the photo's is a NRF51XX 48-pin QFN with a 0.4mm pitch.
I did it a 40% speed and with the bottom camera set to accurate placement.
Used a number 1 nozzle from the front stack of the machine.
-
Yea man really nice to see the machine accuratley placing 0.4mm pitch IC parts! That's really good for a machine in the 3k range!
If it only had 100 feeders ;D
-
From the pictures, the placement looks as good, if not better than the landing pads for the QFN. Maybe just the optics on the pictures, but looks really good!
Yes, I am pretty pleased with it. I must admit that its getting to the stage where the rest of my tool chain is starting to lag. At work we have a really nice microscope that costs more than than a new TVM802B. For that price you can zoom in on the pads of a QFN at 1 feet(30cm). It also has auto-focus allowing you concentrate on the part.
Photos were taken with a $50 usb microscope (http://www.aliexpress.com/item/US-UK-AUS-STOCK-HD-2MP-USB-Andonstar-Digital-Microscope-Webcam-Endoscope-Camera-Metal-Stand-Base/1761453184.html)
-
Detailed write-up of automatic PCB loading:
http://servokit.com/blog/2016/04/17/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-iii/ (http://servokit.com/blog/2016/04/17/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-iii/)
Regards, Axel
-
Hi
The placement on all of the QFN's has looked quite good. Any time you zoom in on a PC board at that level, you begin to see the limits of the fabrication technique on the board. You also start to see the fab limits on the package it's self. If you put solder down, you'll see some limits on your stencils and maybe your printer.
The microscope is *not* the only part of the tool chain this stuff pushes. You are getting to "as good as it gets" for most fab shops at 0.4 mm. I hesitate to use the term "state of the art". There are finer pitch things out there. A lot of pretty capable groups are still challenged by 0.5 mm.
Bob
-
Any opinions regarding getting the larger B version?
It is about 900 euro more but also has 20 more feeders or so.
They have differences other than just the number of feeders. In other words you can't just buy A and upgrade it into B. They are substantially different in the construction.
First of all, B has a slot in the plate for spent tapes on the Left feeder block to be routed. Camera location is different. A has a built-in tray for chips. They have different nozzles.
-
Any opinions regarding getting the larger B version?
It is about 900 euro more but also has 20 more feeders or so.
They have differences other than just the number of feeders. In other words you can't just buy A and upgrade it into B. They are substantially different in the construction.
First of all, B has a slot in the plate for spent tapes on the Left feeder block to be routed. Camera location is different. A has a built-in tray for chips. They have different nozzles.
Hi
Is there any reason to believe that one machine is more accurate than the other one? This is in the context of doing tight pitch parts using the camera.
Bob
-
TVM802A is 2.3k euro on taoboa. I wonder how much it would take for a agent to ship it here.
That would save you a few hundred bucks.
-
TVM802A is 2.3k euro on taoboa. I wonder how much it would take for a agent to ship it here.
That would save you a few hundred bucks.
Hi
No matter how cheap the shipping, there still is likely to be a tax / customs man getting his charges on the deal :)
I suspect that if a dozen or two people could all agree on a delivery point and organize a group order ... the machines would be amazingly affordable. It's not going to work very well for a US to Europe sort of group. It might well work for a group of people within a 4 hour drive of some central point in the EU. The hardest part is always organizing the whole thing and tending to all the details.
Bob
-
For company use you pay the tax but get it back :)
-
Hi
Ok so I'm wandering around the internet and up pops the fact that there are *four* flavors of the TVM802:
TVM802A
TVM802B
TVM802C
TVM802D
It is *abundantly* unclear to me exactly what the difference between the A and the C ( or the B and the D) actually is. This is compounded a bit by the fact the info on the C shows pictures labeled "TVM802A". Likewise, the TVM802D pictures show a machine labeled "TVM802B". Does anybody have an idea of what's going on ?
Bob
-
Is there any reason to believe that one machine is more accurate than the other one? This is in the context of doing tight pitch parts using the camera.
I don't think so. They should be about the same.
-
someone tried 0402 parts already ?
I placed a lot of SOD523 diodes today and it worked very well. SOD523 are not much bigger than 0402, I will give it a try in my next board.
Regarding A,B,C,D type: I have got all the leaflets from the manufacturer and they show only A and B machines.
Harry
-
someone tried 0402 parts already ?
I placed a lot of SOD523 diodes today and it worked very well. SOD523 are not much bigger than 0402, I will give it a try in my next board.
Regarding A,B,C,D type: I have got all the leaflets from the manufacturer and they show only A and B machines.
Harry
Hi
I just noticed the C and D go up today. The pictures don't show C and D markings on the units. I'd bet they are something new. Pricing on each of them is just slightly higher than the A or B.
Here's a link to a search that turns up a bunch of TVM802C's.
http://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20160417172144&SearchText=TVM802C (http://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20160417172144&SearchText=TVM802C)
Everything I see in those offers is a TVM802A. I may have missed something subtle. I'd bet the "real" copy for the other machines is still in the works. It's pretty odd that they have them up for sale, but no information on what they are.
Yes, for about $3,300 (plus shipping) I *could* find out the answer ... sorry ... not doing that just to find out.
=====
Edit 4/10/2016:
Well I decided to take the lower cost option and contacted the company. Their reply:
The C and D are the same as the A and B. The name change is simply a way for them to differentiate the product sold directly by them from the product sold by dealers.
... not very exciting. I was hoping for a raft of new features :(
Bob
-
someone tried 0402 parts already ?
I've populated over 10 PCB's now, each with >100 0402 parts. The first 1-2 were a pain due to alignment tuning and protective film collector tuning, the rest have been without problems, to the point that I can let it populate all the 100nf caps and just come back a few mins later. You must make sure to use the right nozzles: my exp has been smallest nozzle for 0402 and 0603.
My biggest problem now is 1) not enough reel spaces 2) crappy tray setups 3) setting up orientations for new parts, especially expensive IC's that I don't like wasting. Due to the need to still place some parts by hands it's not a fully automated solution in our case, but did make placement a whole lot easier.
For best results
1) make sure both fiducials align
2) run step by step the first time you run it and take the time to tune out each problem
3) don't use the 1/2 nozzle feature. Nozzle 1 *or* 2. I've had weird occasional 90deg rotations with the 1/2 feature and I'm at least 25% sure it's not my fault.
-s
-
thanks for your experience, Sergeil.
That sounds good.
I have a little trouble with transistors in SOT-23.
I tried the smallest nozzle, maybe its too small.
When I move from the tape to the camera, then just before it enters the camera
the part sometimes turns around, one was even flipping top/bottom.
It looks like that there is something magnetic which pulls on the part, but I'm not sure.
-
They just turn from vibration and vacuum being not strong enough. Also, move the pick up location as close to the peeling edge as possible. Parts can make a flip from the pick transport needle hitting on the tape.
I connected external vac source to mine. There is not much to do about vibration, other than not use speeds below 70%. Somehow the vibration is worse at lower speeds.
-
Exactly what ar_systems said: make sure the pickup location is as close to the peel-off as possible. I had flipped sot23 and similar as well before that tuning.
Too small nozzles seem to have too little rotary friction with the part, meaning when it rotates the part doesn't quite rotate with the nozzle (e.g. picking up a LQFP64 with a small nozzle). Use the bigger rubber nozzles and those parts get placed much better.
-
thanks for the explanation.
I can do that by reducing the X coordiate of the left stack a little bit, right ?
Currently I use the values from the CDrom (from the pictures).
I will test that with the single step fuction.
Should I replace the original vacuum pump by an external one ?
If yes, how to choose the right pump ?
-
Today I tried two things: first, I replaced the rubber crap in the peel reels with brass pads. Works much better now.
Second, I tried a new concept for the peeling mechanism. Please don't laugh, it is a prototype. :) And it works like a charm.
So basically we have a motor with a spool on a arm that rotate on a pivot and thus can bring the spool closer or further from the feeder. As it moves further, the steel sheet enters the slot of the photo-interrupter which in turn activates the motor (not seen on the photo, but the black spool is sitting on it). The force with which the device pulls the cover tape is constant, and is set by the spring (rubber band in this prototype). White plastic spring prevents the black spool from turning back when the motor is not powered.
I wonder why this simple method is not used in any of the cheap machines. Once the proper force spring is selected, it basically never needs another adjustment, unlike the shitty friction based peelers. The tiny DC gear-motors can be had for like 4 bucks or less, so for the total cost of like 300 a perfectly consistent peeler can be made.
-
Hi
Ok, so here's an inherently un-answerable question:
How long will it be before something better comes along at a lower price?
Better at a higher price does not count in this case. You can always get better for more money.
Better in this case probably means a significant improvement in terms of:
1) Vision, machines without vision need not be considered. This also includes the whole part recognition / placement process.
2) Feeders, machines with fewer feeders or less functional feeders .. not so much. Quick change feeders would score in this category.
3) Nozzles and basic pick function. Auto change would be a plus in this category, so would handling a wider range of parts.
4) Something really significant in software. Full open source would be a plus here. It would need to be more than incremental improvement or bug fixes.
I am by no means saying it must be an improvement in all three. Simply that it needs to be notably better in one of these categories. I've left out placement speed and number of heads. Both of those are very application specific in terms of their real performance. Coming up with a real speed number means running *your* typical job on the machine. I've also left out board feeders since that sort of gets you out of the "desktop" world.
Yes, it's a very arbitrary list of what's in and what's out. There could be a *lot* of very valid criticism of the picks on this list. Nit picking the list is fine and feel free to do so. My larger focus is on the "when" not on the quality of the list. The why of asking this is fairly simple: Nobody wants to buy a shiny new laptop and find out the new one comes out the next day. I'm saying that the machine becomes scrap when the new one comes out. The "old" laptop still works fine the next day.
So - any guesses? Is there some, known only to a few, "model year" cycle to this stuff? Are we two years away or five years? I will admit I only got to thinking about this after the A->C and B->D model stuff popped up.
Bob
-
I would honestly think that better for less is going to be hard to come by. You don't get much more simple than the low end Neoden TM-240 / Qihe TVM802 units. They are pretty much stripped down to the bone in terms of features vs price vs build quality. Both have pretty good feedback regarding build quality. Both have similar price. If the Chinese are selling at a given price, and the price is relatively even across the market leaders, you really can't expect to get much more for less, unless the cost of the components drop due to certain components going mass market or alternate suppliers selling cheaper parts. Pretty much everything on these units is mass market items such as steppers, drivers, belts, pulleys, usb cameras, etc.
The only proprietary parts are the physical machine base and control board / software. The workmanship / precision put into the machine base is directly related to machine cost. The flatter, or more rigid it becomes, the more it weighs, and the more costly it becomes. They are made in relatively small quantities (ie compared to the mass market parts), so not much is going to change there.
The software / control boards are the only parts that really separate the different vendors, and everyone here is pretty focused on the software and how well it works / flexibility, so one could only assume if we want better software, the cost will only go up in this area. Better software requires more development time / resources.
As we already have seen with the Neoden 4 which costs significantly more than the TM240, and also with the Qihe TVM920, Charmhigh CHMT530P, adding more features drives the cost up. Once you get past a certain point in mechanical design, you have to forgo the hobby parts, and start using lots of custom machined parts. This adds cost really quickly
I am sure that there are ways to build cheaper PNP's than china offers, but in reality would you really want one? You gotta sacrifice somewhere to drive the cost down. Sometimes the sacrificial lamb is staring right at you in terms of features, other times it is in the form of sub-par components. Both will drive you crazy when it comes to PNP's.
If you want to save money, take a look at the OpenPNP software project and the accompanying OpenBuilds PNP hardware. You can build it up in basic form for about $1500. The software is great and moving further forward every month, and the hardware is really dependent on your mechanical skill and budget. You forgo the China labor in favor of your own blood, sweat and tears. You will get more bang for your buck this way (theoretically), but you also have to consider whether you have the time.
Alternative, consider a used piece of equipment. It may not be new, but sometimes you can find a well maintained commercial PNP at a really reasonable price, which can satisfy your requirements. Often the software will be antiquated, but you can always consider using / porting to openPNP to run the machine, and exploit its commercial grade hardware.
As with all things in life, you get what you pay for... Hopefully, what you can afford, can also meet your needs / requirements.
-
I would honestly think that better for less is going to be hard to come by. You don't get much more simple than the low end Neoden TM-240 / Qihe TVM802 units. They are pretty much stripped down to the bone in terms of features vs price vs build quality. Both have pretty good feedback regarding build quality. Both have similar price. If the Chinese are selling at a given price, and the price is relatively even across the market leaders, you really can't expect to get much more for less, unless the cost of the components drop due to certain components going mass market or alternate suppliers selling cheaper parts. Pretty much everything on these units is mass market items such as steppers, drivers, belts, pulleys, usb cameras, etc.
........
Hi
Ok, there's the first "vote". Thanks for the input.
I would score that as "maybe never". The price of chunks of metal does not follow an ever declining curve the way silicon IC's do. There is a lot of metal chunks in a pick and place. There is not a lot of costly silicon in a pick and place. About the only part of the machine that does not apply to is the vision stuff. I can easily see better / cheaper cameras coming along. Do they improve the machine? I suppose they *might*.
Bob
-
How long will it be before something better comes along at a lower price?
Better at the same or slightly higher price would be a start.
The Chinese are too focused on lowering the price and thiking that placement speed is the only important benchmark, so corners get cut, especially with software design & implementation, and I can't see that changing any time soon.
There is such a huge gap between the more established makers and the newer Chinese machines that there has to be a market for something more polished and better supported at a slightly higher price.
There are plenty of areas for things to be improved, and none of them should be that hard to do given a bit of effort:
Auto nozzle change, taller parts, better feeders, better software, flying upward-looking camera.
-
The cost of the cameras is not really relevant, they should not contribute much more than 100$ to the price. Some machines use a PC as a controller, so even from electronics perspective, there is not much room left (as you stated correctly).
More or less all components that go into such a unit have been on the market for a long time, be it motors, bearings or nozzles. A further decline in their cost is not really to be expected.
The Chinese also want to recoup their development cost (even Chinese engineers do not work for free). Given the modest sales figures of these machines (selling a few hundred might already be a major coup), there is simply not much room for further savings.
For the extreme low end, check out the Liteplacer (which comes as a kit). It might give you an idea about the lowest possible cost for the most rudimentary P&P with vision.
Max
-
How long will it be before something better comes along at a lower price?
Better at the same or slightly higher price would be a start.
The Chinese are too focused on lowering the price and thiking that placement speed is the only important benchmark, so corners get cut, especially with software design & implementation, and I can't see that changing any time soon.
There is such a huge gap between the more established makers and the newer Chinese machines that there has to be a market for something more polished and better supported at a slightly higher price.
There are plenty of areas for things to be improved, and none of them should be that hard to do given a bit of effort:
Auto nozzle change, taller parts, better feeders, better software, flying upward-looking camera.
Hi
I deliberately left out the "better at same to slightly higher price" simply because it opens the door to "higher price". You then get into all sorts of debates about the value of each item added. Exchange rates / prices where also was omitted since that gets into other debates. The omission of placement speed also was deliberate. The manufacturers are to focused on being able to advertise some sort of nutty placement speed number. A lot of things get compromised as a result.
I think I would score this one as "it could happen, but not at a noticeably lower price".
Thanks!
Bob
-
The cost of the cameras is not really relevant, they should not contribute much more than 100$ to the price. Some machines use a PC as a controller, so even from electronics perspective, there is not much room left (as you stated correctly).
More or less all components that go into such a unit have been on the market for a long time, be it motors, bearings or nozzles. A further decline in their cost is not really to be expected.
The Chinese also want to recoup their development cost (even Chinese engineers do not work for free). Given the modest sales figures of these machines (selling a few hundred might already be a major coup), there is simply not much room for further savings.
For the extreme low end, check out the Liteplacer (which comes as a kit). It might give you an idea about the lowest possible cost for the most rudimentary P&P with vision.
Max
Hi
I have looked at a number of other machines and pretty much decided that anything less capable then the TVM802 is not going to do a reasonable job. That is a *very* subjective evaluation and in no way a knock on the other gear on the market. Truthfully the TVM802 *barely* does what I think it needs to do. Barely doing the job is still OK.
Again, I think I would score this one as "not going to happen".
So far, nobody seems to believe that the $1500 replacement for the TVM802 is coming out next week :) I certainly do not disagree with anything anybody has come up with. It is good to see that most people have a fairly similar view of where this all is going.
Thanks!
Bob
-
I deliberately left out the "better at same to slightly higher price" simply because it opens the door to "higher price".
I'm not convinced that there is much benefit to lowering prices further - performance and usefulness will inevitably suffer.
If you can't afford a P&P now, chances are you don't really need one.
-
I deliberately left out the "better at same to slightly higher price" simply because it opens the door to "higher price".
I'm not convinced that there is much benefit to lowering prices further - performance and usefulness will inevitably suffer.
If you can't afford a P&P now, chances are you don't really need one.
Hi
Ok, one half of that one I would not agree with. (lowering price does not get more customers). I have never seen any market study on a piece of gear where volume did not go up as price goes down. Yes there may be some inelastic demand gizmo somewhere, but I don't think pick and place machines are it. Drop the price by half and you more than double the number of units sold (provided it's the same quality /type / capability machine). There is always a "if we add another machine will it pay for it's self" decision. Lower price gets more of those decisions turned into a purchase.
I'll score this one as "not going to happen". The real core argument is still that the quality requirement can't be met at a lower price.
Bob
-
Lets remember that the Chinese manufacturers are typically working on very slim margins. Either materials must go down in price, or their labor must go down in price, as the hours to put one together will likely not go down significantly to impact price. I don't see materials dropping, and wages in China are on the rise. Your best bet is to hope for currency swings to save you some money.
So yeah, as everyone is pretty much saying, the cost isn't going down anytime soon.
But from another perspective, consider the benefits and time / cost savings of having a reliable, well built PNP in your office. You should be able to recoup the cost of almost any small, well built machine rather quickly. At that point, any minor savings in purchase price are soon negated.
-
Lets remember that the Chinese manufacturers are typically working on very slim margins. Either materials must go down in price, or their labor must go down in price, as the hours to put one together will likely not go down significantly to impact price. I don't see materials dropping, and wages in China are on the rise. Your best bet is to hope for currency swings to save you some money.
So yeah, as everyone is pretty much saying, the cost isn't going down anytime soon.
But from another perspective, consider the benefits and time / cost savings of having a reliable, well built PNP in your office. You should be able to recoup the cost of almost any small, well built machine rather quickly. At that point, any minor savings in purchase price are soon negated.
Hi
Rather than get into all sorts of messy details, let's just say that not every machine gets used the same way in the same setting. Put another way, you can justify a $250,000 or a $1,000,000 machine with "it will pay for it's self". Been there done that.
"Well built" is pretty elastic and it will move you up from a TVM802 pretty fast. That's where the original "at a lower price" came in. Without that, you are on a very slippery slope. In some cases budgets do matter :)
I'll log this one as another "not getting cheaper any time soon, far more likely to go up in price".
I'm by no means saying that I disagree with any of the basic points made. For a lot of settings, you would be foolish to buy one of these over a machine costing 2 to 5 times as much. The delta in purchase price would quickly be paid back on a better machine that is kept full 80 hours a week 52 weeks a year. You would be banking the delta long before the setup turned 1 year old.
Bob
-
Hi
Rather than get into all sorts of messy details, let's just say that not every machine gets used the same way in the same setting. Put another way, you can justify a $250,000 or a $1,000,000 machine with "it will pay for it's self". Been there done that.
"Well built" is pretty elastic and it will move you up from a TVM802 pretty fast. That's where the original "at a lower price" came in. Without that, you are on a very slippery slope. In some cases budgets do matter :)
I'll log this one as another "not getting cheaper any time soon, far more likely to go up in price".
I'm by no means saying that I disagree with any of the basic points made. For a lot of settings, you would be foolish to buy one of these over a machine costing 2 to 5 times as much. The delta in purchase price would quickly be paid back on a better machine that is kept full 80 hours a week 52 weeks a year. You would be banking the delta long before the setup turned 1 year old.
Bob
Yeah, I agree with you UB. Buying any piece of equipment, be it a table saw, or a PNP, you will find may different products meeting many different needs. You need to fully understand your requirements, and purchase accordingly...
-
The TVM802A has been my first P&P. I have had it just a few weeks and it has been running an hour or two per week. I will go in full production in a week or two.
For the capital investment I need to make about 1500 units and I have already done about 10% of that up to now.
It has been a pleasant budget machine to work with. For me this machine compares to the experience I had with my first 3d printer. I have a commercial 3D printer that was about 4 times the price of an RepRap. It printed stuff with in the first 15 minutes. Even at 4 time the price it has paid it self of with in the first few months.
Getting back to P&P. I could have gone for the DIY version but I would have had more stress keeping it running and making sure production stays online.
Upgrading the software on this machine will make it much better. Not sure how far @ar__systems is with decoding the protocol. Would be interested in writing add on't that use movement of the machine but better camera's or higher components.
My question is, once I pay this machine off where do I go from there? Is there a $10K machine that has quick reel changing and software that is more forgiving. (At $10K and up unfortunately I don't think there will be Open-source options)
-
I will let yo know once my Qihe TVM920 shows up. I will post a complete review of everything from setup to first runs. Hopefully it will fill the gap between the TVM820 (fixed drag feeders) and the Neoden 4 (proprietary fixed feeders). I am confident that the interchangeable Yamaha style CL feeders used on the 920 will make all the difference.
Fingers crossed....
The TVM802A has been my first P&P. I have had it just a few weeks and it has been running an hour or two per week. I will go in full production in a week or two.
For the capital investment I need to make about 1500 units and I have already done about 10% of that up to now.
It has been a pleasant budget machine to work with. For me this machine compares to the experience I had with my first 3d printer. I have a commercial 3D printer that was about 4 times the price of an RepRap. It printed stuff with in the first 15 minutes. Even at 4 time the price it has paid it self of with in the first few months.
Getting back to P&P. I could have gone for the DIY version but I would have had more stress keeping it running and making sure production stays online.
Upgrading the software on this machine will make it much better. Not sure how far @ar__systems is with decoding the protocol. Would be interested in writing add on't that use movement of the machine but better camera's or higher components.
My question is, once I pay this machine off where do I go from there? Is there a $10K machine that has quick reel changing and software that is more forgiving. (At $10K and up unfortunately I don't think there will be Open-source options)
-
I will let yo know once my Qihe TVM920 shows up. I will post a complete review of everything from setup to first runs.
Oh yes, please do!
I'm looking for a smaller replacement for my somewhat ancient Dima Optimat and the TVM920 seems to be a strong contender. I'd love to know if it's as good as it looks.
-
Latest version of automatic PCB loading and unloading:
https://www.youtube.com/watch?v=on6gbG0N9Q0 (https://www.youtube.com/watch?v=on6gbG0N9Q0)
As you can see, it looks very different from the first version and runs much smoother; list of changes I made:
http://servokit.com/blog/2016/05/03/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-iv/ (http://servokit.com/blog/2016/05/03/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-iv/)
Regards, Axel
-
Close-up of PCB pickup and drop off:
https://www.youtube.com/watch?v=BgU4n0lcIpQ (https://www.youtube.com/watch?v=BgU4n0lcIpQ)
Regards, Axel
-
Hi @ ServoKit,
Looks great. Can't wait to see it make its way into your oven eventually.
Out of interest - what are you using for the 'initial lift, and drop actions' on the PCB carriage?
Some detail pics, parts list, would be appreciated.
Thnx
-
Details are here:
http://servokit.com/blog/2016/05/04/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-v/ (http://servokit.com/blog/2016/05/04/automated-pcb-handling-with-the-tvm802a-pick-and-place-machine-v/)
Regards, Axel
-
Surely a belt would be faster and cheaper than a leadscrew?
-
What would be the best way to calculate the needle-to-nozzle offset?
Camera to nozzles can be done with a peace of blue-tac. Press the nozzle into the blue-tac(Note XY), center camera over the blue-tac(Note XY). Deduct two coordinates to get off set.
-
Interesting!!! OpenPnP might be an option in the future...
https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs (https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs)
If only we can decipher the protocol... :-/O
-
@mike: I try to avoid belt drives when possible, components *may* be cheaper but overall the setup means more parts and more work (mounting the motor, regulating belt tension etc.) Also, for 250mm travel, speed is not an issue.
In the meantime I began working on the dispenser; the drive train of the PCB transport can be exploited for Y direction, adding support for X is easy.
Regards, Axel
-
Interesting!!! OpenPnP might be an option in the future...
https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs (https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs)
If only we can decipher the protocol... :-/O
It is done. I'm on vacation now, I'll post it when I get back. Next week.
-
Doing fiducial recognition with Raspberry PI and $25 camera.
Write-up w/ Python code: http://servokit.com/blog/2016/05/09/an-automatic-smt-solder-paste-dispenser-iii/ (http://servokit.com/blog/2016/05/09/an-automatic-smt-solder-paste-dispenser-iii/)
Regards, Axel
-
Interesting!!! OpenPnP might be an option in the future...
https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs (https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs)
If only we can decipher the protocol... :-/O
It is done. I'm on vacation now, I'll post it when I get back. Next week.
this would enable us to save all the machine settings in a file. There are so many numbers without a backup. Currently I save them as screenshots.
-
I did another run with the machine today. 600+ parts with 10 panel changes.
I had one attempt at picking a part not succeeding. Made a small adjustment to the tape peeler and all was good.
With the current settings I can start the job and do other things while it works. Saving me plenty of time.
Working without fid's is also not so bad. With my current set of panels I have used via's on each sub board to do the alignment. I used manual alignment for now and its amazing how each board differs.
I am really pleased with the machine and recommend it as a machine to start with. Cant wait to see what the TVM9XX offers. The removable reel holders sounds very tempting. :)
-
protoneer, I can confirm your experience 100%.
One question,
If a part is not pick successfully then the empty nozzle is dunked into the paste on the board.
This makes problems with further picking.
Does your machine detect empty nozzles ?
How do you clean the nozzles in that case ?
-
Mmmm.... there is a machine setting to check for picking issues.(uses a pressure sensor)
In my case the plastic cover did not remove in time for the nozzle picking the part. It then went straight from the reel to the reject bin. It also has a try limit of 5 attempts if I recall correctly.
Does anyone know of a vibration feeder that might work well with this machine?
-
Mmmm.... there is a machine setting to check for picking issues.(uses a pressure sensor)
In my case the plastic cover did not remove in time for the nozzle picking the part. It then went straight from the reel to the reject bin. It also has a try limit of 5 attempts if I recall correctly.
Does anyone know of a vibration feeder that might work well with this machine?
Hi
I think you could easily shorten that to "Does anyone know of a vibration feeder that works well". :)
Vibratory bowl feeders are tricky beasts. They tend to be a bit sensitive to the exact parts being feed. That makes them a pretty expensive item unless you are doing a *lot* of the same parts. When you go out to custom tool one (so it works well with your parts) you find that there are relatively few people who understand how to do one right. Those that do know generally are pretty fully booked ....Thus not just pricy, but expensive *and* long(er) lead time.
Of course you may be looking for a different sort of feeder ....
Bob
-
Mmmm.... there is a machine setting to check for picking issues.(uses a pressure sensor)
In my case the plastic cover did not remove in time for the nozzle picking the part. It then went straight from the reel to the reject bin. It also has a try limit of 5 attempts if I recall correctly.
...
I had the issue with the plastic cover too, but where can I find and activate this setting ?
-
I had the issue with the plastic cover too, but where can I find and activate this setting ?
Its on the Other tab of the machine settings. Can't recall the wording they used... (Was not very obvious |O)
-
@Harry, you have to activate it in the Machine tab settings ("Use pressure sensor") and for the individual part in the part list. The machine will try 5x to get the part right, then a dialog window comes up with an error msg.
Also, if the nozzle gets into the paste, that shouldn't happen; check the settings for distance from nozzle to PCB.
Regards, Axel
-
With the pressure sensor and smaller tips, a bit of blockage (e.g. a 0402 part that half-slips in or some paste that happens to get in the nozzle) can easily confuse pick-up sensor. Then it'll think every pickup is successful.
You can quickly check if that's working correctly by going to "manual" mode, turning on the vacuum + appropriate nozzle vacuum, and checking that the pressure indicator is off usually and on when you block the nozzle.
-
thanks :-+
that works great It just picks the next part.
I had this problem when the cover tape was not peeled off correctly (my fault, an adhesive tape was on the wrong side and glued on one of the rods).
Here is a new board, just to show the people what this fantasic machine is able to do.
The SMB diodes adjusted in Quick mode, the ICs in Accurate mode and all other parts without camera:
-
Interesting!!! OpenPnP might be an option in the future...
https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs (https://groups.google.com/forum/#!topic/openpnp/ZW4fu7Gy5vs)
If only we can decipher the protocol... :-/O
It is done. I'm on vacation now, I'll post it when I get back. Next week.
any news ? can't wait for this valueable information ;)
-
I have a TVM802B on the way right now and would also be quite interested in seeing the reverse engineered control protocol.
-
looks like we do not get it this way, so why don't we do it by ourself ?
Lets build a team of TVM802 users who are able to work with wireshark.
Wireshark has a nice function "follow TCP stream" which may be useful.
We have to capture the stream, make a little change in the TVM802 settings and then look for changes in the TCP stream.
I am ready to start this project. But the stream is quite long, so more people should work on it to get the information in reasonable time.
We could exchange our results here.
Harry
-
looks like we do not get it this way, so why don't we do it by ourself ?
Lets build a team of TVM802 users who are able to work with wireshark.
Wireshark has a nice function "follow TCP stream" which may be useful.
We have to capture the stream, make a little change in the TVM802 settings and then look for changes in the TCP stream.
I am ready to start this project. But the stream is quite long, so more people should work on it to get the information in reasonable time.
We could exchange our results here.
Harry
Hi
I have a copy of Wireshark if anybody wants to send me a TVM802 :)
===
Having been down this road once or twice:
Wireshark is free, so that's the easy part. The best way to watch things is with a second computer. That way you don't get in the way of the control process on the "main machine". The gotcha there is that you may need an ethernet switch that does port forwarding to see the whole process. Fortunately they are pretty cheap these days. They are also fairly easy to set up.
The objective is to interpret the transactions. That can be done in a number of ways. One is by writing code is incrementally updated as you figure out what is what. Ultimately you have a monitor program that can tell you "send: move head X = 332 Y= 211", "rec: head at location" or some such thing. To do that kind of code with a group, you need agreement on what you can or can't run on your machines. Visual Studio is one (free) option. The nice thing about a program is that it's quite happy to sit there for hours watching the dialog. It can tag the stuff it does not understand so you can come back and work it out later.
Lots of details .... MANY different ways to get the job done.
Bob
-
looks like we do not get it this way, so why don't we do it by ourself ?
sorry guys, I've been busy with other stuff that makes me money :(
pm me your email and I'll forward what I've got so far.
-
Unfortunately the code is very.... bad. My student wrote it and I did not have time to work on it since. That said, it does everything.
-
Unfortunately the code is very.... bad. My student wrote it and I did not have time to work on it since. That said, it does everything.
Could you just chuck it up on a Github repo or something? If you don't want to take the time, and wouldn't mind it being done, feel free to send me what you've got and I will post it for you.
Thanks,
Jason
-
looks like we do not get it this way, so why don't we do it by ourself ?
sorry guys, I've been busy with other stuff that makes me money :(
pm me your email and I'll forward what I've got so far.
Hi @Ar__systems, I have sent you a quick PM. I can setup the code on Github and then people can start working on this. We don't mind about the state of the code, we are use to cleaning up code... :D
By the way, thanks for willing to share this. Once this has been integrated into OpenPNP everyone is going benefit from an even better running machine.
By the way, I have been playing with 0402's the last few weeks and I am very surprised at how well the machine handles it. 0603 and 0402 are almost half the price of the 0805's. Nice cost saving over the long-run...
-
I sent the code to protoneer. He will set it up on Github or wherever, as he suggested :)
-
I sent the code to protoneer. He will set it up on Github or wherever, as he suggested :)
Thanks ar__systems !!!!
I have created a Github repository for it this project ---> https://github.com/TVM802/TMV802-Protocol (https://github.com/TVM802/TMV802-Protocol)
-
Anyone tested it yet? It should compile right away with Visual Studio Express.
-
Yes.. it runs. Also started looking at the code and from what I have seen its going to be very helpful. Thanks again for the contribution.
-
Hey guys,
Just wondering if anyone out there is aware of any purchases of the TVM920 machine and, if so, have there been any reviews or feedback on it?
-
Hey guys,
Just wondering if anyone out there is aware of any purchases of the TVM920 machine and, if so, have there been any reviews or feedback on it?
Hi
Looking only at the price, it appears to be closer to a Neoden 4.
Bob
-
Hi Uncle Bob,
Yes, similar to Neoden4, but with additional, and somewhat more reliable Yamaha, feeders. I intend buying one shortly and just wanted to know if anyone else out there in EEVblog land has also.
Hey guys,
Just wondering if anyone out there is aware of any purchases of the TVM920 machine and, if so, have there been any reviews or feedback on it?
Hi
Looking only at the price, it appears to be closer to a Neoden 4.
Bob
-
the TVM920 is (incl. feeders) about double the price of a TVM802.
Regarding the technical data the placement precision is the same.
If it makes sense to go for the TVM920 depends how well the 920 is adjusted by the factory.
The TVM802 comes with a "basic" adjustment which does not work at all and a manual which covers only a few simple functions in "chinese-english".
After receiving the 802 a lot of work has to be done to understand the machine, to enter all the stack coordinates and
to make all adjustments which usually takes a few days of work. Also every few weeks some adjustments have to be corrected for best placement precision.
This is fine, I don't complain, it is a very good balance between price and quality and I love my 802.
But if I go for a machine which costs >$10k then I would expect a much easier and faster setup. But I don't believe that the TVM920 is much different. Regarding the mechanics
its almost the same as the TVM802.
-
I created a JS script to convert Kicad Pos Files into TVM802 CSV files...
Code is available from : https://github.com/TVM802/KICAD-POS-TO-CSV (https://github.com/TVM802/KICAD-POS-TO-CSV)
Makes it a little easier to work with. Still need to manually set the reels to be used and so on.
-
Hi
Ok, it's down to decision time:
TVM802A
--or--
TVM802B
--or--
save the money and buy nothing
I've hopped back and forth between several use scenarios for this machine. I still may hop again. Right now I have it narrowed down to the "lab machine" and not the "short run production machine".
I've ruled out the "no vision" machines. They simply will not handle the 0.5 mm stuff I need to place. I've ruled out the > $7K machines. I simply do not have the need / budget / cash flow to justify one. The application is short run prototypes in the lab. No production volumes. No 10 panels of 64 boards in one run. If the machine runs 4 hours twice a week .... that's about it. I'd rather spend the time on setup and get it right than spend two days placing parts by hand and get it wrong. (Yes, I've priced out contract assembly ... $250 (setup + shipping parts out + shipping stuff back) x 20 jobs = $5,000 ... you just paid for the machine).
Is the A big enough for what I do? If I split jobs and swap reels, sure it is. If I cobble up a tray system and do this or that, I can handle the big IC's. Can I screw in nozzles every 30 minutes? Sure. It does begin to be a bit of a pain when you do all of that though. There also is some cost involved. I'd love to say that cost is not an issue, but it always is. It's also a lot more expensive to buy an A and then go back and get a B six months later.
What do I mean by 0.5 mm pitch parts? I include 0402 passives in that category, but not 0201's. Fine pitch flat pack packages in the 100 to 200 pin range are certainly in that category. 1 mm pitch BGA's are in. I'd love to think that 0.5 mm BGA's are in as well. Yes indeed, placement speed is not an issue. Doing this at "dead slow" with vision on -- not a problem at all. My time (for this exercise) costs nothing if it's used getting the machine set up. Tuning / tweaking / calibrating is an expected part of this. No, I won't come set up your machine for free :)
Enough background. Here's the questions:
If you have an A, would you get one again or would you go for a B?
If you have a B, would you get an A if you had it to do over again?
If you have either an A or a B, would you *not* get one if you had it to do again?
Do you have a non-vision machine that reliably places 0.5 mm pitch parts?
Then the same set of questions, with the qualifiers of my very low volume and possibly insane fine pitch needs ...
What's the consensus?
Bob
-
Do you have a non-vision machine that reliably places 0.5 mm pitch parts?
:popcorn:
My guess is that .5mm without vision would struggle in the 'reliability category. Can't wait to hear from actual user experience.
-
I have an A and will buy a second one anytime. For a B's price I would rather look at a bigger machine with removable feeders.
The TVM9XX sounds nice but I will only buy one if it still uses TCP. OpenPnp is just a matter of time.
If you have lots of diff parts per board the A looses to th B. My stuff uses lots of the same parts so I am happy with the number of reels.
-
Do you have a non-vision machine that reliably places 0.5 mm pitch parts?
:popcorn:
My guess is that .5mm without vision would struggle in the 'reliability category. Can't wait to hear from actual user experience.
Might be just about doable with mechanical centering, but buying a P&P machine without vision is just stupid.
-
I decided on the B because I felt like even if only 20% of the boards that I make would require me to change reels to fully populate them, that would be enough hassle to justify spending the additional money for the larger machine. Given how some people in this thread have reported that the feeders can be finicky to adjust/calibrate properly, I want to keep reel changes to a minimum and also keep some feeders dedicated to components that are very common (like 100nF, 1µF and 10µF caps).
I bought the PnP along with a stencil printer from a seller on Aliexpress and was able to negotiate the price down to a level similar to what it would have cost to buy them off of TaoBao. The next highest tier of machines like the NeoDen4 and the TVM920 cost around $10-11k including shipping with a similar amount of feeders, so that is more than 2x the price of the TVM802B. I could not justify spending that much more for the limited number of boards I see myself assembling in small production runs.
-
Hi
The TaoBao comment brings up another point:
Where to buy it / who to buy it from?
I don't speak / read / understand Chinese so TaoBao is out. Pretty much anything that I can navigate in English is fine.
Bob
-
I bought it from Aliexpress http://bit.ly/1Up1Nri (http://bit.ly/1Up1Nri). You can send the seller a message and ask for a better price. He was incredibly pushy about making a sale, but did ship it the next day after the payment went through.
-
I bought it from Aliexpress http://bit.ly/1Up1Nri (http://bit.ly/1Up1Nri). You can send the seller a message and ask for a better price. He was incredibly pushy about making a sale, but did ship it the next day after the payment went through.
Hi
I've been down the Aliexpress route on a few other similarly priced gizmos. Based on that experience, I'm waiting for the money to hit my bank before I dive in there.
Thanks!
Bob
-
I don't speak / read / understand Chinese so TaoBao is out. Pretty much anything that I can navigate in English is fine.
You don't have to speak/read Chinese to use Taobao. In this case, you are not even using Taobao to purchase, just to compare the prices.
Aliexpress is where the prices are inflated to cater international customer. The TVM802B is RMB22800 on Taobao from the best seller (usually the manufacturer themselves) without shipping fees which translate to around USD3,474.73. Add few hundreds USD for shipping fee into it, you shall get your best price. Obviously, they will argue they have different quality material, etc, for domestic and international market but I don't buy into that personally. Save few hundred USD, you probably can buy a low-cost stencil printer!
By the way, I don't read Chinese too.
-
I don't speak / read / understand Chinese so TaoBao is out. Pretty much anything that I can navigate in English is fine.
You don't have to speak/read Chinese to use Taobao. In this case, you are not even using Taobao to purchase, just to compare the prices.
Aliexpress is where the prices are inflated to cater international customer. The TVM802B is RMB22800 on Taobao from the best seller (usually the manufacturer themselves) without shipping fees which translate to around USD3,474.73. Add few hundreds USD for shipping fee into it, you shall get your best price. Obviously, they will argue they have different quality material, etc, for domestic and international market but I don't buy into that personally. Save few hundred USD, you probably can buy a low-cost stencil printer!
By the way, I don't read Chinese too.
If you can't read Chinese, how to do you search on TaoBao? If you try entering English in the search box you seldom get many appropriate responses. If you translate with Google translate, and paste in the Chinese, things are slightly better, but not much.
-
If you translate with Google translate, and paste in the Chinese, things are slightly better, but not much.
I would use Google translate in the 1st round. Yes, it will not be optimize BUT you do get some item that is exactly what you want.
From there, you got 2 choice:
- Use the "find similar" option on the correct item (mouse over the item, it will show that "find similar" button). You get a tonnes search result of the same item by clicking the "find similar" option.
- Copy the chinese words of the correct item. Paste it back into google translate and remove any word that too specific. Copy the simplified Chinese word and put it back into Taobao serach.
And in the case you want to something very specific like the TVM802B, just key in that. Your result will be correct and accurate.
After a while, I started to remember more and more Chinese character on the Taobao website menu.
Nothing is perfect but both works wonder for me.
-
If you translate with Google translate, and paste in the Chinese, things are slightly better, but not much.
I would use Google translate in the 1st round. Yes, it will not be optimize BUT you do get some item that is exactly what you want.
From there, you got 2 choice:
- Use the "find similar" option on the correct item (mouse over the item, it will show that "find similar" button). You get a tonnes search result of the same item by clicking the "find similar" option.
- Copy the chinese words of the correct item. Paste it back into google translate and remove any word that too specific. Copy the simplified Chinese word and put it back into Taobao serach.
And in the case you want to something very specific like the TVM802B, just key in that. Your result will be correct and accurate.
After a while, I started to remember more and more Chinese character on the Taobao website menu.
Nothing is perfect but both works wonder for me.
Hi
Chrome and Google Translate let me move around TaoBao pretty well. I can search and sort of figure out what's what. For small stuff, I'll do the "give it a try" stuff to order something. Maybe the translator makes a mistake and I'm out $35 due to a shipping glitch. Doing that with a "many thousands of dollars" item ... not so much.
I do agree that TaoBao gives you a better view of the internal market prices. Since it often decouples shipping from sale, the delivered price may be a bit "exciting". On something small, there is really no paperwork and things just come in. On something big, there will be some paperwork and a nice little bill after the fact collecting the government's share of the transaction. That's true if the paperwork is entered correctly on the other end. If not ... its a 4 hour drive to straighten it out.
Lots of fun
Bob
-
Bob, my decision was for the B and I would do it again.
I have put all the "standard" parts on the back stack, i.e. 1k 10k 100k 1n 100n 1u 10u Si-Diodes and N and P channel Mosfets.
The left stack is used for all special parts I need for a specific board.
This saves me a lot of time since changing reels in the TVM802 is not an easy job.
The precision of the 802 is very good, if carefully adjusted.
The long term precision is not good, so it needs readjustments from time to time.
This makes it an ideal machine for low quantities.
If I need boards in quantities >100 then I go to a professional manufacturer.
-
Bob, my decision was for the B and I would do it again.
I have put all the "standard" parts on the back stack, i.e. 1k 10k 100k 1n 100n 1u 10u Si-Diodes and N and P channel Mosfets.
The left stack is used for all special parts I need for a specific board.
This saves me a lot of time since changing reels in the TVM802 is not an easy job.
The precision of the 802 is very good, if carefully adjusted.
The long term precision is not good, so it needs readjustments from time to time.
This makes it an ideal machine for low quantities.
If I need boards in quantities >100 then I go to a professional manufacturer.
Hi
For the large volume stuff, I have other routes I can take. There is a full blown line upstairs at work. I'll admit that the quick change fancy feeders have me a bit spoiled. It's been a long time (early 1990's) since I had to "fiddle" with parts feeding. The calibration / adjustment stuff is something I'm familiar with from 3D printers. In a lot of ways the TVM802 is closer to my 3D printers than it is to the pick and place gear I've normally used. I'm more than willing to do a bit of fiddling in this case. Since everything that goes on the machine will be designed to work with the machine, I plan to keep the level of "challenge" as low as possible.
Bob
-
Hi
Here's a *very* basic question that I seem to have found multiple answers to ....
As best I understand it, all the 802's share a similar base structure. Reels hang off to various sides of the A or the B. The dimension *including* the reels appears to be 900 to 980 mm. That may or may not include this or that size reel. Hopefully that's correct so far.
How big is the footprint of the machine it's self? My guess is 600 x 600 but that's only a guess.
Does anything outside that footprint need to be supported? Again only a guess, the reel holders simply hang off the base.
If the 980 is correct, does that include any room for a component reel? If so how big a reel?
I'm looking at bench space and the whole "where do I put this beast" part of the process.
Thanks!
Bob
-
I'll measure when I'm in the office.
For now, if you have flaky performance of the visual system, I recommend checking if cameras are in focus. You can't really see that from the tiny pictures of the original s/w. Instead, open the camera in "Camera" application. There you can see full frame and easily see if cameras are focused or not. Both cameras are glued with some red stuff, you need to peel it off before you can turn the focusing rings.
-
for some reason I had to stop a pick&place job, and
want to continue it later at the same position.
How can I run the job from a specific line, not just from beginning ?
Is this possible with the TVM802 ?
-
if you did not close the application the 'done' ticks remain in place. So if start again it will ask you if you want to continue or start over. If you did close it, you can manually set previously populated components as 'done'
-
if you did not close the application the 'done' ticks remain in place. So if start again it will ask you if you want to continue or start over. If you did close it, you can manually set previously populated components as 'done'
WTF? so if it crashes or there's a power failure, you lose the job state? - typical poor Chinese software design.
And of course it should be more than just a case of where it got to in the job, as on errors it ought to carry on and do whatever else it can before stopping, so you can fix all the issues in one visit, but that's a whole other issue....
-
Hi
We run a *lot* of shop floor PC's on UPS's simply to take care of the power failure issue. Loosing state / data when the lights go out is not just a Chinese issue...
Bob
-
thanks ar__systems.
The reason was home made. The 1uF reel got empty during the job, I did not have a new one, so I changed the program to use the 2,2uF reel which also works in my circuit.
When this modification was done, I did not know how to start from the last position.
Marking already populated components as 'done' works fine, but there is another problem now.
When I restart the software then the sort order of the components is different, no idea how the software sorts it.
This makes it difficult to find all the already populated components.
-
Hi
Here's a *very* basic question that I seem to have found multiple answers to ....
As best I understand it, all the 802's share a similar base structure. Reels hang off to various sides of the A or the B. The dimension *including* the reels appears to be 900 to 980 mm. That may or may not include this or that size reel. Hopefully that's correct so far.
How big is the footprint of the machine it's self? My guess is 600 x 600 but that's only a guess.
Does anything outside that footprint need to be supported? Again only a guess, the reel holders simply hang off the base.
If the 980 is correct, does that include any room for a component reel? If so how big a reel?
I'm looking at bench space and the whole "where do I put this beast" part of the process.
Thanks!
Bob
Hi
It's Bob the second request guy :)
Does anybody have some size info? I'm about two days out from placing an order and really need to get all the T's crossed.
Thanks!
Bob
-
Bob,
the footprint is 580mm wide and 640mm deep.
You need a minimum space of 1000x1000mm including the reels (TVM802B, I don't know about the A).
My table is 2000mm wide which is fine for the TVM802 and the monitor/keyboard,
and 800mm deep but the distance to the wall is 200mm so I get the 1000mm needed by the machine including reels.
For the B you need some more space behind the machine. Beding forward to install the back reels is a real pain for the back.
Harry
-
Bob,
the footprint is 580mm wide and 640mm deep.
You need a minimum space of 1000x1000mm including the reels (TVM802B, I don't know about the A).
My table is 2000mm wide which is fine for the TVM802 and the monitor/keyboard,
and 800mm deep but the distance to the wall is 200mm so I get the 1000mm needed by the machine including reels.
Harry
Hi
Ok, that's exactly what I needed. If I can space a bench off the wall, it all works fine for me.
Thanks!
Bob
-
Guys:
My software variant for my 802A is V1.22... ! does anyone have a later variant they can share with me please
Thx
Steve
-
I have decompiled v2.12 of the original control software and made a Visual Studio project for it. The code is quite understandable with some tinkering. If anyone here is interested, PM me your
github bitbucket username and I'll invite you to the repository.
-
Hi
Well, the order for the 802D is making it's way through the every mysterious bowels of Alibaba. We'll see how it all works out....Thanks to everybody for helping out with the decision.
Fingers Crossed !!!
Bob
-
Bob..
What's the difference with the 'D' variant
Thanks
Steve
-
Bob..
What's the difference with the 'D' variant
Thanks
Steve
Hi
That was what had me all tied up in knots a few weeks ago :)
If you buy from one source you pick between the C and the D. From the other main source, you pick an A or a B. The machines are identical (C = A and B = D). It's just some strange form of marketing they are doing.
Bob
-
Ok that makes sense now... I have an 'A' variant that's almost a 'B' variant (I have added the extra feeders already ) I am just waiting on the Juki nozzle adaptors to arrive for the head motors. I have to say its nearly paid for itself in less than 12 months!
Thx
Steve
-
Ok that makes sense now... I have an 'A' variant that's almost a 'B' variant (I have added the extra feeders already ) I am just waiting on the Juki nozzle adaptors to arrive for the head motors. I have to say its nearly paid for itself in less than 12 months!
Thx
Steve
Hi
My conclusion was that if I got an A, I could easily spend the next year turning it into a B. Simpler / easier / (maybe) cheaper to just get the B up front. In the end, I could see only two disadvantages to the B. One obvious issue is the cost, the other was the footprint. I hope it will fit where I'm putting it. The money now vs money later is a "who knows" sort of thing. After *many* months of back and forth (clearly chronicled in this thread) , it was time to finally make a "put up or shut up" decision.
Bob
-
Should have my TVM920 delivered in the next day or two. Will post info / pics as I get it uncrated / setup / test runs complete.
Looking forward to getting a good review posted about this machine.
-
Exciting news @tts!
If anyone else out there has also purchased a TVM920 it would be great to to hear your story. Like I've said previously, I intend buying this same machine - shortly.
Should have my TVM920 delivered in the next day or two. Will post info / pics as I get it uncrated / setup / test runs complete.
Looking forward to getting a good review posted about this machine.
-
All the known issues not withstanding, I actually don't like the layout of the B. A is way more sensible. The up camera is at the front where you are more likely to have the trays, and that space is kind of useless anyway since down camera can't reach it.
But on B the up camera is in the prime location for feeders. I also hate the location of the back feeders on B. I'm going to mill a new waste tape slot on the east site and get rid of the back feeders.
-
I have decompiled v2.12 of the original control software and made a Visual Studio project for it. The code is quite understandable with some tinkering. If anyone here is interested, PM me your github bitbucket username and I'll invite you to the repository.
very well done, thank you !
Now it is possible to add a backup function.
It shows all TCP commands and its easy to see how they are working.
-
Lets hope the owner of this fine machine, gives us some more info. I'm told its extremely good.
https://www.youtube.com/watch?v=uNk2WvenV1A (https://www.youtube.com/watch?v=uNk2WvenV1A)
-
Hi
If that machine is in the same price range as a TVM802A or B, it's a winner. Looking at the feeders, and based on earlier comments, I'd bet it is a bit more expensive.
Bob
-
i believe its about $6000 without feeders.
-
Forgive my ignorance, but what machine is that exactly ? (could not find any reference in the video etc)
-
Forgive my ignorance, but what machine is that exactly ? (could not find any reference in the video etc)
its the 802/802B's big brother
http://www.qihekj.com/en/product/html/?45.html (http://www.qihekj.com/en/product/html/?45.html)
-
TVM920 - new thread
https://www.eevblog.com/forum/manufacture/pick-place-machinetvm920/ (https://www.eevblog.com/forum/manufacture/pick-place-machinetvm920/)
-
Thanks to both of you ;)
-
most of my standard parts are 0603.
I use the nozzle which is recommended for this size.
But there is a big problem with vibrations.
The TVM802 would easily be able to place correctly at 100% speed.
But on the way from pickup to place position the part moves a little bit on the nozzle, probably due to vibration.
So I use only 40 or 50% max speed which is boring.
What can be done to solve this problem ?
Would it help to use a stronger vacuum pump ?
Or should I use a new nozzle, but it is just 2 month old, maybe 50 board.
Or what else could be done ?
-
most of my standard parts are 0603.
I use the nozzle which is recommended for this size.
But there is a big problem with vibrations.
The TVM802 would easily be able to place correctly at 100% speed.
But on the way from pickup to place position the part moves a little bit on the nozzle, probably due to vibration.
So I use only 40 or 50% max speed which is boring.
What can be done to solve this problem ?
Would it help to use a stronger vacuum pump ?
Or should I use a new nozzle, but it is just 2 month old, maybe 50 board.
Or what else could be done ?
Hi
Unless the pump is broke, it is unlikely you will get a lot by switching it out. Going from 80% of an ATM to 100% of an ATM isn't that big a deal.
Take a look at the tip of the nozzle with a good microscope. If it is damaged, you will be able to see the problem. Check the tip and the seal.
Vibration can come from a few dozen sources. One common one is a micro stepping controller that does not do it's job well. A bit of Google investigation on the controller part number will help you dig into that one.
Bob
-
most of my standard parts are 0603.
I use the nozzle which is recommended for this size.
But there is a big problem with vibrations.
The TVM802 would easily be able to place correctly at 100% speed.
But on the way from pickup to place position the part moves a little bit on the nozzle, probably due to vibration.
So I use only 40 or 50% max speed which is boring.
What can be done to solve this problem ?
Would it help to use a stronger vacuum pump ?
Or should I use a new nozzle, but it is just 2 month old, maybe 50 board.
Or what else could be done ?
Hi
Unless the pump is broke, it is unlikely you will get a lot by switching it out. Going from 80% of an ATM to 100% of an ATM isn't that big a deal.
or just under-powered, or too much leakage. Remember we're talking Chinese kit here, so maybe they didn't fit something that was really up to the job. Another possibility is there's some ripple in the vacuum, in which case a reservoir before the vacuum valve may help.
Probably worth experimenting with on the basis that it's a potentially easy fix
Take a look at the tip of the nozzle with a good microscope. If it is damaged, you will be able to see the problem. Check the tip and the seal.
Or maybe roughen the end surface very slightly to increase friction
Also could be something silly like it's not waiting quite long enough during the pick befopre moving off, or not waiting long enough for it to stop before the camera images it.
-
most of my standard parts are 0603.
I use the nozzle which is recommended for this size.
But there is a big problem with vibrations.
The TVM802 would easily be able to place correctly at 100% speed.
But on the way from pickup to place position the part moves a little bit on the nozzle, probably due to vibration.
So I use only 40 or 50% max speed which is boring.
What can be done to solve this problem ?
Would it help to use a stronger vacuum pump ?
Or should I use a new nozzle, but it is just 2 month old, maybe 50 board.
Or what else could be done ?
Hi
Unless the pump is broke, it is unlikely you will get a lot by switching it out. Going from 80% of an ATM to 100% of an ATM isn't that big a deal.
or just under-powered, or too much leakage. Remember we're talking Chinese kit here, so maybe they didn't fit something that was really up to the job. Another possibility is there's some ripple in the vacuum, in which case a reservoir before the vacuum valve may help.
Probably worth experimenting with on the basis that it's a potentially easy fix
Take a look at the tip of the nozzle with a good microscope. If it is damaged, you will be able to see the problem. Check the tip and the seal.
Or maybe roughen the end surface very slightly to increase friction
Also could be something silly like it's not waiting quite long enough during the pick befopre moving off, or not waiting long enough for it to stop before the camera images it.
Hi
The "obvious" thing that neither one of us came up with:
Go to a bigger nozzle, recommendations aren't always right.
That may or may not be possible with a specific small part.
On the B model the nozzles are more or less Juki standard. You probably can find other ones. The rubber on a real Juki nozzle might have more grip.
Bob
-
I place loads of 0603 and 0402 parts with the smallest nozzle, at 100% speed.
Some ideas:
Do make sure you don't have some paste (or even a component!) stuck inside the nozzle. Use a needle to clear it. I had to do it once to make the pressure sensor work well again.
Does the prick needle ever get stuck on one of the reels? If the pickup location is not set correctly (ex slightly too close to one of the ends) that will hurt you. Fixed with re-calibrating the position of the pickup.
Check that your tape collector is not too tight. If it's too tight it will move the tape just a bit between the "prick" and pickup steps, which could mean you're not 100% on the part with your nozzle.
The machine requires a really solid surface, if operating at 100% speed my table is shaking quite a bit but doesn't seem to affect placement too much.
-
I place loads of 0603 and 0402 parts with the smallest nozzle, at 100% speed.
...
do you use the 502 or 503 nozle for your 0402 and 0603 parts ?
-
Or maybe roughen the end surface very slightly to increase friction
No, don't try that. Rough surface means more leakage. Instead you need to polish the surface as much as possible.
On my machine I had the problems with vibration as well and to avoid any guesses about the quality of the built in pump I simply use a big ass external vac pump that I keep in the next room.
-
Or maybe roughen the end surface very slightly to increase friction
No, don't try that. Rough surface means more leakage. Instead you need to polish the surface as much as possible.
On my machine I had the problems with vibration as well and to avoid any guesses about the quality of the built in pump I simply use a big ass external vac pump that I keep in the next room.
If a small amount of leakage makes a difference then the vacuum flow rate is too low.
-
...
On my machine I had the problems with vibration as well and to avoid any guesses about the quality of the built in pump I simply use a big ass external vac pump that I keep in the next room.
When buying a vac pump for that purpose, what are the right parameters for airflow, vacuum and others ? Or buy the biggest one available ?
-
...
On my machine I had the problems with vibration as well and to avoid any guesses about the quality of the built in pump I simply use a big ass external vac pump that I keep in the next room.
When buying a vac pump for that purpose, what are the right parameters for airflow, vacuum and others ? Or buy the biggest one available ?
I use a 1/4hp rotary vane pump from Gast with a 5gal storage tank. It is enough to supply vacuum for a couple of machines in addition to a few manual pickup tools on my benches. It holds about 26in of Vac but I am not sure what the flow rate is. I was more concerned about the depth of vac than the flow rate. The deeper the vac, the more stable the part is. The flow rates are rather tiny on P&P machines.
-
The flow rate is tiny that is true, but that parameter determines how fast the pump recovers the vacuum. Every time a pick up occurs the pump has to swallow a bit of air. (The air inside the tubing from the solenoid to the nozzle.) The faster it does it, the faster the vacuum at the pick up nozzle reaches the maximum. I think the main issue with built-in pump is actually low flow rate. The max vac level is not that bad, I think it was like 560mmHg, and my big one reaches 670mmHg. So only 20% less holding power. But it is very (relatively) slow at recovering vacuum. Keep in mind that if it takes a second to recover, that's almost as long it takes for the part to travel from pick up to placement.
-
Storage tank is actually a very good idea. It allows to get away with a less capable pump.
-
Storage tank is actually a very good idea. It allows to get away with a less capable pump.
I am very happy I added that to the system - and it was not expensive at all. I did have to find a high-flow passive vacuum check valve to prevent the vac from pulling air in from the pump when it cycles off. Regular check valves are useless.
These are fantastic : http://www.vaccon.com/SearchResult.aspx?KeyWords=VCV (http://www.vaccon.com/SearchResult.aspx?KeyWords=VCV)
-
so it may be sufficiant to add a storage tank to the existing pump in the 802 ?
-
I always thought the idea of a vacuum "storage" tank a little odd....
Is it full of "empty"...?
-
I always thought the idea of a vacuum "storage" tank a little odd....
Is it full of "empty"...?
It's completely full of nothing.
so it may be sufficiant to add a storage tank to the existing pump in the 802 ?
Probably. Does the pump run continuous when the machine is on? It kind of acts like the output filter cap on a power supply - smooths out the vacuum.
-
it runs as soon as a pick&place job is started.
I cannot find a storage tank for vacuum, but its available and cheap for compressed air.
These tanks should also work for vacuum ?
-
Any compressed air tank is fine - that is what I use. It's actually a tank from the air brake system on a big truck.
-
Not sure of the TVM802, but on the new TVM920, Qihe recommends (and supplies with the machine) to install an o-ring in the groove of the nozzle before installing in nozzle holder on machine. The holders on the TVM920 are magnetic, so not sure if the TVM802 is the same. They claim (and I agree) that any leakage makes a difference when trying to move fast and keep parts in place.
-
yes, the 802 also has the o-ring in the groove of the nozzle.
I looked for some vac pumps,
Most of them do 26in (abt 85% vacuum),
some pumps (oil) go down to 99,8% vac. This is a very good vac, but I'm not sure if the parts and tubes in the TVM802 can handle that. So I will probably go with the 26in pump.
-
Not sure of the TVM802, but on the new TVM920, Qihe recommends (and supplies with the machine) to install an o-ring in the groove of the nozzle before installing in nozzle holder on machine. The holders on the TVM920 are magnetic, so not sure if the TVM802 is the same. They claim (and I agree) that any leakage makes a difference when trying to move fast and keep parts in place.
I think you are confusing things here. There is no magnetic holder per se. The rotational steppers however are magnetic and will hold the nozzle in place, but that's just side effect. o-ring is a must. Everything I said before was with o-rings in place.
-
yes, the 802 also has the o-ring in the groove of the nozzle.
I looked for some vac pumps,
Most of them do 26in (abt 85% vacuum),
some pumps (oil) go down to 99,8% vac. This is a very good vac, but I'm not sure if the parts and tubes in the TVM802 can handle that. So I will probably go with the 26in pump.
There is little pressure difference between 85 and 99% vacuum. Going from 0 to 85 % is 13PSI difference. from 85 to 100 - only 2PSI. It is not worth it, but the tubes will hold just fine.
-
If 26in does not work, your machine is bad. My machine ill run full speed 3600CPH at 20in just fine.
-
yes, the 802 also has the o-ring in the groove of the nozzle.
I looked for some vac pumps,
Most of them do 26in (abt 85% vacuum),
some pumps (oil) go down to 99,8% vac. This is a very good vac, but I'm not sure if the parts and tubes in the TVM802 can handle that. So I will probably go with the 26in pump.
There is little pressure difference between 85 and 99% vacuum. Going from 0 to 85 % is 13PSI difference. from 85 to 100 - only 2PSI. It is not worth it, but the tubes will hold just fine.
Hi
Speaking of tubes, simply adding a roll of tube to a system like this is a "poor man's" way to add a reservoir.
Bob
-
I can only speak for the TVM920, which obviously is different. The TVM920 actually does have magnetic nozzle holders, as the holder is not directly connected to the stepper motor. It is offset with a belt drive, with about a 3:1 reduction for more accuracy in rotational angles.
But glad to hear the o-rings are in place on the 802's.
Sorry if I added some confusion....
Not sure of the TVM802, but on the new TVM920, Qihe recommends (and supplies with the machine) to install an o-ring in the groove of the nozzle before installing in nozzle holder on machine. The holders on the TVM920 are magnetic, so not sure if the TVM802 is the same. They claim (and I agree) that any leakage makes a difference when trying to move fast and keep parts in place.
I think you are confusing things here. There is no magnetic holder per se. The rotational steppers however are magnetic and will hold the nozzle in place, but that's just side effect. o-ring is a must. Everything I said before was with o-rings in place.
-
Speaking of tubes, simply adding a roll of tube to a system like this is a "poor man's" way to add a reservoir.
Flow rate might be an issue if the tube is small.
-
It is offset with a belt drive, with about a 3:1 reduction for more accuracy in rotational angles.
That is very interesting. I presumed more or less the same design, wrongly.
I wonder if that is really necessary. Without the belt we have 200*8 steps. For a 10mm chip that translates into 0.04mm shift if the angle is off by one full step. But theoretically it should not be off by more than 1/2 step if visual system works 100% reliably. If it does not then higher resolution for angles does not really help anything at all.
I'm just wondering if that makes any difference in some circumstances.
-
I think the reasons for reduction are as follows:
#1, Camera can get a 20mm x 20mm chip in view, so with a larger chip, you have less resolution using a 200 step motor. Micro stepping without inertial load on stepper can be marginal at best.
#2, Run out. Steppers are not known for their shaft tolerances / run out. They are designed to be a driver, not a precision bearing assembly Hence the typical use of flexible couplings. Even a small amount of run out at the motor shaft translates into large problems once a nozzle coupling is added to the end.
#3, Compact assembly height. The TMV920 nozzle holder assembly is within (<5mm) of being vertically even with the front of the motor shaft, It shortens the length of the typical assembly by almost an 20 - 25mm. Shorter assembly translates to smaller run out errors if present.
#4, No need for hollow shaft steppers. Part of the problem with using small NEMA8 steppers for rotation is the limited torque produced, and the drag presented by the rotary coupling used for vacuum connection. This mitigates the problem by increasing output torque using the belt drive reducer, while increasing resolution as a byproduct.
Just my thoughts.
It is offset with a belt drive, with about a 3:1 reduction for more accuracy in rotational angles.
That is very interesting. I presumed more or less the same design, wrongly.
I wonder if that is really necessary. Without the belt we have 200*8 steps. For a 10mm chip that translates into 0.04mm shift if the angle is off by one full step. But theoretically it should not be off by more than 1/2 step if visual system works 100% reliably. If it does not then higher resolution for angles does not really help anything at all.
I'm just wondering if that makes any difference in some circumstances.
-
I looked into the TVM802 and found two pumps, I think there are two vac pumps, one for each nozzle.
Does anybody know how that works and are these pumps also use for blowing ?
To connect an external vac pump, which is the right tube ?
-
I don't remember about blowing, for the moment I left both pumps connected, but plugged their vacuum ends. Both of them can create positive pressure as well, I presume only one is used.
See the attached pic. I used whatever fittings I had at the moment, I guess it can be done in a neater way :)
-
thanks for the picture,
so you simply connected the vac to the two tubes going to the PnP-head.
I will try the same.
Regarding vibrations, I did some tests today and found that it works better with 100% speed.
Maximum vibrations are at about 70% speed.
I am now using 100% for R and C which works well.
The only thing which does not work with 100% (and not with 50%) are SOT23 cases. These parts
"jump" on the nozzle (No. 503) in any direction, even flipping upside down. I have to go down to 20% to get them stable.
-
Creating a new thread specific to this particular machine. TVM802B is mostly identical but has more feeders.
Update on TVM802B: The machine comes with Juki / Juki-compatible nozzles. TVM802A has screw-in nozzles like in Neoden TM220/240 and similar.
Cost of machine as purchased: $3,400 including shipping from China to Germany.
Manufacturer: http://www.qihesmt.com/sell-2403372-pick-and-place-machine-vision-mounter-qihe-tvm802a-qihekj-qihe-smt-machine.html (http://www.qihesmt.com/sell-2403372-pick-and-place-machine-vision-mounter-qihe-tvm802a-qihekj-qihe-smt-machine.html)
Blog post re. purchase on Ali Express, importing & customs: http://servokit.com/blog/2016/03/03/pick-place-tvm-802-a-arrived-today/ (http://servokit.com/blog/2016/03/03/pick-place-tvm-802-a-arrived-today/)
Blog post with unboxing (plenty of hi-res images): http://servokit.com/blog/2016/03/04/unboxing-the-pick-and-place-tvm802a/ (http://servokit.com/blog/2016/03/04/unboxing-the-pick-and-place-tvm802a/)
Blog post with setting up the machine: http://servokit.com/blog/2016/03/05/tvm802a-pick-and-place-setup/ (http://servokit.com/blog/2016/03/05/tvm802a-pick-and-place-setup/)
Will post more information and experiences as we go along.
Regards, Axel
Hey Axel! i see you put your TVM802A up for sale! was there an issue with the TVM802A?
Dave
-
I was hoping biddings were still going but it seems it sold for 2400€, crap.
-
Can anyone tell if the TVM802A can do double-sided boards? I could make a holder to fixate the board when populating the back side, but the board surface would move up by about 2cm. Can the TVM802A handle this, or would I have to modify the baseplate?
Max
@prpplague, @Koen: On the German forum, Axel stated he wanted to focus on SW development. There seems not to have been anything wrong with his TVM802A. I was among to ones who were too slow :(
-
You can, the board is held in place by the edges. If you don't place too close to the edges, you can do double sides.
-
Can confirm, no problem with double-sided population, as long as you don't have components near the edges or have a panel. We do 0402 and other components on both sides without issues.
Note the TVM802X placement becomes quite poor if the board is not sitting all the way down / is a few mm raised (even with the appropriate "board height" program setting).
-
How much edge distance is needed?
-
Unfortunately, Axel (the Servokit guy) has not only sold his TVM802A, but also removed his blog and his YouTube videos, which is a pity. Looking at another YouTube video (https://youtu.be/g8IjoHn_oEg?t=38), I guess that the edge is about 3mm wide (for our non-EU friends: about 120 mil).
Does the TVM802A offer an arbitrary order of placement? I have some caps that are 7.7mm high. If they are placed last, I suppose that this should not be a problem, but even then, I need to set the proper placement order on the panel: the ones furthest to the front being placed first, the ones furthest to the back last. Is the SW capable of this?
Max
-
The software does not support arbitrary order, it always reorders as it pleases, and the logic is not clear.
What happened with Axel? That's unfortunate indeed.
-
Axel has stated (on the German forum) he wanted to focus on SW development. His company seems to be making money with barcode printing and/or scanning solutions, and his hardware excursion might have veered to far off course. Or maybe he simply realized that days are limited to 24h.
The height limit is a problem. Any suggestions how to get around it? Splitting the placement into several jobs?
Oh, and one more thing: the machine is stated to be L 980mm*W 655mm*H 285mm. Does this include the reels? I have rather limited space in my basement :(
Max
-
The only proper solution is s/w modification. I was placing 5mm switches, and even that is a problem. If it drops it for some reason over the feeder block it creates a huge mess as it tries to go to the discard location. It crushes the button and tapes in the process.
-
S/W modification would mean replacing the included S/W with something like OpenPNP (or some self-written stuff)? Not exactly my favourite. But the only alternative seems to be the SmallSMT, which has 10 mm.
Would it help to shorten the metal shield which is between the placement heads and the downward looking camera? This picture suggests so: https://www.mikrocontroller.net/attachment/287193/tvm802a-max-part-height-6.jpg. (https://www.mikrocontroller.net/attachment/287193/tvm802a-max-part-height-6.jpg.)
Max
-
The lowest point is actually the transport needle, you can see sticking out just below the bolt.
-
does anybody know how to configure the TVM802 to make panels ?
it is sufficient to define multiple ficudials for each board in the panel, or is some
other configuration required ?
-
I'm not quite sure *why* DHL has so much fun with TVM802B's. Like others, mine spent a *lot* of time in limbo during the delivery process. I finally gave up on the delivery and drove over to the depot to pick it up. Yikes is this thing HEAVY !!!
Back to figuring out what goes where ....
-
does anybody know how to configure the TVM802 to make panels ?
it is sufficient to define multiple ficudials for each board in the panel, or is some
other configuration required ?
It is pretty simple. Board coordinates of fid are the same for each board, but phisical coords have to be defined separately for each board.
-
does anybody know how to configure the TVM802 to make panels ?
it is sufficient to define multiple ficudials for each board in the panel, or is some
other configuration required ?
It is pretty simple. Board coordinates of fid are the same for each board, but phisical coords have to be defined separately for each board.
thanks, I will try it tomorrow.
-
wow! anyone know what happened to axel and servokit? the website disappeared, and all the videos on youtube have been deleted!
dave
-
wow! anyone know what happened to axel and servokit? the website disappeared, and all the videos on youtube have been deleted!
dave
It seems that Axel has stated (on the German forum) he wanted to focus on SW development and sold the TVM802 machine. But why he deleted all the files and videos that were a good source of related information is something that only he can say :(
-
Has anyone already converted the nozzles from TVM802A to Juki using an adapter?
Paul.
-
Has anyone already converted the nozzles from TVM802A to Juki using an adapter?
Paul.
Hi
On a related topic:
Has anybody tried the other 4 common Juki nozzle sized on a TVM802? Some of them are mighty small ....
Bob
-
Have any of the TVM802B users got any comments on the feeder system?
Are you breaking cover tapes? Does it relaibly feed.
And anyone doing 0402?
-
Have any of the TVM802B users got any comments on the feeder system?
Are you breaking cover tapes? Does it relaibly feed.
And anyone doing 0402?
yes, it feeds reliably, no problem with cover tapes. But needs some adjustment from time to time.
If something goes wrong, the 802B stops and beeps until you fix the problem.
Most problems are self made. With every board I learn a bit and it works better and better.
0402, can work, but I would not do it as my standard parts, 0603 are fine.
For my 0402 parts I always use the camera, even for R and Cs.
Before placing a batch of boards I test the adjustment by placing the most difficult part to a board with double-sided adhesive tape,
a QFN or TQFP to check for any offsets.
Usually the camera rotation needs adjustment. Once adjusted it remains stable for the full batch.
On another day it may need readjustment.
The cam rotation is the only value I adjust regularly, all the other values look to be stable over weeks.
I put a videao on youtube this week which shows some close-up views of the pick&place head beginning at minute 2:30
https://www.youtube.com/watch?v=VsGQ3Gwg9xQ (https://www.youtube.com/watch?v=VsGQ3Gwg9xQ)
-
Thanks, is the machine running quite slowly in this video?
-
Thanks, is the machine running quite slowly in this video?
most parts are placed with 100% speed.
-
Im curious aobut your stencil printer?.. Is that something you made yourself?
-
The stencil looks to me like the one sold by CIF
http://cif.fr/en/manual-operated-stencil-printer-for-smd/292-machine-a-serigraphier-manuelle-pour-cms-tecprint-400-et-600.html (http://cif.fr/en/manual-operated-stencil-printer-for-smd/292-machine-a-serigraphier-manuelle-pour-cms-tecprint-400-et-600.html)
It looks nice, but I think is missing/lacking vertical separation from the PCB?
-
you are right, its the tecprint 500 from CIF.
Last year I tried it with one of these $400 chinese printers. The problem was that its only usable with framed stencils
and its very hard to get this size of frame outside China.
The tecprint can be used with unframed stencils of any size. Two springs keep the stencil flat.
This saves costs for the stencil since I can use smaller ones.
and yes, it is not a fully vertical separation, it uses a hinge, but the results are fine down to 0,5mm pitch.
-
How much did you pay for Tecprint 500 from CIF?
-
How much did you pay for Tecprint 500 from CIF?
it was 1545 Eur
-
What is it's best price without feeder?
-
Thanks, is the machine running quite slowly in this video?
Yes. I would say it is 50% speed.
-
Hi All
new guy here :)
Setting up my 1st pick & place machine - TVM802B
The DVD came with
SucfaceMount802-EN-V2.23Beta2
(not my typo btw) which does not open in windows 7 pro or windows 10.
I just get the 'allow this app' question, then nothing happens.
Probably I'm missing something obvious, its been a few years since I have set up anything like this.
If anybody has some hints or step-by-step instructions or could send me another version of the program that works, I'd really appreciate it.
Thanks
Andrew
-
Hi Andrew
I'd try contacting QiHe directly.
Speak with daisy - I was just speaking with her and let her know about your issue.
Her Skype address is: live:daisyqiu1994
I don't use the 802, but have just ordered a TVM920 model from them.
Good luck!
Hi All
new guy here :)
Setting up my 1st pick & place machine - TVM802B
The DVD came with
SucfaceMount802-EN-V2.23Beta2
(not my typo btw) which does not open in windows 7 pro or windows 10.
I just get the 'allow this app' question, then nothing happens.
Probably I'm missing something obvious, its been a few years since I have set up anything like this.
If anybody has some hints or step-by-step instructions or could send me another version of the program that works, I'd really appreciate it.
Thanks
Andrew
-
Thanks Thommo
I did try to email them 1st, but I got Jim who replied -
"You can find the software in the CD
The software in the CD"
which wasn't really the answer I was looking for :)
I've never used skype, would you have an email address for Daisy?
cheers
Andrew
-
No worries about Skype. They use it for chatting, no worries about making phone calls :)
Skype is easy to use, if you don´t want it on your computer, install the app on your smartphone and remove it once you´re done.
Max
-
I set up skype and got hold of Jim, who sent me the same files:
SucfaceMount802-EN-V2.23Beta2
but still doesn't work. Double-clicking the SucfaceMount.exe gets the computer thinking for a few seconds and then nothing.
Is anybody using this version? Anything special to do or just double click and away it goes?
Does the TVM802B need to be connected and on when the software is started?
Thanks
Andrew
-
Is anybody using this version? Anything special to do or just double click and away it goes?
Does the TVM802B need to be connected and on when the software is started?
Yes it needs to be connected but if it is not the s/w will prompt you to enter new IP add
ress of the machine.
I guess something is wrong with that rev. Ask for the latest non-beta revision.
-
Maybe you have an Antivirus problem blocking the execution?
-
...
(not my typo btw) which does not open in windows 7 pro or windows 10.
I just get the 'allow this app' question, then nothing happens...
Hi Andrew,
this problem is caused by some special chinese characters in the filenames which corrupt the windows explorer.
I solved this on a Linux PC, copying the complete DVD on harddisk and renaming the files with chinese characters. Then windows was able to read the files.
If you cannot do it, just write me a PM with your email address, I will help you.
Good Luck
Harry
-
Thanks Harry
No Linux PC here, so message sent
cheers
Andrew
edit (SOLVED) - Jim from QIHE sent me the previous version of the software today. It worked fine. So back to loading up reels ;D
-
good to hear that it works for you now.
Last week I modified all my PnP files to use quick-camera-mode even for R and Cs.
Some 0603 Rs and Cs where placed out of alignment and needed manual correction before soldering.
One reason are the plastic tapes. Components in plastic tapes are not well positioned, paper tapes are better but still not good enough.
The second problem is that the feeder coordinates change with temperature, realignments where required every few days.
This was solved nicely by using the camera's quick mode. Now all R and Cs are placed perfectly (with 100% speed).
Of course, placing a board needs much longer now, but speed was never an issue when using the TVM802.
my current setup:
Nozzle-1: size 503 (for 0603, 0805, 1206, SOT23 ...)
Nozzle-2: size 505 (for 1210 and most ICs)
pressure check is actived for all parts except 1210 (the 505 nozzle is a bit too big for 1210, but works without pressure check).
quick-cam-mode for all parts except ICs.
accurate mode for all ICs.
speed:
0603, 0805, SOT523 ... 100%
1206 ... 50%
1210 ... 50%
ICs ... 100%
SOT-23 and similar size cases ... 20% (they will flip and jump on the nozzle at higher speed, with one exception:
SOT23 with very flat surface like most MOSFETs then I run it at 100%).
Initial alignment before placing a batch of boards:
install a board with a little piece of doublesided tape at the most critical IC position, then I place just this IC.
I correct any misalignment by modifying the cam-1 coordinates and rotation.
Attention: some misplacements are caused by the cam-1's contrast setting which may also need an adjustment.
If all works well a batch of 50 boards is no problem without any further re-alignments.
Soldering with Loctite GC10 SAC305T4. This solder paste forgives a lot of bad alignments even with big ICs.
My stencil is 100u with 20% pad reduction.
Thats my story :-) I am doing about 50 to 100 boards every week.
How are the other TVM802 users doing ? Tell us.
Harry
-
I also came to conclusion that is better to use quick visual for all R & C. On my machine I noticed that a bit of error is introduced by nozzle rotation - it does not stay in the same spot as it rotates. Visual check fixes that.
I'm using 0.12mm stencil, same paste as you do, but we got quite a few bridges on the last batch, on 0.5mm pitch parts. Will try to make smaller openings in future.
-
the 802 is a nice machine for the money, but we must also talk about its limits.
This is a picture after placing (and before soldering) two CSD97376. These are 0,4mm pitch chips.
Both where placed one immediately after the other. The left one is placed perfectly,
the right one is left shifted by 0,2mm (which is half pitch) and therefore needs manual correction before soldering.
The reason is a bad edge detection by the camera. I tried all threshold setting but no luck.
During "accurate" adjustment I can see the red line jumping. Looks like the contrast with these type of cases is too bad.
I never had this problem with TQFP where the pins (with good contrast) define the edges.
-
Hi
Most of the bottom of that part is a ground pad. That's going to add a bit more complexity to the edge recognition. I'm not saying that's the whole issue. I'm only suggesting that if / when we start playing with LED angles and disks, it will be one more thing to worry about.
For those not already following this on the 920 thread:
The 920 has a different illumination setup. The LED's are angled a bit compared to the 802
There is some thought that a contrasting disk of one or another color behind the part (mounted to the head) is the solution.
FWIW: Back a "a while ago" my day job was picking visual images of things out of clutter with software. The target was different, the idea was the same. It's never easy to do 100% of the time running on autopilot.
Bob
-
I never had this problem with TQFP where the pins (with good contrast) define the edges.
You could try modify/add lighting so the reflection angles become different.
-
Folks!
I just got my TVM802A - got it running some sort of. I found one strange thing which causes troubles. From time to time the machine seems to skip a command. Sometimes a board runs through without any problems. Minutes later the prick or a nozzle stays down and the head moves. Of course this messes up the machine, rips off the belts or crashes the nozzle into one of the screws :-(
Does anybody have an idea what might cause the trouble?
I'm working on a W10 machine. Network is 192.168.0.x so no changes here, I'm connected to my company network, but even when I'm isolated (PC directly connected to TVM802) this happened. Cameras where recognized immediately, no need to install drivers.
Thanks for any ideas, maybe some brain storming might direct me in the right direction. So even if you have no definitive solution idea I need some new ideas to look for.
I had a strange thing in the beginning. The ground wire inside the machine was off. So I had about 85V from the frame against earth. I recognized it when I touched the PC case and the frame of the TVM802. I opened the side found the loose wire. I crimped a new plug and the voltage on the TVM802 body is gone. Maybe there is a problem with the power supply internally but now it looks OK.
Arnold
-
It soulds like the Windows 10 issues people have had where network commands does not get to the machine. Win7 works very well...
-
It soulds like the Windows 10 issues people have had where network commands does not get to the machine. Win7 works very well...
Thats also what I have in mind, Win10 could be the problem.
You can get a Win7 license on ebay for a few Euros, give it a try, may work much better.
Connect the PC to the 802 as short as possible, not through busy network routers.
When the Win7 setup and updating is complete, switch off the auto updater. Automatic updates can cause
immense network traffic in background.
-
Folks!
I just got my TVM802A - got it running some sort of. I found one strange thing which causes troubles. From time to time the machine seems to skip a command. Sometimes a board runs through without any problems. Minutes later the prick or a nozzle stays down and the head moves. Of course this messes up the machine, rips off the belts or crashes the nozzle into one of the screws :-(
Does anybody have an idea what might cause the trouble?
I'm working on a W10 machine. Network is 192.168.0.x so no changes here, I'm connected to my company network, but even when I'm isolated (PC directly connected to TVM802) this happened. Cameras where recognized immediately, no need to install drivers.
Thanks for any ideas, maybe some brain storming might direct me in the right direction. So even if you have no definitive solution idea I need some new ideas to look for.
I had a strange thing in the beginning. The ground wire inside the machine was off. So I had about 85V from the frame against earth. I recognized it when I touched the PC case and the frame of the TVM802. I opened the side found the loose wire. I crimped a new plug and the voltage on the TVM802 body is gone. Maybe there is a problem with the power supply internally but now it looks OK.
Arnold
I had same problem on W10. It actually broke me a couple of nozzles that way. Switched to running on XP machine.
-
Have any of the TVM802B users got any comments on the feeder system?
Are you breaking cover tapes? Does it relaibly feed.
And anyone doing 0402?
yes, it feeds reliably, no problem with cover tapes. But needs some adjustment from time to time.
If something goes wrong, the 802B stops and beeps until you fix the problem.
Most problems are self made. With every board I learn a bit and it works better and better.
0402, can work, but I would not do it as my standard parts, 0603 are fine.
For my 0402 parts I always use the camera, even for R and Cs.
Before placing a batch of boards I test the adjustment by placing the most difficult part to a board with double-sided adhesive tape,
a QFN or TQFP to check for any offsets.
Usually the camera rotation needs adjustment. Once adjusted it remains stable for the full batch.
On another day it may need readjustment.
The cam rotation is the only value I adjust regularly, all the other values look to be stable over weeks.
I put a videao on youtube this week which shows some close-up views of the pick&place head beginning at minute 2:30
https://www.youtube.com/watch?v=VsGQ3Gwg9xQ (https://www.youtube.com/watch?v=VsGQ3Gwg9xQ)
Do you have a method to calibrate the machine? I just got a TVM802A and can not find how to do this!
RichardS
-
Hiya all
So sorry that this is my first post here if I should have first said Hi on a new introductory member thread.
Been reading this thread quite a lot and a few times, hence the SPY part of my chosen user name :)
I would really first like to start by thanking ServoKit and all other contributors on this thread that freely share much valueable and appreciated information
Reading your posts that share information, modifications, knowledge and experience, good or bad, directly resulted in my recently aquiring a TVM802B.
I hope this will allow me to participate here and allow me to now reciprocally contribute.
I do appreciate that there are limits to this equipment, but then at its price point think it is a good entry option and provides a very interesting degree of customisation possibilties, expadability and flexibility that I am able and willing to perform, and I hope will make interesting reading.
I am very comfortable in both software and hardware areas, hardware being electronic and engineering wise.
I have already started on making a quick swappable racking system for the IC stack area
When purchasing, I also aquired a rather interesting Vibratory tube feeder option that I have yet to try out, and will of course let you know more about this with pictures :-+
I am running Sucface V2.23, V2.27 and V.28 on Win7 and annoyingly also see a nozzle left down.
I have 12 Juki Nozzles ranging between the 503 and 506, but just purchased 20 more,
above and below this size range, at prices of mostly $14 each.
I believe and hope I am already about half way through the steep learning curve required.
My only outstanding issue is that I could do with some help with is that I still need to figure how to change the coordinates for componant discarding location.
-
Hiya all
So sorry that this is my first post here if I should have first said Hi on a new introductory member thread.
Been reading this thread quite a lot and a few times, hence the SPY part of my chosen user name :)
I would really first like to start by thanking ServoKit and all other contributors on this thread that freely share much valueable and appreciated information
Reading your posts that share information, modifications, knowledge and experience, good or bad, directly resulted in my recently aquiring a TVM802B.
I hope this will allow me to participate here and allow me to now reciprocally contribute.
I do appreciate that there are limits to this equipment, but then at its price point think it is a good entry option and provides a very interesting degree of customisation possibilties, expadability and flexibility that I am able and willing to perform, and I hope will make interesting reading.
I am very comfortable in both software and hardware areas, hardware being electronic and engineering wise.
I have already started on making a quick swappable racking system for the IC stack area
When purchasing, I also aquired a rather interesting Vibratory tube feeder option that I have yet to try out, and will of course let you know more about this with pictures :-+
I am running Sucface V2.23, V2.27 and V.28 on Win7 and annoyingly also see a nozzle left down.
I have 12 Juki Nozzles ranging between the 503 and 506, but just purchased 20 more,
above and below this size range, at prices of mostly $14 each.
I believe and hope I am already about half way through the steep learning curve required.
My only outstanding issue is that I could do with some help with is that I still need to figure how to change the coordinates for componant discarding location.
Hey Spy Guy, I am new here also, so we probably have a lot in common :-) except yours is B and mine is A (with the juki nozzles)
I had not heard until you saying that it left nozzle down win WIN7... crap!
I am also looking for info on completely tuning this things zero's....
RichardS
-
Hello and congrats for your new machines.
Tuning, thats the hard part, it took me a full week after getting my machine. You will learn a lot during setup.
I did it in April, here is what I remember:
The 802 is totally unconfigured when shipped (at least my machine was unconfigured, the values already entered in the setup unusable, maybe this has changed). You need a full setup before the first run. A starting point are the pictures delivery with the software, these values are not good, but better than nothing. So enter all these values from the jpg files into the system setup.
Next thing is to fine tune the pick-up coordinates of the left stack (and the back for the 802B). Start with the left front reel first. Install a reel and move the nozzle to the position previously taken from the jpg picture. Then correct the Y pos until the nozzle is centered. Now you need to check and correct the prick offset, the prick must exactly hit the hole of the tape. After running a simple PnP job for this reel, the prick will move the tape to the X position. Now the nozzle must be adjusted manually for the X pos over the component. The X pos should be the same for all left stack reel, the Y pos must be adjusted for every reel (back stack upside-down)
For the place postion the camera-1 offset must be adjusted. I do this work before every batch of boards because this setting is not stable over time or temperature. I use a chip with as many pins as possible and a real board, put some doublesided adhesive tape to the place location and then start a test run.
If the chip is not placed correctly I adjust cam-1's x and y offset and the rotation until it is placed perfectly.
There are other settings, so take your time.
BTW I never had a problem with nozzle down. If the nozzle is accidentialy down (i.e. by manual move) the machine beeps and print a warning and stops.
Harry
-
Hello and congrats for your new machines.
Tuning, thats the hard part, it took me a full week after getting my machine. You will learn a lot during setup.
I did it in April, here is what I remember:
The left prick offset is measured from what 0,0? offset from what?
I understand howto zero the nozzle per tape feeder.
Also the nozzle 1 and nozzle 2 vision offset I you can test #1 but what about #2??
Thanks!
RichardS
-
Many Many thanks Harry ;-)
Seeing your super cool custom milled IC stack tray and videos was certainly hugely inspirational and what mostly convinced me to go for the TVM802B.
I have CNC routing equipment but only go through thin <2.0 mm ally sheets so don't really do depths (Z) much
My TVM802B came pre loaded with the settings as per CD Jpg's.
I am guessing that these settings are actually stored in the machine itself via TCP/IP based read and write commands.
So I had a little head start, but still have had to do some tuning, figuring and learning which does take at least a week.
I quickly figured about the Left and Rear tape stack pick point & prick point aligning requirement.
I have however only just started playing with all the camera positioning and relative offsets to nozzles etc.p
I note and appreciate your experienced comments re time and temp causing these to shift and already have the transparent sticky tape aspect up my sleeve ;-)
I don't actually need to put ths machine to work until late Nov, and in the meantime will ensure that by then I will be as knowledgable and conversant as Europeanly possible with it ;-)
If the nozzle is accidentialy down (i.e. by manual move) the machine beeps and print a warning and stops.
Not seen mine do this yet, which is something I would have coded in, and expected, just been stepping through two TFQP placements from a standard waffle tray. After picking one up using nozzle 2 it dropped nozzle one, next step saw it dragging the tray, and me swearing a l
As I stated, in Win 7.
I think that you just have to go into owning and using one of these with an open mind and apprciation of what it is and is not for it's cost and be willing to fill in the gaps yourself and be forgiving and understanding.
-
They are great little machines. I have made just over 1000 units with mine and I am looking to make a small tweak to it.
I use the smallest nozzle a lot and find that the pressure sensor does not detect when the tiny nozzles missed a pick up. The bigger nozzles work very well.
-
hiya protoneer
I use the smallest nozzle a lot
Just to help me and others clarify and fully undestand what you mean, mine came with 4 different Nozzle sizes labeled No1 to No4.
These equate to juki numbers of 503,4,5 and 6.
I quickly learned that Juki has a smaller 500/1/2 as well as a larger 507,8 and a 510 and 511.
So which size on the Juki numbers are you referring to as in "using the smallest"?
-
hiya protoneer
I use the smallest nozzle a lot
Just to help me and others clarify and fully undestand what you mean, mine came with 4 different Nozzle sizes labeled No1 to No4.
These equate to juki numbers of 503,4,5 and 6.
I quickly learned that Juki has a smaller 500/1/2 as well as a larger 507,8 and a 510 and 511.
So which size on the Juki numbers are you referring to as in "using the smallest"?
Yes... That is where the plot thickens. I have a TVM802A that uses non-Juki nozzles.
The smallest I have is able to pickup 0402-0805 where the next up is struggling to pickup 0805 consistently.
Does anyone know where I can buy some spare nozzles for an TVM802A? I asked for a quote form the machine supplier and they quoted $100 per nozzle and $80 shipping. That sounds a bit steep to me but I only have one #1 nozzle left after my Win10 episode...
The nozzles were on Aliexpress a while back but I have not seen them in a while. It also makes me wonder if I can upgrade the machine to Juki nozzles.
-
Yes... That is where the plot thickens. I have a TVM802A that uses non-Juki nozzles.
The smallest I have is able to pickup 0402-0805 where the next up is struggling to pickup 0805 consistently.
Does anyone know where I can buy some spare nozzles for an TVM802A? I asked for a quote form the machine supplier and they quoted $100 per nozzle and $80 shipping. That sounds a bit steep to me but I only have one #1 nozzle left after my Win10 episode...
T.he nozzles were on Aliexpress a while back but I have not seen them in a while. It also makes me wonder if I can upgrade the machine to Juki nozzles
So I have contacts to get the nozzles for around $20 a pop, if you want let me know, where are you in the world I was just going to order some...
My machine 802A just came and had juki installed :-)
R.
-
@SPYGUY
Any hints on prick calibration? Still can not figure that one out well enought.
RichardS
-
today I did some work on the vacuum of the TVM802.
The built-in pump can do 150mbar (2.2psi) which is a very good value for a small pump.
But it takes a lot of time until it reaches this low pressure.
The normal working pressure is about 500mbar (7.3psi).
Then I replaced it with an external pump combined with a pressure tank (2 liters).
This external pump can go down to 10 mbar (0.15psi) in a few seconds. Now sucktion is very well.
However there is another problem now: For 0603 parts I use the Yuki 503. With the lower pressure the pressure sensor does not work any more. It works only between 400mbar and 650mbar (5.8 - 9.4 psi)
With the 504 nozzle it works without problem. Tomorrow I will try to pick 0603 parts with the yuki 504, lets see if it works or not.
Unfortunately there is no adjustment for the pressure detector.
-
So I have contacts to get the nozzles for around $20 a pop, if you want let me know, where are you in the world I was just going to order some...
My machine 802A just came and had juki installed :-)
R.
Thanks. I will have to look into upgrading my machine to take Juki nozzles first. I asked the machine supplier about upgrade kits but they said they will have to do it. (NOT!!!!)
I get the feeling its as simple as fitting a new nozzle holder to the small stepper motors and making sure that the nozzles sits at the same height as the old one. (The new TVM920 software seems to be configurable for things like this.)
Can you tell me what version of the software you are running?
I calibrated the prick with the tray settings. The Prick needs to align with every trays advancement holes. The Prick is tapered and will slip into the hole when properly aligned. (For each tray I moved the head manually and dropped the prick making sure it was aligned.)
After that the Nozzle to Prick offset needs to be set to ensure the nozzle picks up the part in the center. This bit is still not 100% on my machine cause when the part rotates 360 I can see that its not 100% centered.
-
motivated by ar__systems I did a first test run with an external vac pump today.
What a difference !!!
Instead of the (correct) 503 nozzle I use the 504 for the 0603 parts. The 504 is usually too big, but with the better vacuum is works excellent.
Here are some pics of my setup.
Placement is significantly better. Now the last remaining problem is the inaccurate vision of the 802.
BTW, I opened the placement head: the manufacturing quality is very good. I did not expect such a good quality from a Chinese company. Compared with the pictures of the other desktop PnP machines the 802 has the best manufacturing quality.
-
Hi
I suspect that a two stage oil pump is a bit overkill :)
Bob
-
Recently I started to make paneling outside of the TVM's software. So if I have a panel of 10 boards, 10 parts each, I don't give the machine a 10 parts data file and configure it for 10 boards. Instead I it one data file of 100 parts and configure it as single PCB. Saves time on fiducials and such, when you have to restart the job.
-
@tmf
I have not actually done a prick calibration myself yet, but it seems so logical and straight forward enough to me that I feel i understand it enough to offer some comment
I believe it is all about the possibility of a small difference between the position of the prick holes relative to the actual part locations on any particular tape that may vary between reels / suppliers etc.
As a generilized example only
Take an 8mm wide tape, the prick hole might be anywhere between say 6.5mm to 7.5 mm across its width, so you need to adjust the pricks position across the tape width, so it drops in the hole at least correctly in this particular direction.
The other direction is not so important as once it's correct across the width, it will drop into a hole and pull the tape forward repeatably by the assigned pitch, to the point where it then pulls out.
So then you just have to manually set the nozzles pick up point of the part itself centrally, relative to where the prick mechanism has moved the tape to and exited the hole.
@harry4516
wow that's quite some upgrade to the Vacumn system. Thanks for sharing ;-)
Like Bob though, I do wonder if it is not a bit overkill.
Would be nice to know the parts, sources, Part numbers and costs though, because sometimes overkill is nice ;-)
Although I have not checked, I imagine that as the pumps run full time, the pipes up to the solenoid valves in the heads hold vacumn, that is then quickly depleted when the solenoid valves are energized.
I wonder therefore if just tapping in a small accumulator / reservoir might yield a cheaper and less complex improvement.
Re the pressure sensor problem I read so much of.
Although the software accepts only a digial yes / no switch type input, from the digital sensor
I imagine it to be realatively easy to replace the current digital output sensor with an analogue one that has it's ouput fed into a comparator as well as that of a user adjustable potentiometer.
@ar_systems
I actually do both using fram and per PCB fiducials, but prefer outside paneling.
I have panels where only one board failed an electrical check and so can delete this.
On another note re moving the discard coordinates
I just got yet another variation of the V28 software that has new fields in the config sys menu for this. Of course there are also new bugs with this version.
Whats New LOL
-
Hi
I suspect that a two stage oil pump is a bit overkill :)
Bob
sure, this is overkill,
previously I had a membrane pump of the same size. It had two problems. The engine got very hot after a minute and it only started at normal pressure, not if vac was already there.
The new pump costs about $250 and stays cold even after 10 minutes.
-
I use a very overkill pump on my machine with a storage tank and controller. The nice thing is that is cycles every 5-10 minutes and only stays on for 30sec to a minute. It is a rotary type and very quiet even when it is running, but it is much better than a pump that has to run continuously in terms of noise. Originally I thought that it would be a problem since the vacuum is not constant, but so far it is not a problem at all.
-
@rx8pilot , @harry4516 , @ar_systems
Would this vacuum pump do the job? They go for US$100 shipping included..
https://www.aliexpress.com/item/FY-1H-N-150W-1L-Lightweight-portable-air-compressor-vacuum-air-pump-for-vacuum-LCD-separator/32736004424.html?spm=2114.13010208.99999999.304.2cWg7z (https://www.aliexpress.com/item/FY-1H-N-150W-1L-Lightweight-portable-air-compressor-vacuum-air-pump-for-vacuum-LCD-separator/32736004424.html?spm=2114.13010208.99999999.304.2cWg7z)
Also do you need some kind of regulator to limit the vacuum or will the air valves(solenoids) be okay without it?
-
I just have a cutoff switch that I can set to turn off at just under 29in and turn on when it gets to around 22in. It also has a continuous switch but I have never needed it. I use the same vac system for my various pickup tools and to de-gas silicone and polyurethane for over molding cables.
-
@rx8pilot , @harry4516 , @ar_systems
Would this vacuum pump do the job? They go for US$100 shipping included..
https://www.aliexpress.com/item/FY-1H-N-150W-1L-Lightweight-portable-air-compressor-vacuum-air-pump-for-vacuum-LCD-separator/32736004424.html?spm=2114.13010208.99999999.304.2cWg7z (https://www.aliexpress.com/item/FY-1H-N-150W-1L-Lightweight-portable-air-compressor-vacuum-air-pump-for-vacuum-LCD-separator/32736004424.html?spm=2114.13010208.99999999.304.2cWg7z)
Also do you need some kind of regulator to limit the vacuum or will the air valves(solenoids) be okay without it?
this is the same pump as mine. It works ok.
It blows out a fog of oil, you need to separate it from the PnP machine.
I have connected a tube which goes out of the window.
Maybe an oil free pump is a better choice.
@rx8pilot: what size is your storage tank ?
-
I think it is about 10 gallons.
-
Hi
I suspect that a two stage oil pump is a bit overkill :)
Bob
sure, this is overkill,
previously I had a membrane pump of the same size. It had two problems. The engine got very hot after a minute and it only started at normal pressure, not if vac was already there.
The new pump costs about $250 and stays cold even after 10 minutes.
Hi
In this application the difference in pump down between a single stage and a double stage pump is likely to be unimportant. They both will get past 29" with no trouble at all. The single stage (possibly of questionable quality) can be had for < $80 in the same ~ 3 to 4 CFM pumping capacity.
Bob
-
It blows out a fog of oil, you need to separate it from the PnP machine.
I have connected a tube which goes out of the window.
Maybe an oil free pump is a better choice.
Could it be that you over filled the oil?
I am tinkering with the idea of adding a pressure sensor. Something like this: http://www.mouser.com/ds/2/302/MPXV7002-783439.pdf (http://www.mouser.com/ds/2/302/MPXV7002-783439.pdf)
Just to get a sense of that the machine is doing.
The more I think about it, if I can get the machine to report an error with the smaller components it would save me plenty of time baby sitting the machine. (It works great on bigger components to the point that I can work on other stuff while it does its job and when there is an issue it just stops)
-
Does anyone know how many pressure sensors the machine have and where they are located?
(I have about 12 hours of work booked for the machine before I can poke around under the hood)
-
It blows out a fog of oil, you need to separate it from the PnP machine.
I have connected a tube which goes out of the window.
Maybe an oil free pump is a better choice.
Could it be that you over filled the oil?
I am tinkering with the idea of adding a pressure sensor. Something like this: http://www.mouser.com/ds/2/302/MPXV7002-783439.pdf (http://www.mouser.com/ds/2/302/MPXV7002-783439.pdf)
Just to get a sense of that the machine is doing.
The more I think about it, if I can get the machine to report an error with the smaller components it would save me plenty of time baby sitting the machine. (It works great on bigger components to the point that I can work on other stuff while it does its job and when there is an issue it just stops)
Hi
*IF* you want the machine to report a very small part missing (with a pressure sensor) ... you need a low volume vacuum pump.
*IF* you want the machine to pick up and hold a very small part ... you need a high volume vacuum pump.
Essentially you have two requirements pointing you in opposite directions. As the nozzle gets smaller and smaller, the flow through that nozzle (when open) gets less and less. With a high enough volume pump, the entire pressure drop would be at the nozzle (for a really small nozzle).
An alternative is to detect flow rather than pressure. You can also check for the part with a camera.
Bob
-
...
Essentially you have two requirements pointing you in opposite directions. As the nozzle gets smaller and smaller, the flow through that nozzle (when open) gets less and less. With a high enough volume pump, the entire pressure drop would be at the nozzle (for a really small nozzle).
...
exactly, thats the problem.
The yuki-503 nozzle (which is the right nozzle for 0603 parts) works with pressure sensor and the original vac pump very well. It detects if a part is missing.
With my new strong vac pump pressure detection does not work any more, so I use the yuki 504 with a larger hole. I placed a few boards and it works well with 0603 parts. This is a big advantage because the TVM802 has two nozzles only.
Previously I used the 503 and the 505, now the 504 and the 505. The better vac let me use a broader range of different components.
Also, I had big problems with SOT23 before (jumping and rotating on the nozzle while travelling), this problem is solved with the better vac.
Harry
-
The yuki-503 nozzle (which is the right nozzle for 0603 parts) works with pressure sensor and the original vac pump very well. It detects if a part is missing.
With my new strong vac pump pressure detection does not work any more, so I use the yuki 504 with a larger hole. I placed a few boards and it works well with 0603 parts. This is a big advantage because the TVM802 has two nozzles only.
Previously I used the 503 and the 505, now the 504 and the 505. The better vac let me use a broader range of different components.
Also, I had big problems with SOT23 before (jumping and rotating on the nozzle while travelling), this problem is solved with the better vac.
Harry
Thanks for pointing out the high vs low flow.
I have ordered the stronger Vacuum pump so that will solve one issue. My TVM802A does not have Juki Nozzles but I have seen a few adapters that might fit. Will give that a go.
Out of interest @harry4516 , what size Juki Nozzle would you use for 5050 led's?
-
Does anyone know how many pressure sensors the machine have and where they are located?
(I have about 12 hours of work booked for the machine before I can poke around under the hood)
It has two pressure sensor on the bottom side of the circuit board that is screwed to the top of the block.
Both are semiconductor transducers, however schematics is extremely basic. There is no calibration of the sensors, and there is no threshold adjustment.
On my machine I don't have a problem with it not seeing 0603 missing. I think on my machine it is too sensitive even with external pump, and dumps parts that look like picked up relatively ok.
-
Hi
*IF* you want the machine to report a very small part missing (with a pressure sensor) ... you need a low volume vacuum pump.
[/quote]
That's not true. You just need a pressure sensor with adjustable threshold.
-
Hi
*IF* you want the machine to report a very small part missing (with a pressure sensor) ... you need a low volume vacuum pump.
That's not true. You just need a pressure sensor with adjustable threshold.
[/quote]
Hi
As long as you have a pressure sensor that has a perfectly stable threshold, an very small hysteresis, and a set point that will adjust to very close to 29.x" of vacuum .... yes that is correct. With practical sensors, you do indeed have a compromise for very small nozzles.
Bob
-
Hello, what do you guys recommend to make sure the nozzle picked a part at our DIY level ? Vision, a particular pressure sensor, lasers ? Thank you !
-
Vision is the best value for money, as it gives you both alignment and mis-pick detection. Any pick & place without vision is just a toy.
-
IMHO it seems that the same thing is required.
Both based on taking an analog output from a sensor such as listed and converted into a digital output based on comparison input from a user adjustabe threshold value.
1) The softwares digital pressure sensors detect input.
2) The reservoir pressure that can control the internal or external Vacumn pump(s).
This can easily be achieved via a very basic micro controller with A to D's and coffee break coding, or even with just basic comparitors.
Not so much an analog person myself.
You could even get more fuity with a micro that has TCP/IP input that then applies different vacumn values for 1) based on watching co ordinates.
just my humble 2C's worth
-
Folks!
I reported severe problems with the TVM802A powered by W10 about 2 weeks ago. Based on recommendations in this forum I installed a new PC with plain W7 patched it up to the most recent level and started testing. The PC is a 2GB ATOM box which just went out of use. The program does not need much PC power. To make a long story (calibration) short: it works with almost no problems. I have some error messages when I accidentially klick outside the controller program. This does not cause any wrong movements which I had before. I ordered new nozzles as the current ones might got damaged - or I'm not quite sure about them. Sometimes they drop components even speed 50% speed. I do not want to use vision on 1208 resistors.
It's really worth the time to invest time on calibration. Unfortunately I'm still trying to find out what some of the values in the software mean. We should compile some advices on this to help others.
Next upgrade is a touch screen. I used a mouse and later a touchpad keyboard to save space. But when I'm at the machine it might be easier to touch the screen. Especially some confirmation windows are quite annoying. I still have the problem, to stop the machine with the keys on it. So I frequently change back to the computer -> my idea to connect a touch screen.
Arnold
-
...Sometimes they drop components even speed 50% speed. I do not want to use vision on 1208 resistors...
I had the same problem, it was caused by the poor vacuum (in combination with the level of vibrations). I solved that with an external vac pump.
regarding 1206... after some experience you will use vision for all components. Take more time but need less manual corrections.
-
After I align with 2 fiducials I ask the machine to go to say (0,0) I then make note of it, I then move 6.000 inch to the right and it moves 6.014.... I have verified this time and time again, then a 6.000" move in Y and its off approx 0.014" again...... + 0.3 to 0.4mm both axis
I have also tried to not use the fiducials, do not enable them.... and still the same thing....
The 2 point I am moving between on the PCB are exactly correct!
I think the pixels per inch calibration must be off, how do you calibrate this??
Or???
Thanks.
Richard.
-
TFM,
I'd contact the guys at QiHe.
A mate of mine who owns a TVM920 had the same issue.
The gave him an updated version of the App. which rectified it issue immediately.
Good luck - Peter
PS: perhaps one of the other guys may have a later/est version they could send you.
Don't know about the 802, but the 920 has provision to calibrate this and reconfigure.
-
my software for the TVM802 is: Version 2.12
-
Hello!
My SW Version of TVM802 is 2.23Beta2. If somebody wants to have it I can put it on a public space.
Arnold
Sorry folks there was a typo in the version I modified the posting
-
That would be nice.... mine is 2.23beta2 and they sent me the "solution" but when you goto the screen that has the config its missing! So they are probably talking about a newer version.... (you did mean 2.32beta2 or did you mix up 2 characters)
Please let me know when you upload.. looking forward to it!
Thanks!
-
china just sent me 2.27 and it wont even home! |O
RichardS
-
china just sent me 2.27 and it wont even home! |O
No good news. But how have you got access to the update?
-AH-
-
instead of mailing bad updates QiHE should release the source code.
The community could fix the problems in a very short time.
I think thommo from Australia is in close contact with these people. Could you talk about it ?
-
It is useless. The s/w is crap and should be written from scratch.
-
Guys,
I look at it this way.
You spend peanuts and you'll certainly get monkeys - there's no two ways about it. These are cheap-arse machines, and when you see the level of work that's gone into the 920 [can't speak for other models], it's hard to imagine how these companies even survive.
I did my research and the 802 doesn't suit my business case. I've invested a fair amount of time [with a translator] speaking with these guys, and I believe they really want to help, and when they've understood the issues I've presented to them, so far they've responded brilliantly - no complaints. I believe the 920 has a lot of potential and is the only one of the bunch I'd consider investing in for this use-space. The main reason being that they have opted to design around industry standard 3rd party Yamaha component feeders.
But another thing that has also struck me now that I read this complaints - and that is the dependency on 3rd party PC and OS [which the 802 models appear to have]. That's very hard work for anyone to maintain - just ask those that have tried.
Again, the TVM920 is different, and although it doesn't have a high-end commercial MOBO, it does have it's own MiniITX hidden away inside the machine, so what a user experiences is 'more likely' what was experienced in the factory by QiHe R&D guys [so that's another variable out the window].
It sounds to me that many of you are not really concentrating on the 'direct to QiHe' path, instead relying on others [from this forum and elsewhere] to solve the various issues. So I'd really encourage you to be patient and contact them 'directly' - otherwise they cannot learn or experience the issues that are being discovered.
I am happy to share my experiences in regard to the 920 as they occur, but I do not have the time or interest to extend that further - say into the 802 space. I hope you can understand the practicality of what I am saying here. It's not that I don't want to help, rather I need to 'stick to my knitting'. I have provided my links and contacts at QiHe in posts on the TVM920 forum for those wishing to delve.
The answer may just be that, for many of you, the 802 provided sufficient 'visibility' into the potential of handling PnP requirements 'in-house', and now it's time to move up to the 920 - something I'll know more about personally, and soon I hope!
Good luck guys - Peter
-
So I am live online with them now trying to figure it out, china never likes letting secrets out :-)
RichardS
-
My TVM802B came with Sucface V2.23Beta2 on the disc, which was odd, as the pics in the basic printed user guide that came with it showed V2.25.
That and the name typo set the expected standard, which has been maintained :(
I had missing line of manual move mode buttons that included the Pump button, which was actually there just but not shown due to a lack of my screens resolution.
I went through a V2.27 which addded X and Y scaling that was initially set to -0001 causing the No home tmf is probably suffering with, just a quick jerk in the wrong direction, easily fixed by setting both to 1.
I have since had two V2.28's, the latest to add fields to change the discard co ordinates, which does that but has added other errors.
I will drop them on one of our servers if anyone needs this.
re thommos (peters) Peanuts and monkeys post
I fully accept and appreciate the low cost, built to a budget, limited capability and entry level aspect of ths equipment.
But at it's low price, it is good starting point, that attracts DIY buyers, most able and willing to contribute to a forum and thread such as this, which is a huge benefit all round, and I certainly did not purchase this without any expectation of not having to do some work to customise it and make suitible upgrades, based on this thread and it's many inspirational contributors alone, to who I give huge thanks :-+
I just tried out my tube feeder option this weekend and sure enough it does not work right out of the box. So whats new |O
With a seemingly endless chuckle, I discovered that the parts are too low to be picked up by the nozzles at even their lowest travel point.
Just another interesting challenge, and an opportunity to get inventive and creative.
And also another opportunity to reciprocally post and inpire and help other owners out.
-
So far 2.27 is working well for me in trials right now..... the scaling helped tonnes on the 300x300mm pcb..... just placed a SSOP24 part in a far coordinate and its almost dead on.....
RichardS
-
I have sent an email to QiHE,
they resonded immediately and sent me 2.29
I will test it tomorrow and keep you informed.
Harry
-
Folks!
I have got access to TVM 802A and B SW version 2.27. The SW seems to be the same, the only difference is that the B offers more component sources which the A does not have physically. I did about 50 boards with the "B" SW on my TVM802A. It used the upward camera correctly, as you might know the position of the camera is different on both models. So the SW recognices which HW is attached.
This morning I got 2.29 bits per mail directly from Gera. My mail scanners tagged it as "dangerous" but all virus scanners did not complain so I stared it. I just started the SW and played around a few minutes. The x/x Axis Scale was set to 0,001 so the machine did not even home. Already in the 2.27 SW that option was available and I have configured it before. We had that reported on a previous posting already. As I was aware that this was added and I had to readjust it already on the 2.27 SW. I immediately checked the configuration screen (last tab on the SysConfig). I have set both values to 1 and will check calibration. I expect if I have to adjust those values to anything different of 1 I need to recalibrate all other machine values. For now I positioned the head to the camera and strip positions which I have stored on my PC from previous SW revisions. With 1/1 values on the scaling I find the same positions again as before. But I need to check if the stepping is precise on large x/y values.
I have attached a screen shot of the new configuration option.
Finally I asked Gera if I'm allowed to share the bits publicly, up to now I have no response. Might be a time zone issue, I'm sitting in Europe, so it might be too late in China. I plan to offer it on one of my servers of if we get up a WIKI or something I will post it there.
rgds
Arnold
-
Technically the s/w does not need XY scaling option to provide perfect placement. It only complicates setup and adjustment process. All of the positions of the machine (like feeder's positions) are set in absolute coordinates even if they are slightly of. So that's fine. Once the PCB is configured, the exact scaling within the space of the PCB is established by means of fiudicials.
-
Technically the s/w does not need XY scaling option to provide perfect placement. It only complicates setup and adjustment process. All of the positions of the machine (like feeder's positions) are set in absolute coordinates even if they are slightly of. So that's fine. Once the PCB is configured, the exact scaling within the space of the PCB is established by means of fiudicials.
Hello!
You are completely right BUT:
on my machine when I run the head to 0/0 and next to 300/0 I find the head misaligned, less than a millimeter. So If I had a small component far at the right It may be an advantage to change the stuff. I'm only doing short boards under 10cm so I decided to ignore that little error. So the possibility to calibrate the movement itself might help in special situations.
-AH-
-
The math in the software is a bit screwed up I admit.... For example, on a long board (20cm in X direction) I've found that without vision the further right I go, the more off the components get in X axis. So I modified my export script to fix the "scaling error" and got a perfect placement. WITHOUT VISION. Now, as soon I enabled vision for these components, I got X errors again. I can't understand it... |O
All I'm saying they introduced XY scaling to let people workaround some other problems in the sw instead of fixing them :)
-
Hi ar__systems!
I agree with you, I wonder why the errors happen. Driving steppers correctly does not allow to have those misalignments. Nevertheless it's still a fine machine and saves me a lot of time!
-AH-
-
All I'm saying they introduced XY scaling to let people workaround some other problems in the sw instead of fixing them :)
Bingo my thoughts exactly, however since I used this "feature" all my PCB's are on track now ....
However read my next post :-)
RichardS
-
OK so I just ran my first real pcb, its 300x300mm (12x12") with 144 leds, of which 140 were perfectly placed, the other 3 were 25mm or 1: off at least and one was rotated.
I used vision... and when these error occured the image on the screen left was a part that was not centered..... so I think there is some form of error from the camera, not sure why yet.... it was board #1, but I will watch it much more closely next time and have my finger on the pause button!
RichardS
-
Update from the TVM802B machine in Zurich:
Been using it extensively. 0402, QFN, etc components, you name it. Placed a total of around 40-60 PCBs by now, component count 200-300 per pcb (we use the machine to make test adapters for QCing and calibrating the Fotokite Phi. This lets us have a 3-day revision time on these adapters including 2-day PCB external manufacturing)
The machine has been working great except recently it has developed 2 ailments:
1) one of the nozzles places randomly doesn't enable the vacuum anymore. It somehow fixed itself. Not sure yet if it's mechanical, software, or electrical problem
2) The machine has a certain corner where it makes a horrible grinding "step" sound and I've noticed completely messes up pick-up. It seems like a mechanical problem but I haven't found anything loose or any obvious signs of wear. The nozzle head seems to become "non-rigid", and dips down, causing the feeder pin to get stuck consistently in the leftmost back pickup position. It can be "manually" made to work by applying corrective force to the head, but something is definitely not right.
Can anyone comment on experiencing something similar? A bit puzzled here since again, no obvious wear&tear marks or loose fasteners -- though we haven't yet fully disassembled the machine to investigate.
Another fun note: I noticed that if you get PCB's with HASL/immersion tin surface finish, we've had some so reflective as to confuse the fiducial camera (it reflects itself and the fiducial no longer stands out). Turns out there is a simple fix -- apply a bit of clear tape over the fiduacial to make it less reflective!
The software is shit but workable once you know the bugs. Haven't had a chance to use the more recent versions.
-
sergeil:
Sure sounds like you have some physical issues with your particular TVM machine.
It seems they all have some individual personality :-DD
Thre is a contrast adjustment on the down facing camera that may solve your HASL problem.
tmf:
so I think there is some form of error from the camera, not sure why yet
Harry:
During "accurate" adjustment I can see the red line jumping. Looks like the contrast with these type of cases is too bad.
I never had this problem with TQFP where the pins (with good contrast) define the edges.
Having quickly gone through much of the step learning curve required of any owner, I am now also at this camera error / red line jumping point too, in my case on a small BGA device. It seems like the Illumination amount, or angle, Camera or software lacks sufficient contrast provision or detection between the pads and case edges.
I live in hope that I can simply and physically improve this to the required degree somehow ;-)
-
......
2) The machine has a certain corner where it makes a horrible grinding "step" sound and I've noticed completely messes up pick-up. It seems like a mechanical problem but I haven't found anything loose or any obvious signs of wear. The nozzle head seems to become "non-rigid", and dips down, causing the feeder pin to get stuck consistently in the leftmost back pickup position. It can be "manually" made to work by applying corrective force to the head, but something is definitely not right.
Can anyone comment on experiencing something similar? A bit puzzled here since again, no obvious wear&tear marks or loose fasteners -- though we haven't yet fully disassembled the machine to investigate.
........
Hi
More from the global belt driven 3D world rather than specific to the 802, general troubleshooting:
1) Binding could be the issue. Everything needs to slide easily on the rails with the motors powered off. Misalignment of the rail sets (one rail is out) can cause things to bind at one end or the other.
2) Dragging of cables can be the problem. It's the same basic thing, only at the extreme of the cable distance. One link of the drag chain gets messed up and the cable no longer extends nicely.
3) Never rule out good old dirt. A bit of lube (not much) on a rag will sometimes pull a lot of gunk off of an apparently clean rail.
4) Loose end stops switches will allow the travel to go to far. Rather than stoping before it gets to the end of the rail, it crashes and makes noise.
That all assumes the machine is working correctly otherwise. If there are multiple issues, that adds a few items to the short list of things to check ....
Have Fun
Bob
-
after 6 months and hundreds of boards with the TVM802B I came to a conclusion:
the 802 is as it is, it helps me a lot and saves a lot of time.
It has some issues, like the inefficient camera, but I can live with it.
Manual correcting the placement of a few parts before soldering is not a problem.
I will never use it for BGAs (by the way, I would not trust my soldering oven or my soldering knowledge in case of BGAs).
I will continue working with the 802 and in parallel I will watch the progress
of the new smallsmt machine. If this machine is available I will possibly make a new decision.
Until then I am happy with the TVM802.
Harry
-
You spend peanuts and you'll certainly get monkeys - there's no two ways about it. These are cheap-arse machines, and when you see the level of work that's gone into the 920 [can't speak for other models], it's hard to imagine how these companies even survive.
Hope you are right with that 920. Its still ( reletively ) a cheap machine..
-
Yeah - it's definitely still in the 'cheap' machine space Mr Packethead, maybe not at the bottom of the list, but still 'relatively inexpensive' for sure.
So, following my own statement, I don't expect to have it deliver full SMT Production Line capacity or longevity. I understand what I'm buying, and that's perfectly OK with me.
I believe that they've sold a reasonable qty of them now, so looks like they are 'taking off' which will be a nice position and comfort to me too.
You spend peanuts and you'll certainly get monkeys - there's no two ways about it. These are cheap-arse machines, and when you see the level of work that's gone into the 920 [can't speak for other models], it's hard to imagine how these companies even survive.
Hope you are right with that 920. Its still ( reletively ) a cheap machine..
-
One word "BOUNTY"....
There is very nice software being made from the Open PNP guys but no support yet for 820A/B
I have a TVM802A right now and the software and vision is garbage, do you think a bunch of us with 802A/B could put some money into a "bounty" so we might find someone to complete this code for us?? We could use https://www.bountysource.com/ (https://www.bountysource.com/)
Richard.
-
One word "BOUNTY"....
There is very nice software being made from the Open PNP guys but no support yet for 820A/B
I have a TVM802A right now and the software and vision is garbage, do you think a bunch of us with 802A/B could put some money into a "bounty" so we might find someone to complete this code for us?? We could use https://www.bountysource.com/ (https://www.bountysource.com/)
Richard.
Jason from OpenPnP here. To be honest, I don't know if you'll be happy with OpenPnP either. It lacks some of the things that I see folks on here talking about being important for production use. What I would recommend is downloading it and taking it for a test drive. It comes with a fully functional simulator / demo. There are instructions for running through a job at https://github.com/openpnp/openpnp/wiki/Quick-Start (https://github.com/openpnp/openpnp/wiki/Quick-Start)
If you check it out and you feel like OpenPnP would be an improvement on what you have now, let me know, and we can discuss prioritizing support for the TVM. The big issue will be getting me access to a machine. I can knock out the basic driver pretty easily without a machine but making everything work correctly will really require some face to face time with it. I live in Kansas City now, so if there is anyone within an hour or two of there maybe we could work something out?
Jason
-
Nice fast response... always a good sign :)
When you say missing production features, what might some of these be, I think for my case use its not important....
I have run the demo, looks good.
Wish I was closer to you... I am Ottawa Ontario Canada... but if a developer were close I would be more than willing to work with them.
From what I can see the machine mechanically is good, its the software that is very suspect, especially when they need a XY scale factor fudge! mine are like 0.9985 0.9995, ouch!
Richard.
-
Nice fast response... always a good sign :)
When you say missing production features, what might some of these be, I think for my case use its not important....
I have run the demo, looks good.
Wish I was closer to you... I am Ottawa Ontario Canada... but if a developer were close I would be more than willing to work with them.
From what I can see the machine mechanically is good, its the software that is very suspect, especially when they need a XY scale factor fudge! mine are like 0.9985 0.9995, ouch!
Richard.
A few things off the top of my head:
* Not very easy to run a partial job, or continue an aborted job.
* Doesn't show status of each placement during the job run. This is tied to the above and is a long running TODO. Just haven't quite gotten to it.
* Does not use fiducial board scaling. I have not really had people report this being an issue, but it might be for people who are running a lot of boards. By scaling I mean that OpenPnP uses the fiducial to find the position and rotation of the board, but it does not adjust the placements based on the distance between fiducials. That would help handle boards that are not made perfectly, so that could be a concern for some users.
* Bottom vision is pretty new and may be somewhat confusing to new users. UI needs work.
In general, OpenPnP is pick and place software imagined by someone who has never touched a real pick and place. That's me :) I saw the need for the software, no one else was writing it and now here we are. It does work, and there are a lot of people using it for production work, but I feel like if you are coming from a commercial pick and place expectation there may be some disappointment. I guess the one major benefit, though, is that it's open source, so if there's something that really bugs you there is at least the possibility that it can be fixed.
Anyway, I will monitor this thread and the other TVM one and gauge interest. If it seems viable, I suppose the bounty funds could be used to fly me to someone who has a machine :)
Also, as for working with a developer nearby - I am completely open to that. If we can find someone who knows Java and is interested in the task I'd be more than happy to guide them remotely.
Jason
-
One word "BOUNTY"....
There is very nice software being made from the Open PNP guys but no support yet for 820A/B
I have a TVM802A right now and the software and vision is garbage, do you think a bunch of us with 802A/B could put some money into a "bounty" so we might find someone to complete this code for us?? We could use https://www.bountysource.com/ (https://www.bountysource.com/)
Richard.
Jason from OpenPnP here. To be honest, I don't know if you'll be happy with OpenPnP either. It lacks some of the things that I see folks on here talking about being important for production use. What I would recommend is downloading it and taking it for a test drive. It comes with a fully functional simulator / demo. There are instructions for running through a job at https://github.com/openpnp/openpnp/wiki/Quick-Start (https://github.com/openpnp/openpnp/wiki/Quick-Start)
If you check it out and you feel like OpenPnP would be an improvement on what you have now, let me know, and we can discuss prioritizing support for the TVM. The big issue will be getting me access to a machine. I can knock out the basic driver pretty easily without a machine but making everything work correctly will really require some face to face time with it. I live in Kansas City now, so if there is anyone within an hour or two of there maybe we could work something out?
Jason
Very keen on putting money into a bounty but my backup plan is to have a look at this over Christmas when I have some time off.
I agree that access to the machine will be critical.
@vonnieda , I think having openPNP as an option will still add value. The current software for the machine is working but there will always be more features requested.
Jason, do you have an Interface document for people wanting to write drivers for openPNP?
-
One word "BOUNTY"....
There is very nice software being made from the Open PNP guys but no support yet for 820A/B
I have a TVM802A right now and the software and vision is garbage, do you think a bunch of us with 802A/B could put some money into a "bounty" so we might find someone to complete this code for us?? We could use https://www.bountysource.com/ (https://www.bountysource.com/)
Richard.
Jason from OpenPnP here. To be honest, I don't know if you'll be happy with OpenPnP either. It lacks some of the things that I see folks on here talking about being important for production use. What I would recommend is downloading it and taking it for a test drive. It comes with a fully functional simulator / demo. There are instructions for running through a job at https://github.com/openpnp/openpnp/wiki/Quick-Start (https://github.com/openpnp/openpnp/wiki/Quick-Start)
If you check it out and you feel like OpenPnP would be an improvement on what you have now, let me know, and we can discuss prioritizing support for the TVM. The big issue will be getting me access to a machine. I can knock out the basic driver pretty easily without a machine but making everything work correctly will really require some face to face time with it. I live in Kansas City now, so if there is anyone within an hour or two of there maybe we could work something out?
Jason
Very keen on putting money into a bounty but my back plan is to have a look at this over Christmas when I have some time off.
I agree that access to the machine will be critical.
@vonnieda , I think having openPNP as an option will still add value. The current software for the machine is working but there will always be more features requested.
Jason, do you have an Interface document for people wanting to write drivers for openPNP?
No specific document, but there are a few resources that will help:
The API documentation for the driver interface: http://openpnp.github.io/openpnp/develop/org/openpnp/machine/reference/ReferenceDriver.html (http://openpnp.github.io/openpnp/develop/org/openpnp/machine/reference/ReferenceDriver.html)
The source code for the simplest working driver: https://github.com/openpnp/openpnp/blob/develop/src/main/java/org/openpnp/machine/reference/driver/GrblDriver.java (https://github.com/openpnp/openpnp/blob/develop/src/main/java/org/openpnp/machine/reference/driver/GrblDriver.java)
Source code for all the drivers: https://github.com/openpnp/openpnp/tree/develop/src/main/java/org/openpnp/machine/reference/driver (https://github.com/openpnp/openpnp/tree/develop/src/main/java/org/openpnp/machine/reference/driver)
I *think* someone has taken a stab at writing a driver. I'll need to go digging, though, as I'm coming up empty right now. There are some comments from Cri S in this thread on the OpenPnP mailing list: https://groups.google.com/forum/?utm_medium=email&utm_source=footer# (https://groups.google.com/forum/?utm_medium=email&utm_source=footer#)!msg/openpnp/ZW4fu7Gy5vs/gxpi9WrJFwAJ
Another thing to note, since the machine is entirely TCP/IP based (correct me if I'm wrong there) I could potentially work with the machine remotely with someone's on site assistance. We could open a port to the machine and I can just connect to it from here.
Jason
-
Nice fast response... always a good sign :)
When you say missing production features, what might some of these be, I think for my case use its not important....
I have run the demo, looks good.
Wish I was closer to you... I am Ottawa Ontario Canada... but if a developer were close I would be more than willing to work with them.
From what I can see the machine mechanically is good, its the software that is very suspect, especially when they need a XY scale factor fudge! mine are like 0.9985 0.9995, ouch!
Richard.
A few things off the top of my head:
* Not very easy to run a partial job, or continue an aborted job.
* Doesn't show status of each placement during the job run. This is tied to the above and is a long running TODO. Just haven't quite gotten to it.
* Does not use fiducial board scaling. I have not really had people report this being an issue, but it might be for people who are running a lot of boards. By scaling I mean that OpenPnP uses the fiducial to find the position and rotation of the board, but it does not adjust the placements based on the distance between fiducials. That would help handle boards that are not made perfectly, so that could be a concern for some users.
* Bottom vision is pretty new and may be somewhat confusing to new users. UI needs work.
In general, OpenPnP is pick and place software imagined by someone who has never touched a real pick and place. That's me :) I saw the need for the software, no one else was writing it and now here we are. It does work, and there are a lot of people using it for production work, but I feel like if you are coming from a commercial pick and place expectation there may be some disappointment. I guess the one major benefit, though, is that it's open source, so if there's something that really bugs you there is at least the possibility that it can be fixed.
Anyway, I will monitor this thread and the other TVM one and gauge interest. If it seems viable, I suppose the bounty funds could be used to fly me to someone who has a machine :)
Also, as for working with a developer nearby - I am completely open to that. If we can find someone who knows Java and is interested in the task I'd be more than happy to guide them remotely.
Jason
Typically I find all boards I have ever used are very accurate across them no scaling needed :-)
Bottom vision is important to place parts accurately so need to watch that one....
It would be interesting to send the 802 commands over ethernet to move say 300mm and see if it really moves 300mm cause if we need to add a fudge factor to openpnp.... damn! why they need it.... no real idea other than they do not have a handle on floating point rounding....
Yes it would be nice to have a list of placed parts and be able to check off already placed or re-place any parts as you can with the current software, that is for sure good.
All these items can be address by bounties, from people like me that are willing to help out financially....
Richard.
-
Another thing to note, since the machine is entirely TCP/IP based (correct me if I'm wrong there) I could potentially work with the machine remotely with someone's on site assistance. We could open a port to the machine and I can just connect to it from here.
Camera's are USB....
Willing to help anyway I can....
Richard.
-
Nice fast response... always a good sign :)
When you say missing production features, what might some of these be, I think for my case use its not important....
I have run the demo, looks good.
Wish I was closer to you... I am Ottawa Ontario Canada... but if a developer were close I would be more than willing to work with them.
From what I can see the machine mechanically is good, its the software that is very suspect, especially when they need a XY scale factor fudge! mine are like 0.9985 0.9995, ouch!
Richard.
A few things off the top of my head:
* Not very easy to run a partial job, or continue an aborted job.
* Doesn't show status of each placement during the job run. This is tied to the above and is a long running TODO. Just haven't quite gotten to it.
* Does not use fiducial board scaling. I have not really had people report this being an issue, but it might be for people who are running a lot of boards. By scaling I mean that OpenPnP uses the fiducial to find the position and rotation of the board, but it does not adjust the placements based on the distance between fiducials. That would help handle boards that are not made perfectly, so that could be a concern for some users.
* Bottom vision is pretty new and may be somewhat confusing to new users. UI needs work.
In general, OpenPnP is pick and place software imagined by someone who has never touched a real pick and place. That's me :) I saw the need for the software, no one else was writing it and now here we are. It does work, and there are a lot of people using it for production work, but I feel like if you are coming from a commercial pick and place expectation there may be some disappointment. I guess the one major benefit, though, is that it's open source, so if there's something that really bugs you there is at least the possibility that it can be fixed.
Anyway, I will monitor this thread and the other TVM one and gauge interest. If it seems viable, I suppose the bounty funds could be used to fly me to someone who has a machine :)
Also, as for working with a developer nearby - I am completely open to that. If we can find someone who knows Java and is interested in the task I'd be more than happy to guide them remotely.
Jason
Typically I find all boards I have ever used are very accurate across them no scaling needed :-)
Bottom vision is important to place parts accurately so need to watch that one....
It would be interesting to send the 802 commands over ethernet to move say 300mm and see if it really moves 300mm cause if we need to add a fudge factor to openpnp.... damn! why they need it.... no real idea other than they do not have a handle on floating point rounding....
Yes it would be nice to have a list of placed parts and be able to check off already placed or re-place any parts as you can with the current software, that is for sure good.
All these items can be address by bounties, from people like me that are willing to help out financially....
Richard.
Agreed - I don't think there are any showstoppers. It was more of just a warning that it may need some work before it completely suits your needs.
I will spend a little time in the next night or two digging up everything I've saved or seen about these machines. I am pretty sure someone did some data captures, and I think someone tried to write a driver. In addition, I think I have some information about the cameras some where. I'll gather this all together and put it in a Google doc or something we can share and edit.
It would be good if you can get in touch with other TVM owners and see who would be interested in assisting the effort. Goal 1 would be for me to be able to visit a machine somewhere. I live in Kansas City, MO and would be willing to drive a couple hours to do this. Barring that, we can do things remotely.
One other note: Depending on how serious folks are about a bounty, if you can gather enough funds to buy me a machine, I will absolutely get it working under OpenPnP. That's a pretty big bounty, but I figured it was worth mentioning :)
Jason
-
I have a fully functional driver. Did not have time to work on anything beyond that. Someone I shared it with created a github project.
-
I have a fully functional driver. Did not have time to work on anything before that. Someone I shared it with created a github project.
https://github.com/TVM802 (https://github.com/TVM802)
-
I have a fully functional driver. Did not have time to work on anything before that. Someone I shared it with created a github project.
I have a fully functional driver. Did not have time to work on anything before that. Someone I shared it with created a github project.
https://github.com/TVM802 (https://github.com/TVM802)
There it is! :) I knew I had seen something. Are either of you using this at all currently?
-
There it is! :) I knew I had seen something. Are either of you using this at all currently?
I tried it on my TVM802A but it did not respond at all....
The file below seems to contain the golden nuggets...
https://github.com/TVM802/TMV802-Protocol/blob/master/p1/hal-api.cpp (https://github.com/TVM802/TMV802-Protocol/blob/master/p1/hal-api.cpp)
-
Can someone post the .exe for windows 7 64 bit and I could try it also....
Richard.
-
Hi all...
I started working on a API for the TVM802. Is there anybody with a TVM802B that has a bit of NodeJS skills and can help with testing?
Code : https://github.com/TVM802/TVM802-WEB-API (https://github.com/TVM802/TVM802-WEB-API)
Currently it connects to the machine and toggles the beeper as a test. At this stage I just want to make sure the code works on a "TVM802B" as I have a "A" version.
-
Hi all...
I started working on a API for the TVM802. Is there anybody with a TVM802B that has a bit of NodeJS skills and can help with testing?
Code : https://github.com/TVM802/TVM802-WEB-API (https://github.com/TVM802/TVM802-WEB-API)
Currently it connects to the machine and toggles the beeper as a test. At this stage I just want to make sure the code works on a "TVM802B" as I have a "A" version.
I should help, I have not done NodeJS before but I am a very quick learner and have lots of experience under the belt :-)
Keep up the efforts!
RichardS
-
Hi harry, I have seen you have problems with up looking camera.
Can you publish two images of the problematic part , one as is and the other
Without the red down looking lights. No additional black paper.
If you can use BMP or eventually png that is better as jpeg.
Having this two images I can make a video filter that blacks out the
Green color. Eventually its necessary to disable red down light with
Reed or microswitch , I can say it after seeing image.
-
Hi harry, I have seen you have problems with up looking camera.
Can you publish two images of the problematic part , one as is and the other
Without the red down looking lights. No additional black paper.
If you can use BMP or eventually png that is better as jpeg.
Having this two images I can make a video filter that blacks out the
Green color. Eventually its necessary to disable red down light with
Reed or microswitch , I can say it after seeing image.
Should they have a filter that only shows the bright white parts anyway?? This is an idea.
RichardS
-
I placed about 1000 small SMT leds today and used vision on them all, these are very tiny leds, so vision is best, I used quick vision and about 5 times out of 1000 the image it took was mis-read and the led place either totally rotated 180, or off by 10mm, you could see the camera misread the image and then the placement so astray.... :-( anyone know tips to perfect the camera?
RichardS
-
Question:
Down camera, should part #'s on back of IC be in focus?
or Should the pads be better focused?
Just trying to tweak the cameras for better operation.
RichardS
-
pads should be in focus
-
Question:
Down camera, should part #'s on back of IC be in focus?
or Should the pads be better focused?
Just trying to tweak the cameras for better operation.
RichardS
Down-facing camera is usually only fo rfiducials, so should focus on the PCB surface
-
right duh! :palm: (as dave would do)
-
Hi folks, just wanted to let you know that we've begun an effort to support the TVM802A/B and TVM920 machines in OpenPnP. If that is a topic that interests you, please come join the discussion at https://groups.google.com/forum/#!topic/openpnp/735aZn-GFRw.
Thanks,
Jason
-
Hi, i want know the max depht of Z and the distance between the center of the ball bearings of the Z cam wheel.
It is possible to get this info from some owner of 802 ?
Thanks
-
35mm between and Z max of 20mm which the software has limited to 16mm right now. And the diameter of the rollers that push down are 17mm, I suspect these measurements do not need to be overly exact....
Hi, i want know the max depht of Z and the distance between the center of the ball bearings of the Z cam wheel.
It is possible to get this info from some owner of 802 ?
Thanks
-
35mm between and Z max of 20mm which the software has limited to 16mm right now. And the diameter of the rollers that push down are 17mm, I suspect these measurements do not need to be overly exact....
Hi, i want know the max depht of Z and the distance between the center of the ball bearings of the Z cam wheel.
It is possible to get this info from some owner of 802 ?
Thanks
The distance should be precise, ideally with caliper. It can be measured with rollers and then removing the roller diameter. For now i assume radius of 17.5mm .
-
I have calipers, measured the DIAMETER again and its 16.99mm
center to center is 51.3-16.99 = 34.31
35mm between and Z max of 20mm which the software has limited to 16mm right now. And the diameter of the rollers that push down are 17mm, I suspect these measurements do not need to be overly exact....
Hi, i want know the max depht of Z and the distance between the center of the ball bearings of the Z cam wheel.
It is possible to get this info from some owner of 802 ?
Thanks
The distance should be precise, ideally with caliper. It can be measured with rollers and then removing the roller diameter. For now i assume radius of 17.5mm .
-
Rather than hacking/rev eng the existing board/ controller, how about just putting some new code in it ?
That's what i am going to do with the TVM920..... the hardware is not sophisticated, nor has many layers than STM32 pinouts (TVM92) cannot be figured out in a few minutes...
I am not going to even bother with trying to use the existing code in the microcontroller PCB.
IMO you just inherit all the issues.... but I have a very bold POV.
I have no idea what controller is in a 802B....
glen
-
That's probably the best approach as you could spend a lot if time reversing the protocol to find it has inherent priblems. Also you don't know if it will change in future versions. The functionality is simple enough that writing from scratch may be quicker than reversing anyway.
-
Reverse engineering the stock controller has the primary benefit of remaining compatible with the stock software should other options not work out, or if you just want to do some testing. For OpenPnP, I strongly support a reverse engineered solution so that OpenPnP can work out of the box with the TVM machines. It will give people the option of trying a new software stack without irreversibly modifying their machines. On top of that, the work is already quite far along.
-
There is also probably nothing that can not be "tweak" in code if there is something inherently wrong with the controller, hacking the controller code at this point is not what I am sure a lot of 802 users want to do... the current software works for me very decently, there are some tiny issues I would like fixed and OpenPNP will probably be able to address them better....
Even if the controller does has a drift issue due to bad math, we can understand the error and account (calculate) for it in the new control software.
For me stock 802A is the way to go....
I am curious what controller is in mine, but I do not have time to rip it apart... :-)
RichardS
-
What are this issues, apart tall components, top vision for resistors or caps , nozzle toolchanger , missing autocalibration using vision and
to improve vision image preprocessing.
-
tmf: If the LEDs came off tape, how on earth did it put it in 180 deg ????
I would not be confident of the code in the micro....
I don't agree it is irreversibly modifying their machine . If you can put new code in, you can likely (as I doubt they have jtag read fuses set) pull the old code out...
I guess though, on the other hand someone has t actually DO IT ! :-)
g
-
Yes... It is likely that the firmware can be retrieved and re-flashed if needed.
I have a few of the T962 re-flow ovens and the original firmware could not even do proper time tracking. I found a solution online that showed how to flash it with better firmware and now these little ovens are worth double the money I paid for them.
If someone with the skills and time wants to take it apart and build better firmware I am all for it. The better we understand these machines the better they will perform in the future. For now I am putting time and effort into the openPNP driver that will use the existing firmware. (I use my machine for production work and will only risk new firmware updates once its well tested or I have a second machine)
-
Just tested, c2j translation works to some extend.
How should openpnp driver handle buttons, LEDs, vacuum and prick sensors?
-
Abstraction layer needs to be thought out.
This is probably off topic. QiHe now have been monitoring these threads and we (I) am filling them full of junk.
-
tmf: If the LEDs came off tape, how on earth did it put it in 180 deg ????
Parts rotate on the nozzle in movement due to vibration.
-
There is also probably nothing that can not be "tweak" in code if there is something inherently wrong with the controller, hacking the controller code at this point is not what I am sure a lot of 802 users want to do... the current software works for me very decently, there are some tiny issues I would like fixed and OpenPNP will probably be able to address them better....
Even if the controller does has a drift issue due to bad math, we can understand the error and account (calculate) for it in the new control software.
For me stock 802A is the way to go....
Controller does not have a drift. All math errors are in S/W. However I'm now thinking I'd be better off with a new driver/controller entirely. The way original 802 controller works, it creates too much vibration.
-
There is also probably nothing that can not be "tweak" in code if there is something inherently wrong with the controller, hacking the controller code at this point is not what I am sure a lot of 802 users want to do... the current software works for me very decently, there are some tiny issues I would like fixed and OpenPNP will probably be able to address them better....
Even if the controller does has a drift issue due to bad math, we can understand the error and account (calculate) for it in the new control software.
For me stock 802A is the way to go....
Controller does not have a drift. All math errors are in S/W. However I'm now thinking I'd be better off with a new driver/controller entirely. The way original 802 controller works, it creates too much vibration.
I would say one step at a time, yes it creates vibrations, but we can first get OpenPNP to work with the current commands, then someone can tackle firmware and use the same commands to smooth out the vibrations.... or swap out the hardware and use a known hardware to OpenPNP...
RichardS
-
.... However I'm now thinking I'd be better off with a new driver/controller entirely. The way original 802 controller works, it creates too much vibration.
I think thats the right way to go. Buildig a new and better controller costs less time than hacking the existing one.
-
anyone got any photos, pseudo schematics posted anywhere of its insides ?
how similar to TVM920 ? (microcontroller talking to ATOM PC on ethernet), expansion IO board on ribbon talking to misc single IOs and feeder solendoids, closed loop stepper interface...
g
-
Anyone got a Lasagne recipe?
English-glen. that is offtopic :scared:
-
Ar__systems, as you surely have tested different acceleration and speed settings,
do you think this is because of simplified stepper accel/decel algorithm or because of
some meccanical stickness or slight disallinement?
-
Yeah, from what I read, the driver does not have current control and as a result applies to much torque at lower speeds. So the stepper completes full transition from one state to another faster than necessary for the given speed. In other word, it moves in small jerks instead of rotating continuously.
There is a bit of play in the nozzle lowering mechanism, but that does not seem to have significant effect.
-
Folks!
Does anyone have a hint for me: I'm playing with the camera parameters, and found no solution. I'm using the upward camera to look on a SOT23. Sometimes it takes 1-3 corrections and the machine places the part. Next time it requires several minutes i.e. 100'reds of retries to identify it as OK and next it places it correctly. I'm using SW version 2.29
Could anyone post the configuration of the camera page?
Thanks
-AH-
-
I have also found SOT23 to be troublesome, when it messes with it adjusting the "100's" of times, sometimes it flips it 90 and 180 out in its attempt.....
RichardS
-
I have also found SOT23 to be troublesome, when it messes with it adjusting the "100's" of times, sometimes it flips it 90 and 180 out in its attempt.....
I played a lot with the parameters so it does not rotate any more but it slows down speed dramatically. So again has anybody found a good configuration combination to recommend?
Thanks!
-AH-
-
yes, I found a solution, but you will not like it:
I set the cam to quick mode and reduce the speed to 20% (for SOT23 only)
This is very slow but works for me.
The problem with SOT23 are vibrations. Nozzle suction is too weak for the vibrations, so the SOT23 are dancing and flipping.
After replacing the built in vac pump with an external one, its much (!) better. Now I can run SOT23 with 100% speed
using 504 nozzles.
-
The problem with SOT23 are vibrations. Nozzle suction is too weak for the vibrations, so the SOT23 are dancing and flipping.
Thanks for the recommendation I'll try it.
Regarding the vibrations, I have seen that in an earlier posting. I have not that experience on my machine. I do MELF or SO23 with 80-100% speed and I have no problem loosing a component. I played around with the nozzles to find a component / nozzle combination that works fine. There are vibrations visible if the nozzle is too small, that the component dances around as you described. With a suitable nozzle even round ones like MELF work fine.
I had to readjust the prick offset for some strings for some reason. Quite interesting that this happens only for specific feeders. I had to redo that several times, only +/- 0,5 to 1mm
-AH-
-
The 90 and 180 are not vibration for me, there are exactly out 90 and 180 due to camera error.
If the quick takes a picture and in that picture it can not outline the pins correctly, (due to lighting and contrast) its not going to work either, at any speed.
RichardS
-
The 90 and 180 are not vibration for me, there are exactly out 90 and 180 due to camera error.
I doubt your assessment is correct. AFAIK the s/w is not capable of making visual corrections of more than 45 deg.
More likely the initial rotation was due to incorrect pick or the package bounced in the pocket.
-
The 90 and 180 are not vibration for me, there are exactly out 90 and 180 due to camera error.
I doubt your assessment is correct. AFAIK the s/w is not capable of making visual corrections of more than 45 deg.
More likely the initial rotation was due to incorrect pick or the package bounced in the pocket.
Interesting... I will keep my eye on this and try and see why then, does not happen that often so could be a bit before I see it again.....
RichardS
-
Can someone please send to me the latest software of the TVM802A PNP?
Thanks
-
Can someone please send to me the latest software of the TVM802A PNP?
Just contact Gera directly via mail. I got the SW within hours and got asked not to share it with other or put it publicly on the net. Current version is 2.29
-AH-
-
Does anyone have ideas on how to remove lock-tight on up facing camera lens?
I hooked the camera's up to openPNP and tried the part recognition pipeline. Seems to work very well but because the part is not focused the alignment is out by about 0.1mm on one side.
-
Does anyone have ideas on how to remove lock-tight on up facing camera lens?
I hooked the camera's up to openPNP and tried the part recognition pipeline. Seems to work very well but because the part is not focused the alignment is out by about 0.1mm on one side.
Surely there's a setting to determine the Z height when holding parts over the camera?
-
Does anyone have ideas on how to remove lock-tight on up facing camera lens?
I hooked the camera's up to openPNP and tried the part recognition pipeline. Seems to work very well but because the part is not focused the alignment is out by about 0.1mm on one side.
Surely there's a setting to determine the Z height when holding parts over the camera?
Good point... The machine head is Cam based and I should be able to move the inspected nozzle a few mm in the negative direction(Part needs to farther away) to focus it.
-
I removed on mine. The stuff is tough, but I'm tough too. I used long nosed pliers and screwdrivers.
-
Surely there's a setting to determine the Z height when holding parts over the camera?
Yes there is but the focus point was still too high. It could not focus on taller (taller than 0603 that is) parts at the highest nozzle position.
-
Is the part height set correctly? This should be being used to adjust the focus distance for different part sizes
-
Is the part height set correctly? This should be being used to adjust the focus distance for different part sizes
Yep, check part height and camera height in the settings. When OpenPnP wants to image a part with the bottom camera it uses the camera's Z location (set in camera properties) and the part's height to calculate where it should place the nozzle. When you set the camera Z, it should be with no part on the nozzle and with the nozzle bottom in focus over the camera. Then when OpenPnP images it adds the part height to this value.
-
Is the part height set correctly? This should be being used to adjust the focus distance for different part sizes
Yep, check part height and camera height in the settings. When OpenPnP wants to image a part with the bottom camera it uses the camera's Z location (set in camera properties) and the part's height to calculate where it should place the nozzle. When you set the camera Z, it should be with no part on the nozzle and with the nozzle bottom in focus over the camera. Then when OpenPnP images it adds the part height to this value.
I wonder if the stock software does this??
RichardS
-
Today I tested out our own feeder and tape stripping system on the TVM802. The stripping works independently of the TVM, so it works fine with the stock s/w. I made a completely new feeder block that solves multiple issues of the original one. And don't even get me started on the TVM's stripping mechanism...
So far we only ran test jobs. Placed over 500 of 0603 caps and close to 100 of SOT23. I think there was one discarded components in total.
My feeder block packs tapes tighter (3 mm spacing instead of 5 of the original), so more lanes can fit into the same space. It has 18 of 8mm lanes and some space for configurable lanes, which allows setting up an arbitrary combination of tapes of any width. (That part is not tested yet.) If you fill configurable space with 8mm tapes only it gives 11 more of 8mm tapes. So L side can host 29 of 8mm tapes in total.
My stripping mechanism provides constant pulling force and it supposedly will never require any adjustments. Well maybe once when you install a new reel with a different peel force. However suitable peel force for the tapes that I checked is in fairly wide range, so I don't think it will be ever required.
-
I wonder if the stock software does this??
Yes it does
-
Nice! Any images?
Today I tested out our own feeder and tape stripping system on the TVM802. The stripping works independently of the TVM, so it works fine with the stock s/w. I made a completely new feeder block that solves multiple issues of the original one. And don't even get me started on the TVM's stripping mechanism...
So far we only ran test jobs. Placed over 500 of 0603 caps and close to 100 of SOT23. I think there was one discarded components in total.
My feeder block packs tapes tighter (3 mm spacing instead of 5 of the original), so more lanes can fit into the same space. It has 18 of 8mm lanes and some space for configurable lanes, which allows setting up an arbitrary combination of tapes of any width. (That part is not tested yet.) If you fill configurable space with 8mm tapes only it gives 11 more of 8mm tapes. So L side can host 29 of 8mm tapes in total.
My stripping mechanism provides constant pulling force and it supposedly will never require any adjustments. Well maybe once when you install a new reel with a different peel force. However suitable peel force for the tapes that I checked is in fairly wide range, so I don't think it will be ever required.
-
Hi All,
I got my TVM802A a month ago and except all known bugs I got one very specific. I can't fix it and want to ask community about it.
After 1-2 weeks of using my TVM802A I found it's hit nozzle into stack bolt. I don't see this crash but hear it and when I turned around, I found that nozzle is broken.
Ok, I order 10 nozzles ;)))))
After that I continue to use this machine but collisions occurs regular.. more and more frequently. At the begining I guess it's all because PC and Windows... I found information here that TVM802A may skip commands from PC (W10) by ethernet. But I use Win7. In any way I tried reinstall Win7 and even more - I tried several different PC and even one with Win XP. XP don't work correctly at all. So I return to Win7.
As I already wrote, collisions happens more and more frequently. Now it happens every days ((( Todays it was 2 collisions but I finally see it and understand one important thing - It not seems only like escaping of command. It's looked like something absolutely wrong happens.
First collision:
1) nozzle pick up chip 0805 (right action)
2) nozzle coming to camera (right action)
3) recognize chip (right action)
4) move nozzle down by Z-axis (under camera plastic box level) (why ?)
5) than it's move and hit nozzle to camera box side.
Second collision:
1) head come to stack (right action)
2) move prick down (right action)
3) machine move head to camera (.....why ?)
So first time machine make wrong action, which can't be at all. Next time machine must move prick by X-axis, than pick up chip and only after that move head to camera.
Anybody have ideas what happens and how to fix it?
-
Hi!
I have crashes as well but eventually once a week using it for 2-3 hours a day. What SW version are you using. My experience is that one of the SW updates reduced the number ob mismovements.
Do you have any other activity on that PC during usage?
-AH-
-
In the first sequence it can still be skipped commands. I.e. after the camera it was supposed to move to the component location and then nozzle down, but it just went straight to nozzle down.
I'm using XP SP3. Also, I recommend not to have any other equipment on that network. I have only PC and machine connected to a switch.
-
I have crashes as well but eventually once a week using it for 2-3 hours a day.
It's also frequently.
I use is every day 4-6hours
My experience is that one of the SW updates reduced the number ob mismovements.
Do you have any other activity on that PC during usage?
Sorry, SW.. what do you mean? SurfaceMounter app version? I will check it tomorrow.
And of course I don't run any other applications on PC. It's stay alone with fresh Win7 without any other applications. I even switch off WiFi )))
In the first sequence it can still be skipped commands.
I don't think so, because after camera it's tried go to the PCB, like nothing had happens.
-
what version of s/w do you use?
-
ar__systems, sorry, but english is not my native language and I don't sure what exactly you mean when ask about s/w, but I guess software.
I check QiHe Vision Mounter version is V2.30
-
I am using the TVM802 since 9 month and never had any crash or wrong movement.
My software is version 2.12, I will not update it since it works perfectly.
your problem may be caused by missing commands over the network, like the well known Win10 problem. However, I cannot understand how a command can get lost in a network. This can only be an issue of the PC software or the firmware in the machine itself.
So we need to compare the firmware version too, but I have no idea how to read this version number.
-
harry4516, yes I also think about machine controller. When spring festival in China will be over, I will try to order another controller from qihe but if it will not help..
In this case only one way I see is order another machine like neoden4 to etc.
Do you know some one who has successfull experience of using this machine with free software??
-
I am using the TVM802 since 9 month and never had any crash or wrong movement.
My software is version 2.12, I will not update it since it works perfectly.
your problem may be caused by missing commands over the network, like the well known Win10 problem. However, I cannot understand how a command can get lost in a network.
I can tell you how. Normally two separate commands arrive in separate packets. But W10 can be merging some packets if they are issued close enough. Then the machine does not recognize them. I've confirmed that with WireShark when I worked on reversing protocol.
-
I can tell you how. Normally two separate commands arrive in separate packets. But W10 can be merging some packets if they are issued close enough. Then the machine does not recognize them. I've confirmed that with WireShark when I worked on reversing protocol.
Do you successfully start machine under free software as a result of reversing protocol ?
-
I can tell you how. Normally two separate commands arrive in separate packets. But W10 can be merging some packets if they are issued close enough. Then the machine does not recognize them. I've confirmed that with WireShark when I worked on reversing protocol.
interesting fact. So it is an issue of the controller firmware which cannot separate the packets. Maybe QiHE should be able to solve this problem.
Many people use TCP as a packet oriented protocol (like UDP), but it is stream oriented (the application is responsible to find the start and end of an information). This makes problems if packets are sent too close.
Looks like I am lucky, because I use one of my oldest PCs for the TVM802, this thing is too slow to make this kind of problem.
-
I can tell you how. Normally two separate commands arrive in separate packets. But W10 can be merging some packets if they are issued close enough. Then the machine does not recognize them. I've confirmed that with WireShark when I worked on reversing protocol.
Do you successfully start machine under free software as a result of reversing protocol ?
No. I've worked out the protocol and made basic application that allows to control everything, but not past that. In the hindsight, the machine has limitations that go beyond the crappy software, so even porting it to OpenPNP would not be sufficient to make it useful.
In particular, the stepper controller is not smooth enough. It creates too much vibration. As a result parts, especially ceramic caps dance on the nozzle which results in frequent discards and/or misplaced parts.
-
No. I've worked out the protocol and made basic application that allows to control everything, but not past that. In the hindsight, the machine has limitations that go beyond the crappy software, so even porting it to OpenPNP would not be sufficient to make it useful.
In particular, the stepper controller is not smooth enough. It creates too much vibration. As a result parts, especially ceramic caps dance on the nozzle which results in frequent discards and/or misplaced parts.
Bad stepper drivers and nozzle vibration is the smallest problem. I will be happy if machine will stop to destroy itself and nozzles.
I tried yesterday to reduce machine speed to 60% and every components speed to 60%. So all movements seems very slow, but finally after 4 hours of working machine again broke nozzle. I hate it (
-
maybe you should ask the manufacturer.
-
I already ask support but they don't know what happens and don't give me any recommendations (just only useless like check program step by step and etc). It seems like my case not interesting for them. I will send them videos after 6.FEB but I not sure that it will help me.
-
howdy
so you say Win10 is concatenating short packets ? that's not actually bug, really.
is the TVM802 TCP/IP or UDP ?
I gather TCP, because is it usual to concatenate packets. sounds like the controller TCP implementation in the controller PCB is a bit lacking.
You can also the mode of the TCP interface with IOCTL / SOCKET OPTION commands to prevent Nagle algorithm etc .
or in the registry....
take a look at this advice- it deals with disabling Nagle and a few other network gems
https://www.reddit.com/r/killerinstinct/comments/4fcdhy/an_excellent_guide_to_optimizing_your_windows_10/ (https://www.reddit.com/r/killerinstinct/comments/4fcdhy/an_excellent_guide_to_optimizing_your_windows_10/)
http://www.speedguide.net/articles/windows-8-10-2012-server-tcpip-tweaks-5077 (http://www.speedguide.net/articles/windows-8-10-2012-server-tcpip-tweaks-5077)
glen
-
Before Christmas, we got our TVM802B too but until yesterday we had no time to try :-//
Placing 0603 works quite well but with 0402 we have some real problems.
Maybe someone can tell me if the described behavior is the same for him.
We changed the Feed to 2 instead of 4 as there are two parts between the 4mm.
Every second part, the brick hit the hole of the smd tape.
If I change the Feed to 4 for 0402, every second component is skipped in the SMD tape.
So it looks as if something is wrong. The delivered software has version 2.30!
Has anyone else experienced such a behavior?
What type of nozzle have been successfully used for 0402? Type 502 or Type 503?
With Type 502, I have seen quite often that the component is picked but vertically but I think this is only due to not proper alignment.
Anyway, until the feedrate for 0402 is as it is for us right now, 0402 is not really useable as 50% waste of parts are not ok.
Thanks!
-
Hello!
Regarding the feeder: I was forced for some 1208 tapes to adjust the position of the brick for that single position. It is done in "system config". I guess that might help you as well.
-AH-
-
@huebsch
Well do you mean PrickCorrect in the LeftStack tab or 'Left Prick Offset' in the Other tab?
I tried the PrickCorrect (first column x-axis) from -0.01 to -2 and +2 in 0.05 steps. Larger than 2 does not work.
But anyway we have not seen any change in the position. Changing the second column (y-axis), we saw an effect.
As we only have the 1.30 software, we do not know if there is a bug in the implementation.
1208 has also only one component between two holes in the tape, right?
-
@steffenmauch Yes that's what I was recommending. The 1208 strips have only one component per hole. my corrections where only on Y axis necessary.
-AH-
-
howdy
so you say Win10 is concatenating short packets ? that's not actually bug, really.
is the TVM802 TCP/IP or UDP ?
I gather TCP, because is it usual to concatenate packets. sounds like the controller TCP implementation in the controller PCB is a bit lacking.
TCP. I did not say it was a bug, I said machine can't handle it :)
-
@steffenmauch, do you already run you machine?
A also has 2.30verion software. Is it work fine for you or you also have troubles with it?
-
@Sugrob: well, placing 0603 works, but I need to adjust the pick up position to be more precise.
Nevertheless, 0402 does only work when the Feed is set to 4 instead of 2. As with this setting,
every second component is not used.
I don't know whether this is a bug of 2.30 as I do not have another version.
Tried to contact a lot of different persons, but still no real response.
Hope to get in contact with qiHe tomorrow when the holidays in china are ending ...
-
@steffenmauch, do you already run you machine?
A also has 2.30verion software. Is it work fine for you or you also have troubles with it?
Try connecting the machine to the PC directly, i.e. not through a switch or a hub.
-
Try connecting the machine to the PC directly, i.e. not through a switch or a hub.
it's connected directly. I never use hub.
On several sites I found information about laptop is not so good work with CNC controller, but can't find a reason.
All of 3 PC which I test with my machine was a laptops. How do you think, is it may be a reason of frequently collisions?
-
Not sure how laptop or not can be a factor. We used an older XP laptop
.
-
Try connecting the machine to the PC directly, i.e. not through a switch or a hub.
it's connected directly. I never use hub.
On several sites I found information about laptop is not so good work with CNC controller, but can't find a reason.
All of 3 PC which I test with my machine was a laptops. How do you think, is it may be a reason of frequently collisions?
There is only one thing I can recommend,
look for an old PC, maybe a dual-code, with Windows 7.
Install Win7 and updates, then disable all further updates, disable indexing (!).
The problem is the firmware in the machine. But you can help with such a PC to get it working.
At least , these are my experiences.
-
@Sugrob: well, placing 0603 works, but I need to adjust the pick up position to be more precise.
Nevertheless, 0402 does only work when the Feed is set to 4 instead of 2. As with this setting,
every second component is not used.
I don't know whether this is a bug of 2.30 as I do not have another version.
Tried to contact a lot of different persons, but still no real response.
Hope to get in contact with qiHe tomorrow when the holidays in china are ending ...
I would not expect 0402 placement work reliably on this machine. The feeders just don't provide enough accuracy. I fought with it for a year and I gave up.
Drag feeders are not hopeless, imho they can be made to work reliably. But they must be closed loop to verify position of the tape after the movement. Either by means of down camera, or by some other mechanism. Without that they have too much play for 0402.
-
Ok short update regarding software version 2.30.
We received version 2.32 and with this 0402 does work with the right feed rate.
Looks as if they have broken that in 2.30; but they have not confirmed it :scared: just gave me the new version and said try this one.
We adjusted the camera offset in the settings and now placing 0402 capacitor works quite well.
0402 resistors are sometimes rotated by 90° and they fail much often during pickup.
-
Folks!
I just found a release document on the WEB which I machine translated to english. That document just goes up to rev. 2.31 2.32 is not included...
2016/12/8 V2.31
1, increased PCB software "Easy-PC" files into formats compatible with 11.10
2, mount complete the prompt box the next time you start automatically remove 11.11
3, manual, does not run when the nozzle down 11.11
4, increasing Visual precision adjustable bias (element coordinate deviations, element angle deviation) 11.28
5, modify 0402 needles suction bug 12.7
2016/11/5 V2.30
1, change the maximum acceleration
2, the keys function increase file editing (ESC, space, Delete)
3, delete the symbol prompt information "confirmed"
4, increasing the light switch
2016/10/8 V2.29
1, modify the file "posted" order bug
After stopping the machine needle 2, modified firing pin bug
3, scale accuracy of 0.0001
4, stack after stack, 6, 7th against (Beta2)
5, 802A disk cannot be located, set before the current bug (Beta3)
2016/10/5 V2.28
1, revised separate pin offset is not in place of the fabric stack Bug
2, open to give up other location data in the system setup parameters
3, increase the length of 16mm feed
2016/09/02 V2.27
1, modify components parameters table "description" column, do not select the same data change bug
2, depending on the camera on "Visual components" adding "Red Cross"
2016/08/20 V2.26
1. the software implemented two versions a, b, b, 802A, two versions
2. differences:
A software (1) system setup small stacks in the back behind the bar, and other parameters in the PIN offset parameter
(2) in the element information table in "stacks" column in less material behind stack options
(3) when the CSV file is loaded, a loaded b-file will report an error
2016/08/05 V2.26
1. increase the percentage of EFAC coefficient, fix x, y coordinates
2016/07/19 V2.25
1. amendments to the results did not automatically detect save bug
2016/07/09 V2.24
1. fixed limit trigger the buzzer doesn't alarm
2. the factory parameters depending on the test button to add the light source open
3. close the page when testing the nozzle automatically stopped
2016/06/27 V2.23
1. fixed stack-pin on the left offset parameter error
2016/06/07 V2.22
1. increase depending on camera "photo delay" or "camera exposure" parameter
2016/05/27 V2.21
1.IC stack sorting optimization
2016/05/04 V2.20
1.IC stack sets the current serial number can be read
2. swing PIN card
3. Add a separate needle for each stack offset
4.PCB configuration, automatically switches to manual mode and depending on the camera mode
5. change camera acquisition algorithms.
6. depending on the camera light on only when the test opens.
2016/03/21 V2.12
1. fixed two nozzles while feeding, the second device in the stack in the back angle errors
2016/02/21 V2.11
1. front material stack up to 30
2.ACC12:X10 Y7 Z15
ACC11:X7 Y6 Z12
2015/11/16 V2.10
1. the maximum acceleration
2. tray can use two suction nozzle and mounting
2015/11/12 V2.09
1. When you don't have to check the material, not in place waiting for inspection
2015/09/28 V2.08
1. fixed after the capture device, when moving to the posted position, determine the pressure, and z-axis down
2015/08/31 V2.07
1. fixed 2nd after Visual suction head, process errors of judgment
2015/08/24 V2.06
1. modify the English translation
2015/08/21 V2.05
1. fixed coordinate the IC Tray problem
2015/08/19 V2.03
1.24mm the strip width
2015/07/25 V2.02
1. fixed IC tray when Y=1, coordinates of confusion
2. fixed coordinate system parameters page camera location
3. modify LED errors
2015/05/28 V2.0
1. fixed needle drop BUG after a pause. The new version firmware.
2015/05/10 V1.25
1.IC Tray under plate increments
2015/04/29 V1.24
1. fixed IC Tray 1 sequence problem
2015/04/10 V1.23
1. fix rear stack 11~30 stack 1/2 suction nozzle only when a material problem
2015/04/07 V1.22
1. fixed load CSV file when the file because the device is empty alarm
2015/03/20 V1.21
1. increased needle fabric stack offset
2015/03/06 V1.20
1. fixed after the 2nd nozzle dump, repeat posted 1th bug
2.12MM set distance
2015/03/04 V1.19
1. fixed when the file is read, not read into the PCB configuration BUG
2015/02/27 V1.18
1. the parameters go to the coordinates from the manufacturer depending on the camera system parameters
2015/02/24 V1.17
1. can support Protues coordinates the export file
2015/02/12 V1.16
1. fixed camera manufacturers parameter coordinate automatic test BUG
2. fixed stack separate sort BUG
3. fixed read Candes coordinate export file error
4. increase can read the decimal point as a comma-file
2015/01/10 V1.15
1. fixed IC Tray data bug
2014/12/26 V1.14
1. increase the generated array functions
The position of the 2.PCB configuration increases the z-axis origin location
3. fixed the right-click menu sets the current camera position bug
2014/12/04 V1.13
1. modify the English version of the stack display error bug
2014/12/04 V1.12
1. revision of German regional data decimal points are replaced by comma bug
2014/12/02 V1.11
1. modify 0402 stack feed 2mm, needle drop coordinates error bug
2014/11/29 V1.10
1. support 802B double stack SMT machine
2. improved pre-stack functions
2014/10/30 V1.08
1. improved feature by camera device coordinates (consider MARK point transform)
2. Add a right-click menu and set the coordinates must be in a State of origin has been found
3. added support for Allegro export file
2014/10/27 V1.07
1. modify the parameter settings page, position run without prompting z is not zero-bug
2. change when the file is loaded, bottom mirrored error bug
3. remove manual run time, request the z axis at zero
2014/10/24 V1.06
1. modify the selected row appears error bug
2014/10/24 V1.05
1. revision of PCB configuration-5th Board Gets the camera position bug
2. change SN number display bug
3. modify the 1th nozzle nozzle feeding, feed 2mm,2 4mm, 2 a good nozzle suction error bug
4. Add the home button icon
5. increase the version number display
-
Well somehow the information below is not surprising but anyway:
- do not used red solder mask when trying to use fiducials with the TVM802b
- also do not used white solder mask
Due to the red leds being used for fiducials recognition (and maybe some color filter digitally or optically), you will not see them on the image in the software. |O
On option is to use a black permanent pen and mark the circles of the fiducials.
Then, the circles are visible again.
One good tweak would be to use RGB leds and change the color depending on the solder mask. Shouldn't be too hard but the firmware of the internal controller must be modified.
Does anyone have a good image of the controller board of the TVM802a/b?
-
I never put soldermask over the fiducials, so no problem with color.
-
The fiducial has no solder mask on top; but it has HAL lead free finish.
-
We did some more pick and place jobs. Placing 0402 capacitors works ok, but 0402 resistors are pure pain.
Any suggestions except replacing the vacuum pump? Tried 502 and 503 type of juki nozzles ...
-
I just received v2.32 of the SurfaceMount software from the Aliexpress seller I bought my TVM802B from. He sent it to me over Skype, saying that he can't send it over email, which I found odd since the size of the archive is only 1.9MB. As soon as I tried to extract the archive, Windows Defender quarantined the executable (SurfaceMount.exe) since it detected Trojan:Win32/Fuery.B!cl in it. Submitting the file to virustotal.com also comes up with multiple hits for a trojan. Did anyone else find that this file was infected? It may be a false positive, but that seems unlikely for this type of a program. I'm guessing that his email service may also have rejected the file because it detected malware, although it may just reject the sending of executables.
-
I can confirm that our anti-virus program also detected that trojan. But interestingly only after a few days.
Thus, the anti-virus program did a self-update and than transferred the executable into quarantine.
I hope that it is a false-positive ... but you can't be sure. It is also V2.32 as 2.30 has that nasty 0402 bug.
-
Yeah, I think its bug. I hope they will fix it soon.
-
yes, I found a solution, but you will not like it:
I set the cam to quick mode and reduce the speed to 20% (for SOT23 only)
This is very slow but works for me.
The problem with SOT23 are vibrations. Nozzle suction is too weak for the vibrations, so the SOT23 are dancing and flipping.
After replacing the built in vac pump with an external one, its much (!) better. Now I can run SOT23 with 100% speed
using 504 nozzles.
got any info on what vac pump you replaced it with?
thanks
Dave
-
My solution for plastic tape troubles.....
First off, new here and just received my machine the other day 802A. So far so good. Not too many major troubles. Up and running in less than a day.
But when it came time to feed a 12mm plastic tape, the feed pin would deflect the tape and I would get no feeding at all! |O
So I 3D printed a small piece of plastic, that would support the tape, so that the feed pin would not deflect it. Added some 3m tape to keep it in place.
Works like a champ!!!!
With this, I should be able to change the parameters of the 3D file, to support other plastic tapes width and height.
-
...
got any info on what vac pump you replaced it with?
thanks
Dave
I am using an oil running VAC pump which give a very deep vacuum.
A good oil filter is mandatory at the air outlet.
I connected the VAC pump to a small 1 litres air tank, and then with 10mm tubes directly to the tubes in the TVM802.
The result is good. No more problems with moving and jumping parts on the nozzle. Running all jobs with 100% speed.
One disadvantage with the ext. VAC pump: the pressure sensor does not more work with 503 nozzles (but works with other nozzles).
So missing parts (0603) are not detected and sometimes the nozzle dunks into the solder paste. If that happens, the nozzle must be
cleaned very carefully since any rest of solderpaste reduces suction significantly.
-
I just received v2.32 of the SurfaceMount software from the Aliexpress seller I bought my TVM802B from. He sent it to me over Skype, saying that he can't send it over email, which I found odd since the size of the archive is only 1.9MB. As soon as I tried to extract the archive, Windows Defender quarantined the executable (SurfaceMount.exe) since it detected Trojan:Win32/Fuery.B!cl in it. Submitting the file to virustotal.com also comes up with multiple hits for a trojan. Did anyone else find that this file was infected? It may be a false positive, but that seems unlikely for this type of a program. I'm guessing that his email service may also have rejected the file because it detected malware, although it may just reject the sending of executables.
Can anyone supply me a new version of the application (V2.32). My computer got defective and now I'am without controlling SW.
-
Try to contact with manufacturer. They send me 2.32 thru skype.
My antivirus told that 2.32 has trojan.
Anyway give me email and tomorrow i will sent 2.32 for tvm802b.
-
My email is adas@stipte.nl for anyone who can send the PC software
Thanks in advance. Ad
-
I know this was an older part of the discussion, but it is possible to disable packet coalescing on Windows 10.
There's a guide here:-
http://www.speedguide.net/articles/gaming-tweaks-5812 (http://www.speedguide.net/articles/gaming-tweaks-5812)
-
Hi there!
About a week ago we recieved QiHe TVM802b, and since that moment we are trying to set it up. A few times we assembled boards with good result. But at one moment the machine started to place components with noticeable mistake. It looks like a system mistake, because all placed components have similar offset in one direction. We changed software from 2.31 to 2.32, we changed nozzles, we tried to adjusting nozzles offset, we tried with different projects, but result is the same (you can see attached pics). Can you give me any advices? I have a real headache with this problem and don't know in what way I should think...
-
You need to adjust "vision offset"
-
You need to adjust "vision offset"
Thanks, this really helps me. Not on the first try, but after a few iterations of adjusting vision offset I've got right result with all components excluding big IC (soic28 on the white PCB). It's focusing mistake, such big chip sometimes doesn't fully recognized by the camera. Camera skips one row of legs (as it would be IC with 26 pads instead of 28). And this issue results in misplacement. I tried to set the height of this IC in the PnP file, but it doesn't help me. Maybe anyone can share own expirience with the big ICs placement?
-
camera might be out of focus. Peel off the red lok-tite and make sure camera can focus perfectly on the part.
-
camera might be out of focus. Peel off the red lok-tite and make sure camera can focus perfectly on the part.
Excuse me, not shure that I understand you correctly. Did you mean to check and align camera physically? You are right, component already looks shifted from the first shot (please see the picture), but how I can repair it?
-
Hm, no, on your pic it looks like SOIC28 does not even fit into the frame! Weird.
-
Hm, no, on your pic it looks like SOIC28 does not even fit into the frame! Weird.
Yes, sometimes camera fit soic28 right from the first shot, but after 2-3 iterations shift appears.
Only a few times the machine placed this IC right.
-
Do you have any thoughts how I can solve this problem? :scared:
-
Excuse me, not shure that I understand you correctly. Did you mean to check and align camera physically? You are right, component already looks shifted from the first shot (please see the picture), but how I can repair it?
you cannot repair it.
I use the 802 for one year now, and this kind of camera error with SO packages is totally normal for this machine.
I got used to it, a short manual correction before soldering and everything is fine.
I did a video on youtube showing this problem.
Harry
-
Folks!
I just got SW Version 2.35. It slowes down the rotation speed of the heads. This gives more precission when the component needs to be rotated.
-AH-
-
Hi
Now if there just was a rational way to get the software updates ....
Bob
-
Now if there just was a rational way to get the software updates ....
The only way is to contact them per mail and ask for an update.
-AH-
-
Just a note from a newbi: I got my TVM802C directly from Qihe via EMS/TNT (great packing!) about 3 weeks ago.
I can't really contribute much here (yet). Just one thing about the software, mine came with SucfaceMount802-EN-V2.23Beta2 (folder size 4Mb) and the SucfaceMount.exe in it is 1.2Mb. No, I did not have a typing error there.
The software didn't run at first on my brand new (on old hardware) Windows 7 and crashed for two reasons:
• The .NET installation had errors, the original MS one (3.5) also didn't work for me and they sent me a new one, that worked.
• The example CVS crashed the software because of Chinese characters in the (used for the stacks)
One question I have, is there any community produced manual? The handbook is in Chinglish, which is really hard to understand.
One more note: I read somewhere here (or think I read somewhere here) that they have problems with round SMD diodes (melf/minimelf). This might be related to the wrong nozzle type. Qihe supplies Juki type 503, , 504, 505 , 506 - ideal for melf are the types 510 and 511. Have a look at the Juki nozzle guide (http://avipre.com/products/smt&tht/juki//Juki%20-%20Nozzle%20Catalogue-Rev-C3.pdf).
-
Hi
One of the nice things about this machine is the ability to use Juki style nozzles. That gives you a lot more ability to customize.
Bob
-
...I read somewhere here (or think I read somewhere here) that they have problems with round SMD diodes (melf/minimelf). This might be related to the wrong nozzle type. Qihe supplies Juki type 503, , 504, 505 , 506 - ideal for melf are the types 510 and 511. Have a look at the Juki nozzle guide (http://avipre.com/products/smt&tht/juki//Juki%20-%20Nozzle%20Catalogue-Rev-C3.pdf).
When I was new to the TVM802 I also tried Minimelf diodes. I orderd the special Juki nozzles for round diodes. But I gave up very soon. There are many alternative flat diodes which fit on the same footprint and work much better.
-
Well, I tried some minimelf with the 803 nozzle it it seems OK. They don't fall off.
But I have two issues with vision.
1) In most cases it works, but sometimes vision does not recognize the correct PCB marking. It's should be the cross in the two images.
2) Vision seem to have a very hard time with SOT23 parts.
What is your experience and is there any way make it smooth?
-
Another one issue I met. Sometimes the machine goes in cycle while making soic8 (maybe other ICs too) full vision control. It might take up to 15-20 minutes for one IC. And it happens always suddenly. The machine can assemly a series of 10-20 equal boards without this issue and suddenly goes in cycle on the next one. My first thought was about problems with image contrast and I tried to reduce ambient light in the room. But this doesn't help.
-
Another one issue I met. Sometimes the machine goes in cycle while making soic8 (maybe other ICs too) full vision control. It might take up to 15-20 minutes for one IC. And it happens always suddenly. The machine can assemly a series of 10-20 equal boards without this issue and suddenly goes in cycle on the next one. My first thought was about problems with image contrast and I tried to reduce ambient light in the room. But this doesn't help.
I have that issue sometimes with SOT23, can take sometimes (rarely though) a few Minutes. And then place it in a 45° angle.
Other issue I have is that vision sometimes does not find the PCB marker and recognize something wrong.
And sometimes that suddenly one axis is 2-3mm off.
For the vision I know Qihe is working on it. So I wonder if my software is still the latest - and how you update the software anyway? Where you get it from?
-
Hi yellofriend,
Re: machine having vision issues with the PCB Marker - if you are using that '+' shape as your PCB Marker/Fiducial then that will explain it.
A Marker/Fiducial should conform with the industry standard [which is a circular shape with good Solder Mask relief].
There is plenty written on the subject - you should look into it.
Good luck - Pete
Another one issue I met. Sometimes the machine goes in cycle while making soic8 (maybe other ICs too) full vision control. It might take up to 15-20 minutes for one IC. And it happens always suddenly. The machine can assemly a series of 10-20 equal boards without this issue and suddenly goes in cycle on the next one. My first thought was about problems with image contrast and I tried to reduce ambient light in the room. But this doesn't help.
I have that issue sometimes with SOT23, can take sometimes (rarely though) a few Minutes. And then place it in a 45° angle.
Other issue I have is that vision sometimes does not find the PCB marker and recognize something wrong.
And sometimes that suddenly one axis is 2-3mm off.
For the vision I know Qihe is working on it. So I wonder if my software is still the latest - and how you update the software anyway? Where you get it from?
-
@thommo
Sometimes you need to work with the material you have.
But the cross to dot change is already on my list for the next PCB order.
-
Hi, I am a new owner of TVM802B, the machine works fine but I have detected several problems.
Almost all my circuits have white mask, the camera does not go well with this color it is necessary to work with darker colors.
Sometimes the components (0805 and 0603) get trapped in the nozzle although perhaps this is because it could have been messed with tin.
I use windows 10, so far I have no problems, only in step mode gives me errors in the camera.
A lot of initial work on configuring the prick offset, the factory settings were not correct and the tape was not moving well.
The camera only accepts small components, can not put QFP144?
-
The camera only accepts small components, can not put QFP144?
If the software was any good it should be able to deal with this by imaging multiple times & moving the part over the camera.
But we're talking Chinese software so maybe not.
You also need pretty high angular resolution to place QFP144s, which the hardware may or may not have, but still no excuse for not handling larger parts sensibly.
-
for vision someone should try if installing a directX filter is possible.
If it's possible, then it is no problem avoiding false recognition on some type of parts, otherwise only better lightning could resolve things like false angle of sot23.
-
The camera only accepts small components, can not put QFP144?
Largest I have is a 80-Pin TQFP and that pretty much fills the vision screen. So I wonder if the PnP vision can handle larger parts.
Additionally, the placement is not so good. Sometimes the PnP is half pin off.
What experience you have with TQFP or similar? I use the 505 nozzle for that, but somehow the 505 can't be automatically adjusted (I guess the whole size is too big for the vision). Qihe ask me to adjust manually, but I see no function for that - and their Chinglish manual is hard to read.
BTW, latest software is English802B-V2.36 - will update mine soon. SOT23 issue not fixed yet.
-
for vision someone should try if installing a directX filter is possible.
If it's possible, then it is no problem avoiding false recognition on some type of parts, otherwise only better lightning could resolve things like false angle of sot23.
SOT23s are inherently different from other parts, as the pin grid is not rectangular. With most parts all you need is to identify the perimeter of all the pins to get the position and rotation, but for a SOT23 you need to identify each pin seperately. Lighting can be an issue as it is less critical when you just want to identify the overall outline.
-
V2.36 for download.
https://www.mediafire.com/?3d2232wwvwzgrwe (https://www.mediafire.com/?3d2232wwvwzgrwe)
-
BTW, latest software is English802B-V2.36 - will update mine soon. SOT23 issue not fixed yet.
I updated to V2.36 and don't notice any difference. Is there a change log?
-
Anybody else got a virus alert from sucfaceMount.exe?
Avira gave me a TR/Confuser.56b259 warning.
Anything to worry?
PS: No, I did spell the filename correctly ;-)
-
magaq
-
...How do you get the pin to travel far enough on these larger reals? For examples, in L21 I have inductors that need twice the amount of tape movement as compared to L1 to have components lined up correctly for pickup.
go to the setup menu where you enter the XY coordinates of each feeder, there is a selection "Feed" where you can set the prick's travel distance.
-
Does anyone have the vibration feeder for ic stiks? Captures of the appearance?
-
I purchased a vibration feeder from the manufacturer, it works properly.
https://youtu.be/pvL5D8sqK1w
-
I´m installing a new computer with Win7, first works ok but slow OS. I have reinstalled the OS for cleanup and now the surfacemount program does not start, stop working in start. I tried with another computer and it happens to me the same, I tried with another compilation of win7 and it happens the same. The program works for me in old installations of win7 but if I do new installation of OS no longer works, does anyone else? Does anyone know the solution?
-
Solved after install microsoft visual c++ 2010 redistributable package :phew:
-
hi, don't worry about.
this is not real virus .
it's because of the encrypted of .exe
and antivirus mistake for virus . so don't worry about it.
-
there is a file named "802-V2.xx Software Update Log.doc"
this is the log , tell us what's new of this new software. and what we do in this new sofware.
but attention please , when you updata from some old versions , you must browse the log . bacause you should chage some data
if not change , the machine will not working normal.
so it's important to browse this log , before updata to new version. :) :) :)
the newest one is 2.37. (to be the 2.0 verions)
in fact the newest one is 3.03 but not support old machine . (the newest machine have updata the hardware.)
-
there is a file named "802-V2.xx Software Update Log.doc"
this is the log , tell us what's new of this new software. and what we do in this new sofware.
but attention please , when you updata from some old versions , you must browse the log . bacause you should chage some data
if not change , the machine will not working normal.
so it's important to browse this log , before updata to new version. :) :) :)
the newest one is 2.37. (to be the 2.0 verions)
in fact the newest one is 3.03 but not support old machine . (the newest machine have updata the hardware.)
Hi qihe_kim,
Please tell me where I can get the latest stable version of software? I'm your customer and I use TVM802B. My current software has version 2.35. Is there any subscription list or page where I can check the new software versions?
-
I purchased a vibration feeder from the manufacturer, it works properly.
https://youtu.be/pvL5D8sqK1w
Is this actual machine placement speed shown or your video had been accelerated?
I have tvm802b and even using 100% speed it doesn't move so fast.
-
you could email to our after sales ,and show your SN number
and tell him what's the software version now.
want to updata to the newest software.
email: after_sale1@qihekj.com :) :) :) :)
-
you could email to our after sales
...a webpage with a link to the software, plus a changelog would be a good, professional idea, no?
-
Does anyone have experience adjusting focus of the up camera on the TVM802?
-
Just recieved the new software version but I have following issue. The program easily doesn't work :)
I can see only this picture. Microsoft visual c++ 2010 redistributable and the latest Net framework already installed. Version 2.35 started before without any issues.
Any thoughts?)
-
hi, copy the detial in to txt file , and the shis detial to us.
after_sale1@qihekj.com
we will reply you as soon as possbile.
:)
-
I purchased a vibration feeder from the manufacturer, it works properly.
https://youtu.be/pvL5D8sqK1w
Is this actual machine placement speed shown or your video had been accelerated?
I have tvm802b and even using 100% speed it doesn't move so fast.
No, this is slow (70%) i can work more fast, something you're not doing well.
-
This video is in real speed at 100%, amount 2000 cph mixed vision and no vision.
https://youtu.be/yQW1T2a6ZD4
-
the machine in this video is MUCH faster than my 802 at 100%.
My 802 has this high speed only when moving for fiducial detection.
During normal PnP (100%) its much slower.
Any idea ?
Harry
-
There are only two settings where to regulate the speed, individually to each component in the database and the master speed that is the bar that is just above the movement cursors, this bar can be hidden sometimes if the fonts of the system have been enlarged. The speed master is adjustable from 0-120% and can be modified during work.
-
Please share screenshot where we can change master speed.
-
If you have selected enlarged fonts in windows, this option can be hidden.
-
BTW, latest software is English802B-V2.36 - will update mine soon. SOT23 issue not fixed yet.
I updated to V2.36 and don't notice any difference. Is there a change log?
I'm replying to an old post here but hopefully it may be of use/interest. I believe that this version was produced at my request, it adds a 'point with camera' option if you right-click on a specific component; essentially the camera shows you where it intends to place the part - a big confidence-booster without having to do a trial placement. I asked about a 'show me where you're going to pick up from' but unfortunately the camera can't move as far as the pick-up location. That's a pity since in my experience getting the pick-up position accurate is probably the most important single parameter.
-
Oh and my 2c on SOT23 is not to use vision. Recommended pad sizes are typically quite big and so can accommodate a relatively significant offset or rotation, and this is typically squared up during reflow. Essentially if the nozzle picks the part up accurately you're home and dry.
-
If you have selected enlarged fonts in windows, this option can be hidden.
Thanks so much !
I never have seen this slider before.
-
Oh and my 2c on SOT23 is not to use vision. Recommended pad sizes are typically quite big and so can accommodate a relatively significant offset or rotation, and this is typically squared up during reflow.
Sometimes using less technology certainly helps ;-) Easy to miss the most obvious solution. Thanks.
-
I use vision for SOT23, works well.
Have problems with vision only with SO14 or SO16 chips.
@ManuFerHi: you are my hero !!!
Your hint with the speed regulator is the best of all :-)
Tested a board with 100% settings today, wowww, what a difference.
Placement is still ok with this high speed.
I never have seen this slider before until you gave us the information about font settings.
-
I have updated today to 2.37 (thank you Kim for sending the software),
there are many new settings compared to my old 2.12 version.
Could somebody please explain what these functions are doing:
* Template Parameter. I think it has to do with the fiducial-Template. But what is this setting doing ?
* x/y offset threshold and angle offset threshold: I entered the default values, is there anything to do or should I simple go with the defaults ?
Harry
-
I have huge problem with The Vision camera lag, the FPS is so low, that it cannot position right on the make locations.
I change computer and it not computer problem.
anyone have some idea ?
Tamir
-
that happens to me with a Dell XPS1530 laptop, the lower camera has lag, even formatting again and installing another operating system, I had to change the computer to solve it.
-
Hello, I am the new proud owner of a TVM802B. Please excuse my lack of knowledge.
So far, I have made some progress with the machine but apparently the TVM802B does not successfully recognize the Mark Point on the Test board.
I have read a few posts citing proper camera calibration as critical. Would one of you TVM802B experts be willing to describe the calibration process? Frankly, the manual leaves much to be desired.
Perhaps this is important, the OffsetW/H values dither between (533.33, -480.00) and (-1, 5.23) when a recognition is made. Stepping through the program, after what appears to be a successful recognition, the head moves slightly from the Mark Point.
Two screen captures are attached.
I am running the latest V2.37 software. Also, have there been any updates to the internal controller's firmware?
Any help will be much appreciated.
Thanks,
Mike
-
Exposure setting is critical for fiducial recognition, with automatic vision tag enable, Go up step by step until you see that recognize it.
-
Thank you, i will try change the exposure, if it will not help i will change computer, and update.
By the way, which operation system are you using ?
-
Does the Comp Threshold in the System Config->Camera tab setting affect the Mark Point recognition? In general, what range of values seem to work well?
-
I was able to get the TVM802B to successfully recognize the Mark Point by changing the System Config->Camera Distance/Pixel settings to (0.0351, 0.0351) on camera 1 and (0.0159, 0.0159) on camera 2. The original settings of the machine I bought were all set to 1. The Software Update log document also shows them as set to 1.0000. I wonder why?
Since I bought the machine second hand, I question all the default settings. Is there a description other than what is described in the videos that details what each setting does and its default value? Is the Chinese documentation better that the English version. I have a lot of friends at work who can help decipher it for me.
The next issue I have encountered is unreliable pickup on feeder 1 of the left stack. I guess I need to mess with the X,Y coordinates.
Thanks,
Mike
-
Finally I have achieved that the IC pitch 0.5mm are well centered, the software has a lot of bugs and you have to learn to live with them.
For some strange reason, so that they are centered I have to put 1º of more in accuracy mode, if the component goes to 90º I have to put 91º and this way with all those that are in accuracy mode.
-
...changing the System Config->Camera Distance/Pixel settings to (0.0351, 0.0351) on camera 1 and (0.0159, 0.0159) on camera 2. The original settings of the machine I bought were all set to 1. The Software Update log document also shows them as set to 1.0000. ...
your setting is normal, same here. I don't know why all the documents show 1.000, but 1.000 does not work at all.
And there is no usable manual, no english and no chinese.
I'm using the TVM802B since 1 1/2 years now and I'm still learning.
But it is doing a good job. I have made about 2000 boards until now without any major problems.
The only big change I did was to replace the two tiny vacuum pumps with a powerful vac pump, that improved precision a lot. There is some discussion about this issue here in this thread.
-
the TVM802B does not successfully recognize the Mark Point on the Test board.
Do the upper LEDs work? For my machine they didn't work. Apparently they don't test the machines before shipping.
-
Well, I finally caved in, and ordered a TVM802A. So the anxious wait on the shipment has begun.
In the meantime I need to sort out a bench/table for it. Is a normal sized 48" x 24" bench sufficient? Or are there any issues with depth, and overhang?
-Hans
-
I'm so far pleasantly surprised by this machine. It's not without its issues of course, but given the price, I'm impressed.
Very few complaints. A couple minor things I think would be easier with the software but still quite serviceable.
At this point the only real complaint I have is when it homes the machine, and just slams into the end stops/switches at operating speed.
Since I'm sure these are questions people will have, my machine (TVM802A) is equipped with the Juki nozzles and came with software v3.03
I wonder what the latest changes have been.
Some interesting changes on mine from previous discussion too. The 12mm/16mm feed slots have been moved to the front, close to the camera.
Also, mine has a knob under the head which adjusts the brightness of the down-facing LED's.
There's also a switch on the left side near the front that I cannot figure out what it does.
-
Hi everyone! :)
I have one of this TVM802 and I have one issue... After a offset calibration the nozzle and the camera goes to the same places, but when I introduce "mark1" and "mark2" on "PCB config" only the camera goes to the right place, the nozzle has an offset similar to the PCB desviation (I mean, if TVM802 has its 0,0 point, and my PCB start in 1,2 so my "local" 0,0 is 1,2 point, the offset is 1mm ,2mm) .
You know what is the problem?
Thanks
PD: This is the latest version (2.37) shared by the support email after_sale1@qihekj.com :
http://www.filedropper.com/english802b-v237nocheck (http://www.filedropper.com/english802b-v237nocheck)
-
Hi! I have new tests:
This is a coordenates map (escaled). Each color square means a mark in the PCB: left top, right top, left down and right down. And at the right there are the real measures with and without calibration (PCB fiducial mark1 and mark2):
-Black: Altium designer coordenates
-Pad name (PadULC means Pad Up Left Calibrated, PadULW means Pad Up Left Without calibration)
-Green: Coordenates measured with camera2 with calibration active on it.
-Blue: Coordenates measured with nozzle on it.
-Red: Coordenates measured with camera2 without calibration on it.
My conclusions are:
-Real points(altium coordenates) and coordenates with calibrated vision (camera2) are pretty accurate.
-When calibration is active only shifts vision, NOT nozzle. I expected to be move both vision and nozzle, but only vision is recalculated.
I'm lost... :-[
-
Try to adjust X/Y Axis Scale (SYS CONFIG->Other tab)
For this I use top camera and vernier caliper placed instead PCB.
-
Well, this has become frustrating. Trying to set up a Win10 mini-pc to run my 802a, and the software absolutely refuses to open up and run. I just get a generic "the software has stopped running" error, and that's it.
I can run on my normal PC, but that's across the room and I wanted a dedicated system. This is a brand-new mini-pc, currently loaded with Win10 Enterprise LTSB (also tried vanilla Win10 home), completely fresh install. So no extra software in there other than Win10. I did try installing Java, as a random guess, but that had no effect.
-
Try install Microsoft Visual C++ 2010 Redistributable Package (x64)
https://www.microsoft.com/en-US/download/details.aspx?id=14632 (https://www.microsoft.com/en-US/download/details.aspx?id=14632)
-
I'll give that a try tonight. Would be nice to build up a list of requirements for the software, considering it's not really documented at all.
I'm also waiting on a reply from Qihe to see if they have a hardware/software requirements list.
The other piece of the puzzle that was required was .net framework 3.5
Thankfully that self-identified and was easy to install after making sure the Win10 was fully updated.
-
Unfortunately adding the C++ redistributable didn't help in my case. Tried all the versions that are on my main PC (which can open the software), with no success.
Also tried an earlier revision of the surface mount program. Again, no success.
Waiting on a response from Qihe as well.
I may still end up having to grab a crusty old Win7 machine after all. Which would be unfortunate as a mini-PC could fit inside the pick and place and give me a real clean result.
-Hans
-
Qihe got me sorted out. It requires the C++ 2010 Redistributable Package 32bit, not the x64
You were definitely on the right track there, definitely appreciate it.
Took me a bit to get the vision drivers all sorted out but it's running now.
Also turned off the packet coalescing as well.
-Hans
-
my machine (TVM802A) is equipped with the Juki nozzles and came with software v3.03
I wonder what the latest changes have been.
Can you share screenshot of the software please ? i have 2.36V
-
I'll see what I can do, it's not on my main network. Any particular feature you'd like to see?
I also have a rev change log that shows the changes in the 3.XX series up to this point.
As I understand it, there were some hardware changes and older version machines can't run this version software.
-
my machine (TVM802A) is equipped with the Juki nozzles and came with software v3.03
I wonder what the latest changes have been.
Can you share screenshot of the software please ? i have 2.36V
V3.03? Qihe send me the latest one just today and it's according to what they send V2.37.
The RAR pack now has a change log in docx format (I made it PDF)
-
Quick question: how you do the offset between the fiducials and the parts? Seems I forgot how to do that.
-
my machine (TVM802A) is equipped with the Juki nozzles and came with software v3.03
I wonder what the latest changes have been.
Can you share screenshot of the software please ? i have 2.36V
V3.03? Qihe send me the latest one just today and it's according to what they send V2.37.
The RAR pack now has a change log in docx format (I made it PDF)
Apparently there was a hardware change, and v2.3x would be the branch for the older hardware, while the 3.0x is for the newer hardware.
I haven't pulled mine apart yet to see what's on the inside.
-
TVM920 FOR SALE
Melbourne Australia
see comment in TVM920 Forum in the EEVblog
-
How do you do panels with that PnP?
My PCB has 4 panels, I setup the fidicials on the PCB setup, I enabled the 4 PCBs. The PnP gets parts for 4 PCBs - but places all parts in PCB #1.
What did i do wrong?
-
There are two types of coordinates, MARK and REAL, mark is the coordinates in cad design pcb software and real is the own pick and places coordinates. In mark field (PCB1, PCB2, etc.) there should always be the same value and you just have to change the real coordinates of p&p. See the picture.
-
Mhm, in 'Mark' I put the location of the 4 individual fiducials on each of the 4 PCBs, and in REAL I put the real location of on each of the 4 PCBs. Not correct?
-
Not correct, all mark fields must have same coordinates, and in the real coordinates i see some errors too.
-
Thanks for share, good information ;D
-
Not correct, all mark fields must have same coordinates, and in the real coordinates i see some errors too.
I still don't get it. Now I set all gerber mark points to PCB1 - now the PnP finds the 4 PCBs in the panel - but now the X has about 5mm offset, even though the real PCB mark points are found very accurate.
-
Show me the PCB coordinates sheet please.
-
Show me the PCB coordinates sheet please.
That's the coordinates according to Gerber.
-
No this, the PCB tab.
-
No this, the PCB tab.
Yes, so what coordinates you mean? That is it the actual placement of the parts.
-
This, the coordinates of PCB config tab.
-
I'm now working with the TVM802B for almost 2 years and have manufactured several thousands of boards.
From my experience the machine is mechanically very well done. Even after 2 years all is working well.
The big problem is the camera which does not detect all parts as good as required.
This could be improved by an additional field in the parts list: an X and Y correction-offset for CAM1.
Maybe the responsible Chinese people take this as an idea for future software releases.
Harry
-
Does anyone have a link to QiHe Vision Mounter 2.37 (or whatever the newest version is)?
I just purchased this machine and I'm wondering (and hoping) it can place SOT-45 with no issues. My design relies heavily on them. Does anyone have experience with SOT-45 using the TVM802?
Thanks!
-
to be honest, I never have seen SOT-45 case.
Can you please give us the part number of this component.
As a general rule of thumb, the TVM802 can please components in 0603 size quite well.
Its not ideal for smaller components, in which case you will need to correct it under the microscope before soldering.
-
to be honest, I never have seen SOT-45 case.
Can you please give us the part number of this component.
As a general rule of thumb, the TVM802 can please components in 0603 size quite well.
Its not ideal for smaller components, in which case you will need to correct it under the microscope before soldering.
Apologies, I meant SC-74/SOT-457
https://www.digikey.com/product-detail/en/nexperia-usa-inc/BCM857DS115/1727-4884-1-ND/2531388 (https://www.digikey.com/product-detail/en/nexperia-usa-inc/BCM857DS115/1727-4884-1-ND/2531388)
-
I have put many of this type, generally small components go well, the problem with vision is when the components are large.
Even so it is essential to make a visual review before reflow.
-
Sharing this version (v3.06) of the software that was provided with my 802A
fyi, seems most versions of this software are detected as malware, use at own risk... I cannot verify it as I have not received my machine yet. but since this is newer than other versions of the software that I have seen, I wanted to share it.
https://1drv.ms/u/s!AjSyWl01MM23jogmKNNmTgaelR9pcQ
-
Hey all,
I just got mine yesterday and calibrated it pretty easily, actually. Thought it would be a nightmare.
One question: With the test alignment board, I'm trying to use the bottom side with multiple fiducials and split between PCB1 and PCB2 (in essence, get the machine to place the same thing on "two" mapped PCBs). I map out the fiducial coordinates for PCB1 Mark 1 & 2 and PCB2 Mark 1 & 2. When it goes to "PCB 2" it aligns the fiducials correctly, however it starts placing components on top of PCB1 (where it already placed components).
Any ideas?
Thanks!
-
https://www.facebook.com/groups/213033402766980/ (https://www.facebook.com/groups/213033402766980/)
We've also made a Facebook group for owners of TVM802B. Discussions, questions, sharing scripts for pick and place export (Diptrace, etc.).
-
I have been playing around with this TVM802B machine and made several successful builds with it. However, I am still not very satisfied with it. As a couple of other people mentioned here, the vacuum pump system is not really well designed. It puts my components in the trash area on my second nozzle about 50% of the time, while they are perfectly in place. If I disable the use of the pressure sensor, I can place components just fine, so I conclude that the vacuum is actually working
I seem to have an older version of the TVM802B, I can't run the 3.x software that a couple of guys shared, only the 2.xx versions. Apparently the guys in China made a new FW for the machine, and I guess a few new HW parts too! There are a lot of features in the 3.x that I don’t have access to, like the vacuum configuration for each nozzle for example, or aiming at components when setting up the IC pickup place for trays or strip holders.
A lot of people have big plans, making a new software, writing a new FW as well. That's all very good, but humanity will be on Mars before that project is complete! ;)
At this point I have not had my machine taken apart yet, but I am getting there. I read some other guy said it was a mechanical pressure sensor that was either high or low. Is there anyway to "configure" this little HW thing, or maybe replace it with something a little more refined? Maybe somehow be able to adjust at what pressure it thinks the component is not there? The new FW/SW has an upper and lower threshold, that would be nice.
-
In my TVM802 machine the pressure sensors like this https://www.ebay.com/itm/Sensore-Pressione-MD-PS002-Vacuum-Sensor-Absolute-Pressure-Sensor-For-Arduino/122884435359?epid=1031456898&hash=item1c9c7bad9f:g:4AkAAOSwYwJaRM3M (https://www.ebay.com/itm/Sensore-Pressione-MD-PS002-Vacuum-Sensor-Absolute-Pressure-Sensor-For-Arduino/122884435359?epid=1031456898&hash=item1c9c7bad9f:g:4AkAAOSwYwJaRM3M)
+op Amp + Comparator. I think adjusting comparator level you can change pressure sensitivity.
-
I also started taking the machine apart today, but didn't get as far as you to dissect the actual assembly. But yeah, that's what it must look like underneath the board. Did yours break down after looking at enough pressure sensors I concluded that it is indeed the one you posted the link to.
I followed the hoses back to the pump and it seems like one of the pumps actually isn't performing very well, they are both Parker D861-22-01.
I took a few pictures of the board, trying to figure out what is going on, but maybe you have better pictures now that you have the board off the mount? I only measured a few voltages as well and saw that when there a component is on the nozzle pin 7 on both the OP amps go from ~0V to around 1.4V-2V using nozzle 503, and U1 seems to be for Nozzle 1.
The PC SW does not show anything else than if there is pressure or not, but since the new version of the SW has thresholds for the nozzle pickup, I doubt that the OP amps are actually working as a comparator, but instead measuring the sensor, and amplifying the signal to be somewhere between 0-5V so that micro can deal with it? I could be wrong, would need to measure all the pins
But whether it is a comparator sending a digital signal back or the microcontroller is actually doing an ADC measurement or the sensor reading, I guess I can get around this by adding a potentiometer to one of the resistors so I can control the voltage leaving the board.
If you have some good pictures of the little board on both sides, please do post them
-
But whether it is a comparator sending a digital signal back or the microcontroller is actually doing an ADC measurement or the sensor reading, I guess I can get around this by adding a potentiometer to one of the resistors so I can control the voltage leaving the board.
On 802 the comparator sends out a binary signal. Adding POT is possible, but then you need to keep in mind that for different components you might require different thresholds. Say, ceramic caps are usually more leaky, and a pressure level of a well seating cap might the same as that of a mispicked resistor.
-
Why not use openPnp, there exist a driver for the standard firmware , however i don`t know if it works with latest one.
Further there are some people that have converted it using arduino mega as alternative controller using a protoshield for plugging in the existing connectors and
remaining necessary circuits.
-
Do you use this setup?
-
No, i had for some time a cmht28 that i have converted to openpnp adding top and bottom camera.
Had written the openpnp driver at that time you'r reverse engeenered the protocol, and helped some people to convert the tvm802 to openpnp using
6axis grbl firmware. OpenPnp is slower, assembly is more accurate and further it is possible to mount nozzle changer (approx 200$).
-
Hi.
Does anyone have experience placing TQFP100 and TQFP144 chips with this machine?
I also have doubts about the limitations in relation to the height of the components, such as aluminum capacitors. For example, if we can place aluminum capacitors with a height of 12 mm.
I recently bought this machine, but after seeing the Neoden4, I think I will replace the TVM802 with a Neoden4, it is more expensive but I can safely place TQFP100/144, although there are limits with the height of the components that can not exceed 10 mm .
regards
-
No, i had for some time a cmht28 that i have converted to openpnp adding top and bottom camera.
Had written the openpnp driver at that time you'r reverse engeenered the protocol, and helped some people to convert the tvm802 to openpnp using
6axis grbl firmware. OpenPnp is slower, assembly is more accurate and further it is possible to mount nozzle changer (approx 200$).
I want to try openpnp with TVM802B although it gives me a bit of fear, afraid of making mistakes and having the heads or prick hit with something.
Is it possible to control the prick with openpnp?
-
Hi.
Does anyone have experience placing TQFP100 and TQFP144 chips with this machine?
I also have doubts about the limitations in relation to the height of the components, such as aluminum capacitors. For example, if we can place aluminum capacitors with a height of 12 mm.
I recently bought this machine, but after seeing the Neoden4, I think I will replace the TVM802 with a Neoden4, it is more expensive but I can safely place TQFP100/144, although there are limits with the height of the components that can not exceed 10 mm .
regards
Hola Luis.
QTF100 yes it fits in the vision of the camera, but TQFP144 no, I at least I have not got it you can only place it without vision. Anyway, even if you use vision, my experience is that all CIs with 0.5mm pitch must visually inspect them and correct their position before the reflow. Sometimes the machine hits but others do not, the small integrated ones place them much better than the big ones.
That is one of the reasons why I want to try openPNP, and a bug that has the original software, when it places two components simultaneously with both nozzles, the first nozzle does not stop sucking until the second does so it stays stuck in the nozzle until the second nozzle places its component. It only happens sometimes but this bug is very annoying.
-
.
Thanks Manu.
Even before buying the machine, I was warned by the chinese that the TVM802 did not accurately place these chips with such a fine pitch.
My idea was to position with the machine, and then review and adjust with the microscope.
But shortly after I bought the TVM802, I saw the Neoden4, and I liked it a lot. My idea is to sell the TVM802 to a customer of mine, and buy the Neoden4.
The TVM802 I still have it in the unopened box, I think that as it is, it's going to go for my client, brand new. I do not know if it's worth it, take it out, install it and try it, because if it does not position well the TQFP144, I prefer the Neoden4.
Do you know what limitations the TVM802 has, in terms of the height of the components? This is with respect to the electrolytic capacitors, I have some of 13mm height.
.
-
I have been using the machine for 9 months, I have done a lot of assembly, I think I know it well enough.
If the component you will use in feeder left and back, max height will be about 4mm, if you are going to use it on a tray you can reach 6mm forcing a lot, I have tried to put 10mm but can not even get it out of the tape anymore from 6mm you can touch with other components of the PCB.
The software allows you to work at any height, but you encounter physical limitations.
-
About the height of the component, in Neoden they tell me that they will make me some customized nozzles for the Neoden4, to place components with a height of up to 10 mm. I think they are shorter nozzles.
I recently asked about the nozzles for the TVM802, but these seem to be all standard, do not customize them or make shorter models to solve this limitation.
I do not understand why these machines have these limitations, so that the Z axis could move more, let's say about 2cm, it would be enough to be able to easily place high components such as electrolytic capacitors and some power inductors.
-
you could place the PCB lower, there is enough space to the body of the machine.
Of course, this only makes sense if you have a lot of big caps or inductors since you will
also need a big nozzle and usually a better VAC pump to hold heavy parts.
I'm using an external VAC pump for a year now, and it works MUCH better then the tiny build in pumps,
even for very small components.
-
Hey Harry,
It was your videos on Youtube, your reviews on this blog that caused me to put money into this machine too. Your Eagle to TVM converter got me going much faster than I would have been otherwise. Thanks!
I have the pumps hanging out of my machine at the moment, wondering if adding another small pump would make me happy or if I should follow in your footsteps once again. I have a couple of other vacuum pumps that I could put in serial with the existing dual chamber pumps. But I have a feeling you might already have been down this road and concluded that an external pump is the way to go?
One feature I guess you can't really use then, is the pressure? Or are you controlling the vacuum in the external system to be in the "correct" range?
I spend some time trying to ohm through the little board that sits on top of the nozzles and measures the vacuum. The attached schematic is what I found. It might be wrong, but at least this makes sense.
The vacuum itself is setup up as a wheatstone bridge. One side of the bridge is fed into the opamp (blue) which just works as a buffer as the amplification is close to 1, then in the green box the signal gets amplified/compared, the output driver will drive the pulled up 3.3V signal to ground once the opamp signal is high enough. At least something like that.
I don't know exactly how the sensor works, if one side of the bridge is ambient pressure serving as a voltage reference through one opamp, and the other the sensed pressure? But I am guessing that the 10k resistors in the diagram can be made smaller, to amplify more, giving a signal earlier with the weaker pump? Will try this over the weekend
I guess the feeders mechanism could be better, but as a whole the machine works pretty good. I have serious issues with the SW/FW though, it could be so much better, that's really a shame. Therefore I am leaning a bit towards trying to do a conversion to the OpenPNP now after someone commented about the Smoothie boards and swapping out the control boards.
This guy did the hard work it seems, has pictures and wiring of most things. https://github.com/johngrabner/tvm802B_2_smoothie_and_OpenPnP
Reverse engineering the FW as some people started is of course possible, but making a couple of adapter boards with the correct connectors which connected easily to the existing TVM802 connectors and to the smoothie-boards is probably much easier.
Watching a few videos from OpenPNP, I especially like the option to teach the machine about strips with components, using the camera to find the exact center of very small components, that looked very neat. It is very cumbersome to setup to the machine when having multiple small strips of components.
-
Hey Harry,
It was your videos on Youtube, your reviews on this blog that caused me to put money into this machine too. Your Eagle to TVM converter got me going much faster than I would have been otherwise. Thanks!
I have the pumps hanging out of my machine at the moment, wondering if adding another small pump would make me happy or if I should follow in your footsteps once again. I have a couple of other vacuum pumps that I could put in serial with the existing dual chamber pumps. But I have a feeling you might already have been down this road and concluded that an external pump is the way to go?
One feature I guess you can't really use then, is the pressure? Or are you controlling the vacuum in the external system to be in the "correct" range?
I spend some time trying to ohm through the little board that sits on top of the nozzles and measures the vacuum. The attached schematic is what I found. It might be wrong, but at least this makes sense.
The vacuum itself is setup up as a wheatstone bridge. One side of the bridge is fed into the opamp (blue) which just works as a buffer as the amplification is close to 1, then in the green box the signal gets amplified/compared, the output driver will drive the pulled up 3.3V signal to ground once the opamp signal is high enough. At least something like that.
I don't know exactly how the sensor works, if one side of the bridge is ambient pressure serving as a voltage reference through one opamp, and the other the sensed pressure? But I am guessing that the 10k resistors in the diagram can be made smaller, to amplify more, giving a signal earlier with the weaker pump? Will try this over the weekend
I guess the feeders mechanism could be better, but as a whole the machine works pretty good. I have serious issues with the SW/FW though, it could be so much better, that's really a shame. Therefore I am leaning a bit towards trying to do a conversion to the OpenPNP now after someone commented about the Smoothie boards and swapping out the control boards.
This guy did the hard work it seems, has pictures and wiring of most things. https://github.com/johngrabner/tvm802B_2_smoothie_and_OpenPnP
Reverse engineering the FW as some people started is of course possible, but making a couple of adapter boards with the correct connectors which connected easily to the existing TVM802 connectors and to the smoothie-boards is probably much easier.
Watching a few videos from OpenPNP, I especially like the option to teach the machine about strips with components, using the camera to find the exact center of very small components, that looked very neat. It is very cumbersome to setup to the machine when having multiple small strips of components.
There was a member here and on the openpnp group that made a driver that sat between openpnp and the tvm. That way you didn't need to change out the control system. I don't know if he ended up making it opensource. Glen English I think was his handle.
Openpnp works pretty good too and depending how proficiant you are with Java can be customised to your needs unlike the provided software.
-
The other weekend I made 4 boards with about 192 components and I am guessing there were something like 250 discarded 0603 components, mostly capacitors and inductors, but only a few resistors. I kept trying to adjust the pickup point for the stacks, adjusting the prick correction. I eventually started running only with nozzle 1 as the other nozzle would discard almost any components, which were sitting on the nozzle just fine. I figured it was time to spend a couple of hours to fix some of these items:
1. Nozzle 2 pickup place
I also messed around with the nozzle/camera calibration built in functions. Big surprise: it doesn't work very well. Tried many different 503 nozzles, mostly the same offset calculated. When the pickup place for nozzle 1 is perfect, nozzle 2 always picked up the component too far to the right. But since it was mostly a problem on the left stack, the back stack was a bit better. So I tried adjusting the "calibrated" values manually, and what do you know, it started to work much better!
2. Vacuum pressure signal
The nozzle was picking up much more accurate now, but still no vacuum pressure signal. I'm still not sure if the 2nd pump is actually worse, or it is the vacuum sensor that has a problem, but.... I already mapped the vacuum sensor board components and so I tried to change the R10 (earlier attached schematic) from 10k to 8k, thinking that this was the way to go. Now I measured even less voltage on the transistor basis pin, so went with a 13k instead and suddenly it started working a lot better.
3. Vision aligned components off a bit
With the learning from the nozzle 2, started playing around with the camera aligned components. I often had the problem that bigger TQFP parts ended up falling in between the solder pasted pads. Turned out to be the same "calibration" that was the problem. Corrected the location of the nozzle 1 compared to the camera and now it stated putting components so very accurate.
Later that evening I ran 8 of the same boards and had a total of only 6 discarded components. There are more components lost due to problems with the reel tape not being pulled off properly all the time, but it's a big step forward
When I run the machine at very high speed, the thick aluminum plate where the nozzles, camera and prick sit, will randomly hit the screws on the back stack, around B10-B16, when not using the camera correction. I guess the acceleration is so fast and abrupt that the head actually twists a bit and when it hits, the stepper engine loses a few steps. I guess these are open loop-steppers? Took the metal shell of the head piece, to reduce the weight, have not seen it since. May have bent the prick a bit because whenever this happened, the machine got stuck when trying to pick up the next component, but the location was now off.
There is a play in the linear bearing on the right side, one end of it can move somewhat. Maybe that causes the head to hit those screw now and then.
I have seen it try to go all the way too the left after picking up a component. The alarm sounds and the machine stops. Sometimes it totally misplaces a component, but the next ones are ok. That's kind of annoying because I have to visually inspect every single component before the oven. The USB connectors on the machine are very crappy, causes the cameras to fail quite often.
On the positive side, I don't think I have even seen the PC SW crash. V3.x has some nice features that I would like to use, they could even be ported back to V2.x, but China Jim seems to have stopped responding - the 1 year warranty period is over. Stuck at v2.37
-
Hi,
I plan to buy one next week but I read only about the problems with this machine.
My question:
Is it worth buying? Should I go ahead or start looking for another one?
Thank you
Peter
-
Hi,
I plan to buy one next week but I read only about the problems with this machine.
My question:
Is it worth buying? Should I go ahead or start looking for another one?
Thank you
Peter
The truth is, these are bottom-dollar pick and place machines, so you need to have appropriate expectations. You also need to give yourself time to learn the personality of the machine and work out any issues that affect your needs. Different boards, different parts, different designs, they all have varying needs. So it all really depends on what YOU need the machine to do.
But even full-on industrial machines that cost hundreds of thousands of dollars have issues. Any pick and place machine is going to need engineering time to program and tune, repair, and debug. Doesn't matter the price, source, or age of the machine. It's the price of admission to surface mount manufacturing. If you ever worked on a Fuji CP series machine you'd know exactly how incredibly complex and maintenance intensive a machine can get. Compared to the old CP's, which were industry standard workhorses for a decade or better, this itty bitty Qihe on my desk is a dream to keep going.
I've done a few hundred boards on mine now, mostly just rows and rows of good sized passives. Some QFP32's and such. Probably 15,000 placements so far I'd guess. Working within the limits of the machine I've been pretty happy with it. It's not without its quirks but for my needs it's a very good fit.
The only area that's caused me any actual issues are in the prick needle. It was getting hung up just slightly during motion, causing parts to jump from the pocket, particularly when advancing two components at a time. I took a file and tapered the point, and the problem went away instantly. Now the only real lingering issue I have right now is that the opto sensor which detects the needle being extended sometimes doesn't work. The software compensates but it does dramatically slow down the unit. I don't know if it's overheating or getting dirty just yet. I haven't been to the point yet where I am ready to pull the electronics off the bottom of the head.
-
The truth is, these are bottom-dollar pick and place machines, so you need to have appropriate expectations.
I actually have. I need to place some tqfp64 parts and many passives. For me the speed is not important at all. I have to prototype boards I design and I hope that this machine will help me with it. Probably even the slowest machine will be faster than me. My eyes are not as good as they used to.
-
I did not have much success with TFQP. The machine will routinely place it at an angle. They are very easy to bridge and hard to clean up the bridges.
-
I did not have much success with TFQP. The machine will routinely place it at an angle. They are very easy to bridge and hard to clean up the bridges.
That is actually the main reason of buying it.
-
Hi. Are you generally happy with the machine or buying you would consider as a mistake?
-
I would not call it a mistake... I learned a lot from it. But I stopped using it after I bought TVM920.
-
920 does not fit my budget :( . Bus as I understand I should look for the alternative.
-
I'm pretty sure all alternatives in 4K range will be about the same. If you look at them, they all use identical or similar hardware, even their feeders are the same. The s/w could be better or worse. Then again, your experience might be different. There is a guy here who's user name is a digikey's part number, and he seem's to be satisfied with his machine, which is a different but similar to 802.
-
Thank you.
To buy, or not to buy, that is the question :)
-
For my own use, I've been pretty happy with it. Zero regrets, and it will cover my needs nicely until I'm ready to move up to a larger machine.
Eventually I will outgrow it, and then I'll need a 4+ head machine with better/more feeders.
Software isn't the most intuitive, and I do need to look into a better way at bringing in placement files from diptrace and kicad. Any issues I've had there are mainly just my own lazy self not doing any work to figure out how.
There have been times I wished it had a bit more component range, but having a 3D printer and making custom parts trays keeps things workable for me.
In my case it hasn't been the total number of feeders, it's been the limited height in the 12mm and 16mm wide slots. They just don't deal well with parts more than a couple mm in height.
However, once I put a taper into that the advance pin, nearly all my problems went away. Feeds much more reliably now. Though I do still need to track down an issue with the opto sensor that detects the advance pin. I think it overheats eventually and starts to always think the pin is retracted.
-
However, once I put a taper into that the advance pin, nearly all my problems went away. Feeds much more reliably now. Though I do still need to track down an issue with the opto sensor that detects the advance pin. I think it overheats eventually and starts to always think the pin is retracted.
Thank you. Could you explain it better. Where and what is this advance pin?
-
...I spend some time trying to ohm through the little board that sits on top of the nozzles and measures the vacuum. The attached schematic is what I found. It might be wrong, but at least this makes sense.
The vacuum itself is setup up as a wheatstone bridge....
thanks for sharing the schematic.
My external VAC pump is an oiled vac pump and does a very deep vacuum. It is connected to a pressure-equalizing canister. Then it is spread into two tubes and connected to the original tubes in the TVM820. The 820's vac pumps are unpowered now which makes the machine a lot quieter.
The pressure sensor works fine with the 504 nozzle and larger. But the 503 (which I use for 0603 parts) is detected as OK even if no part is on the nozzle. So I need to reduce the sensivity of the sensor and I will try it using your schematics.
modifying the amplification of the first amp (blue box) should do the job.
-
Well, broke the advance pin on mine, and just bent the spare. Sent an e-mail to Qihe about ordering a replacement, but figured I'd check with you guys to see if anybody else has had this issue and may have identified a stronger/better replacement from a different machine.
-Hans
-
I would not try to make the pin stronger. If fact I would make it weaker. If you broke it means you messed up something, or, more likely, the s/w messed up on its own :) If you run into the same situation again with a stronger pin, something else will break, and could be worse.
-
talk to qihe via skype, that's the fastest way to reach them
-
Hello guys,
I'm negotiating to purchase a second-hand TVM802A, but I don't understand what's the differences between TM802A and TVM802B.
Which are the formats that this machine can't manage?
I'm very newbe of smd world and I need to buy all necessary to can assembling the boards correctly, and then, I ask you which solder paste and reflow oven do you use to have good quality job.
Many thansk at all.
Regards.
Stefano
-
1. The difference is in the number of feeders.
2. If you buy the machine check if it is 3.x version - this software is taking the mnt files directly so no converters are needed
3. Solder paste - Pb free if you want to sell it. ROHs compliant.
4. Conveyor type is the best.
I have the machine for couple of days and I am quite happy. Everything works as expected.
-
1. The difference is in the number of feeders.
2. If you buy the machine check if it is 3.x version - this software is taking the mnt files directly so no converters are needed
3. Solder paste - Pb free if you want to sell it. ROHs compliant.
4. Conveyor type is the best.
I have the machine for couple of days and I am quite happy. Everything works as expected.
What version do you have, TVM802A or B?
Ordered this year? Did you received also a extra box with parts or just bare machine?
Thanks,
TJ.
-
Does anybody have experience with the TVM802B-X? What are the main differences? I couldn't find much info on it. I understand it was released quite recently.
-
eh..... difference ver the website introduction pictures
BX type has Built-in industrial computer and high voltage motor drive system inside
More stable than the old type？？
link
https://qihesmt.en.alibaba.com/product/60757615921-804211664/Products_TVM802A_X_desktop_smt_pick_and_place_machine_bga.html?spm=a2700.icbuShop.prewdfa4cf.40.5b047bbbRfgUV6
-
Maybe what he said was the needle.It is next to the down camera.
-
My TVM802B is constantly generating "X-Axis Into Positive Limit" alarms. There does not appear to be any problem to cause these alarms. Is there a way to adjust the limit?
Also, my TVM802B was purchased in August 2015. Is it possible to upgrade the firmware?
-
My TVM802B is constantly generating "X-Axis Into Positive Limit" alarms. There does not appear to be any problem to cause these alarms. Is there a way to adjust the limit?
Also, my TVM802B was purchased in August 2015. Is it possible to upgrade the firmware?
Maybe the sensor of X-Axis or the mainboard is broken.And it is difficult to upgrade the firmware.
-
"Axis into positive limit" means that the microswitch has been closed by the head moving all the way to its limit (X or Y axis). When you do a 'home' operation the machine does this for both axes and then presumably backs away a little, until the switches open again. If the machine sees that the switch is closed you'll get this warning; I think you'll find that it tends only to happen when the head is in the home position; if so then there's nothing to worry about. If it happens when the head in any other position then there's a problem (stuck microswitch).
-
I can confirm this. It happens when the head is already in the home position at power up and can be ignored.
-
After one year of using this machine, one of the pumps stopped working. Sticker on the pump tells me that it's Parker / Hargraves PN: D861 - 22 - 01. I tried to find it on the official web site using this PN, but nothing was matched with this request. Has anybody meet the same problem? Have you tried to find any replacements? Need to say, that I don't want to use external pump at this moment.
-
Hye guys the TVM802 VIDEO looks impressive
LINK
https://www.youtube.com/watch?v=GbfL4k15UJk (https://www.youtube.com/watch?v=GbfL4k15UJk)
-
After one year of using this machine, one of the pumps stopped working. Sticker on the pump tells me that it's Parker / Hargraves PN: D861 - 22 - 01. I tried to find it on the official web site using this PN, but nothing was matched with this request. Has anybody meet the same problem? Have you tried to find any replacements? Need to say, that I don't want to use external pump at this moment.
You can connect our aftersale to buy the pump.
E-mail:after_sale1@qihekj.com
-
After one year of using this machine, one of the pumps stopped working. Sticker on the pump tells me that it's Parker / Hargraves PN: D861 - 22 - 01. I tried to find it on the official web site using this PN, but nothing was matched with this request. Has anybody meet the same problem? Have you tried to find any replacements? Need to say, that I don't want to use external pump at this moment.
You can connect our aftersale to buy the pump.
E-mail:after_sale1@qihekj.com
Thanks for your response. I've written to your e-mail just.
-
After one year of using this machine, one of the pumps stopped working. Sticker on the pump tells me that it's Parker / Hargraves PN: D861 - 22 - 01. I tried to find it on the official web site using this PN, but nothing was matched with this request. Has anybody meet the same problem? Have you tried to find any replacements? Need to say, that I don't want to use external pump at this moment.
You can connect our aftersale to buy the pump.
E-mail:after_sale1@qihekj.com
Thanks for your response. I've written to your e-mail just.
You are welcome.My colleague has already replied to you.Hope your question will be solved.
-
talk to qihe via skype, that's the fastest way to reach them
Any question about TVM802A/B can connect the aftersale.
Provide free technical support.
Skype ID:huang
E-maill:after_sale1@qihekj.com
-
What is the TVM802D? What does the D model mean?
-
What is the TVM802D? What does the D model mean?
The TVM802D is the same as the TVM802B.
-
What is the TVM802D? What does the D model mean?
The TVM802D is the same as the TVM802B.
Great, but what's the difference between. If I want to buy one, why I should choose a B or a D.
-
The TVM802D is the same as the TVM802B.
Why do you have two different names for the same machine?
-
The TVM802D is the same as the TVM802B.
Why do you have two different names for the same machine?
In the beginning,the software of TVM802B language is Chinese,and the software of TVM802D language is English.
Later, in order to unify the model, TVM802D was changed to TVM802B.So the model of machine is same.
-
What is the TVM802D? What does the D model mean?
The TVM802D is the same as the TVM802B.
Great, but what's the difference between. If I want to buy one, why I should choose a B or a D.
You can only buy the TVM802B now,because we unified the model.And the TVM802A/B had made a lot improvements and upgrades.Also now we have the new model TVM802AX/BX.You can learn about them through this link：https://www.alibaba.com/product-detail/New-style-automatic-led-assembly-pcb_60694587058.html?spm=a2700.7724838.2017115.1.8f3744a4UmFq1o&s=p
-
Ok, thank you for for your reply. The new models have built in computer, is that all?
-
Ok, thank you for for your reply. The new models have built in computer, is that all?
Yeah.
1.Built in computer.
2.High voltage drive.
-
From research and development to the present, we are constantly improving the TVM802, including hardware and software modifications and upgrades.Our software version is from 1.XX to the current 3.XX, but if you are using 2.XX version software, you can upgrade to the latest 2.39 version, but can not use 3.XX.Because the hardware of 2.XX machine is not the same as 3.XX.In the latest software, we have perfected some defects and added some features.If you want to get the latest software and know more,you can connect to our aftersale.The e-mail address :after_sale1@qihekj.com.
Also you can visit our website to know more about the latest machine:www.qhsmt.com
-
These are V2.39 Update log.If you want to know more,you can connect this e-mail:after_sale1@qihekj.com
1、Modify the BUG that the data “Angle” which is in the list “Sub Pcb Reference Point Pos” didn’t have any effect;
2、Add one new algorithm for the recognition function of mark point ：“Image Template”.And optimize the interface.
(1)、The “Size” in the “TemplateParam” corresponds to the size of the Red Box in the Down Vision.And select the appropriate image as the template by adjusting the red box size.
（2）、Click the button “Get” to get the image template of Mark.
（3）、Click the button “Test”。The Bule Box is the image which matches the image template. Below the vision screen,“OffsetX” and“OffsetY”reflect the offset of the matched template.
(4)、The data of “ImageMark” are saved to the CSV file.
3、Optimize the interface:
(1)Move the interface of “Run”“Manual”“Auto”;
（2）Add a quick menu of “PcbArray”;
(3) Add the button which could switch the Vision;
(4) Modify the method of change the speed,now use the button to add and subtract the speed.
4、Optimize the function of “PcbArray”, and the components in the list of each board are independent and can be changed manually.
About V3.10
1、Modify the BUG about “Aimed With Camera”;
2、Add the function that the Z Axis could return to the origin home,and check all the time if the Z Axis is at the origin home.
-
I am very honored to communicate everything about TVM802 with everyone here.This is a public platform that can ask questions and leave suggestions.At present, the hardware and software of TVM802 are continuously updated and upgraded, so as to better meet the needs of customers.Thank you very much for your valuable suggestions.
-
Hi
I am going to buy a TVM802B and I am now talking to someone from
https://www.zjyingxing.com/ (https://www.zjyingxing.com/) on skype.
They are now asking for a deposit via paypal which I find a bit strange.
Is this legit? Anyone bought it there?
Thanks a lot!
-
Hi
I am going to buy a TVM802B and I am now talking to someone from
https://www.zjyingxing.com/ (https://www.zjyingxing.com/) on skype.
They are now asking for a deposit via paypal which I find a bit strange.
Is this legit? Anyone bought it there?
Thanks a lot!
Hi,yingxing is one of our agents.You can buy the machine from them,and the machie is produced by our company.We will also provide follow-up services.
-
Thank you for the confirmation, looking forward to work with the machine
-
Hi,
I am having a couple of issues setting up my TVM802B. It is a newer model with firmware 3.09.
- When turning it on there is a very loud hum, even when the safety switch is in stop mode. It's not the pump I think, because I can turn that on and off in the software. Is this normal? Sounds like transformer hum. This is on 220V.
- USB recognized one camera as STK1160. There are no further cameras unlike stated in the user manual.
The driver supplied is for a STK1150, so windows 7 would not install this one. I installed a driver for STK1160. Both cameras work with the viewer tool included in the driver, but not with the PnP software.
In the software I can only see a red cross on black background
thank you
-
Hi ubbut.
In the setting/camera menu you have a tab that you can choose the serial number cameras, unfold the menu and choose the correct serial.
-
Hi ubbut.
In the setting/camera menu you have a tab that you can choose the serial number cameras, unfold the menu and choose the correct serial.
Thanks. I am unable to find this setting somehow? Where is it?
-
Sysconfig / Camera
-
Looks different in my version of the program :(
-
Looks different in my version of the program :(
Hi ubbut,
There is no serial number cameras in the V3.09 software,you just need to install the camera driver correctly,and then the camera will work.
Please check the e-mail I replied to you.
And the sound is normal.It cames from positive pump,and it is on 220V.
-
My TVM802B is constantly generating "X-Axis Into Positive Limit" alarms. The unit is now almost unusable.
Does anyone have a service manual or schematics so I can troubleshoot the problem?
What is the part number of the photo interrupter? I suspect the problem is related to noise, but it will be helpful to know which part is used.
If it is a noise problem, I'd like to desensitize the detection circuit so it will be more immune to noise spikes.
Thanks,
Mike
-
Is there a bug in the new v2.39 software? After using the "Prick MoveTo" function a few times the prick stops functioning and the unit needs to be power cycled to fix it.
Anyone else seen this problem? BTW, the new software is great. It adds a lot of new features.
Thanks,
Mike
-
Hi,
on my tvm802b the center of Camera1 is not in the center of the lens of this camera - it shifted a few mm up. Is the way to correct it?
I still cannot to tune the visial control well ...
-
Is there a bug in the new v2.39 software? After using the "Prick MoveTo" function a few times the prick stops functioning and the unit needs to be power cycled to fix it.
Anyone else seen this problem? BTW, the new software is great. It adds a lot of new features.
Thanks,
Mike
Please check the e-mail I replied to you.And we will check the software,any updates, we will notify.
-
Hi,
on my tvm802b the center of Camera1 is not in the center of the lens of this camera - it shifted a few mm up. Is the way to correct it?
I still cannot to tune the visial control well ...
Hi pl116,what's the Serial Number of you machine?And you can connect to our aftersale.He will help you solve the problem.His e-mail address is after_sale1@qihekj.com.
-
Hi,
on my tvm802b the center of Camera1 is not in the center of the lens of this camera - it shifted a few mm up. Is the way to correct it?
I still cannot to tune the visial control well ...
Hi pl116,what's the Serial Number of you machine?And you can connect to our aftersale.He will help you solve the problem.His e-mail address is after_sale1@qihekj.com.
SN: 76-6A-46-F5-BF-0B
on the pictures below - current position , and where is real optical center ... I think to fix it - I need the "offset of optical center" option or something like this in the camera driver.
-
Hi,
on my tvm802b the center of Camera1 is not in the center of the lens of this camera - it shifted a few mm up. Is the way to correct it?
I still cannot to tune the visial control well ...
Hi pl116,what's the Serial Number of you machine?And you can connect to our aftersale.He will help you solve the problem.His e-mail address is after_sale1@qihekj.com.
SN: 76-6A-46-F5-BF-0B
on the pictures below - current position , and where is real optical center ... I think to fix it - I need the "offset of optical center" option or something like this in the camera driver.
In the Sys Config,Choose the Nozzle,you can find the Nozzle2 Offset.Click Test.Follow the steps it prompts.
-
Does anyone have a backup of the TVM802B system controller flash image? The STM32F107 on my board got damaged so I installed a new part and can connect to it via ST-Link. I evidently screwed up the backup image I extracted from the defective part.
-
Never mind. I ended up buying a replacement Mainboard from QiHe.
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
Hi pl116,you can adjust the value of Camera Position to slove it.It is just some offset.
-
Hi,
I am having a couple of issues setting up my TVM802B. It is a newer model with firmware 3.09.
- When turning it on there is a very loud hum, even when the safety switch is in stop mode. It's not the pump I think, because I can turn that on and off in the software. Is this normal? Sounds like transformer hum. This is on 220V.
- USB recognized one camera as STK1160. There are no further cameras unlike stated in the user manual.
The driver supplied is for a STK1150, so windows 7 would not install this one. I installed a driver for STK1160. Both cameras work with the viewer tool included in the driver, but not with the PnP software.
In the software I can only see a red cross on black background
thank you
I have the exact same hum! And it's driving me nuts! and yes it's transformer, i took the side cover off and touched the transformer with a wood stick and it was vibrating like crazy. Odd thing for me is it only just started happening after moving to a new workshop. We have tried power filters but did nothing. I was told it might be because here in Australia our mains are 240V not 220V and if you run it off a special UPS that puts out 220V it might stop it?
Could maybe remount the transformer on some rubber dampers but I think to get to it you need to pull the whole top desk off :-\
-
Is there a bug in the new v2.39 software? After using the "Prick MoveTo" function a few times the prick stops functioning and the unit needs to be power cycled to fix it.
Anyone else seen this problem? BTW, the new software is great. It adds a lot of new features.
Thanks,
Mike
I think I found the same thing. I tried out the"Prick MoveTo" function out and after the pick had gone down I clicked it again, the pick didn't come up but the head moved back to the part and ripped up the tape really good :palm: It did it twice then hasn't done it again?
I think they should change it so once the pick is down, you click it again and it just brings the pick up, then click it again and it moves back to the part.
That way if the pick get stuck on the tape you have a chance to fix it before it wrecks everything.
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
Hi pl116,you can adjust the value of Camera Position to slove it.It is just some offset.
There are no problem with any offsets - I mean the camera itself has lens and sensor not centered each other. and there is no options "offset for image" - what to consider the center of the image on the camera. In other words - first you need to move the red cross in the optical center of the camera, and than checking camera position, offsets and so on.
-
Hi,
I am having a couple of issues setting up my TVM802B. It is a newer model with firmware 3.09.
- When turning it on there is a very loud hum, even when the safety switch is in stop mode. It's not the pump I think, because I can turn that on and off in the software. Is this normal? Sounds like transformer hum. This is on 220V.
- USB recognized one camera as STK1160. There are no further cameras unlike stated in the user manual.
The driver supplied is for a STK1150, so windows 7 would not install this one. I installed a driver for STK1160. Both cameras work with the viewer tool included in the driver, but not with the PnP software.
In the software I can only see a red cross on black background
thank you
I have the exact same hum! And it's driving me nuts! and yes it's transformer, i took the side cover off and touched the transformer with a wood stick and it was vibrating like crazy. Odd thing for me is it only just started happening after moving to a new workshop. We have tried power filters but did nothing. I was told it might be because here in Australia our mains are 240V not 220V and if you run it off a special UPS that puts out 220V it might stop it?
Could maybe remount the transformer on some rubber dampers but I think to get to it you need to pull the whole top desk off :-\
If the sound is too loud,maybe the screws are loose.You can check the screws firstly.
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
Hi pl116,you can adjust the value of Camera Position to slove it.It is just some offset.
There are no problem with any offsets - I mean the camera itself has lens and sensor not centered each other. and there is no options "offset for image" - what to consider the center of the image on the camera. In other words - first you need to move the red cross in the optical center of the camera, and than checking camera position, offsets and so on.
The Camera Position coordinates determine where the nozzle center moves.So adjust it can mvoe the nozzle to the center of camera.
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
Hi pl116,you can adjust the value of Camera Position to slove it.It is just some offset.
There are no problem with any offsets - I mean the camera itself has lens and sensor not centered each other. and there is no options "offset for image" - what to consider the center of the image on the camera. In other words - first you need to move the red cross in the optical center of the camera, and than checking camera position, offsets and so on.
The Camera Position coordinates determine where the nozzle center moves.So adjust it can mvoe the nozzle to the center of camera.
:) it looks like chat with the call-center ivr ... just read www.eevblog.com/forum/manufacture/pick-and-place-machine-tvm802a-tvm802b/msg1728800/#msg1728800 (https://www.eevblog.com/forum/manufacture/pick-and-place-machine-tvm802a-tvm802b/msg1728800/#msg1728800) and please post your picture here. I am talking about how to fix the camera , because optical focus of camera1 not at its center of digital sensor. and you use hardcoded center of the camera sensor as the focus position (red cross).
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
Hi pl116,you can adjust the value of Camera Position to slove it.It is just some offset.
There are no problem with any offsets - I mean the camera itself has lens and sensor not centered each other. and there is no options "offset for image" - what to consider the center of the image on the camera. In other words - first you need to move the red cross in the optical center of the camera, and than checking camera position, offsets and so on.
The Camera Position coordinates determine where the nozzle center moves.So adjust it can mvoe the nozzle to the center of camera.
:) it looks like chat with the call-center ivr ... just read www.eevblog.com/forum/manufacture/pick-and-place-machine-tvm802a-tvm802b/msg1728800/#msg1728800 (https://www.eevblog.com/forum/manufacture/pick-and-place-machine-tvm802a-tvm802b/msg1728800/#msg1728800) and please post your picture here. I am talking about how to fix the camera , because optical focus of camera1 not at its center of digital sensor. and you use hardcoded center of the camera sensor as the focus position (red cross).
Sorry, I understand what you mean now. Could you please give me your email address? I will send you the steps to fix the camera.
-
hi,
I would like to know only I have not good "Camera1" or there are more people ?
Please check it : 1. Move Nozzle1 at camera1 position. 2 Move Nozzle1 down to at last -10. If the nozzle hole keeps at red cross - camera is OK, if not - you have not good camera.
I have such camera - I cannot make the fine place for packages large than so8 ...
Hi pl116,you can adjust the value of Camera Position to slove it.It is just some offset.
There are no problem with any offsets - I mean the camera itself has lens and sensor not centered each other. and there is no options "offset for image" - what to consider the center of the image on the camera. In other words - first you need to move the red cross in the optical center of the camera, and than checking camera position, offsets and so on.
The Camera Position coordinates determine where the nozzle center moves.So adjust it can mvoe the nozzle to the center of camera.
:) it looks like chat with the call-center ivr ... just read www.eevblog.com/forum/manufacture/pick-and-place-machine-tvm802a-tvm802b/msg1728800/#msg1728800 (https://www.eevblog.com/forum/manufacture/pick-and-place-machine-tvm802a-tvm802b/msg1728800/#msg1728800) and please post your picture here. I am talking about how to fix the camera , because optical focus of camera1 not at its center of digital sensor. and you use hardcoded center of the camera sensor as the focus position (red cross).
Sorry, I understand what you mean now. Could you please give me your email address? I will send you the steps to fix the camera.
I sent you PM
-
Folks!
I've got SW version 2.39 the other week from qihe, now the upway camera can't position the IC any more. It moves the component visually to the center, spinns it into the correct position, even if I misplace rotation on the component tray, it rotates it to correct position. Then the head moves out of the center in down direction on the screen. It continues to do so until it reaches the physical sensor switch and fires an error window. I`m not usinmg the camera for ICs frequently so I have`n recogniced it before. Any idea what I can do? I checked the various screen shots here in the forum and in the documentation to figure out what I`m doing wrong. I`m sure it is a config problem - so who has a hint for me?
Thanks
-AH-
-
Folks!
I've got SW version 2.39 the other week from qihe, now the upway camera can't position the IC any more. It moves the component visually to the center, spinns it into the correct position, even if I misplace rotation on the component tray, it rotates it to correct position. Then the head moves out of the center in down direction on the screen. It continues to do so until it reaches the physical sensor switch and fires an error window. I`m not usinmg the camera for ICs frequently so I have`n recogniced it before. Any idea what I can do? I checked the various screen shots here in the forum and in the documentation to figure out what I`m doing wrong. I`m sure it is a config problem - so who has a hint for me?
Thanks
-AH-
they moved camera threshold option from camera's setting to stacks - did you set it for ic stack where you picked up IC?
-
Thanks pl116!
I tried the threshold field with low single digit values. Now I tried higher values that is the solution - THANKS!
-AH-
-
about new in SW 2.39 and threshold option for each feeder - my option, the better way is to keep threshold option for camera1 as the default option and than if I need extra correction for some parts - use threshold option in the stack options.
-
I have some reels where the hole location is on the other side. Meaning if normally the part is on the right side and the hole on the left, here it is mirrored. Tried to correct this with prick correction, but it only goes up to +-2. How do you handle this? thanks
-
Then your tape has been taken from another reel and wound on backwards. Take it off and put it back the right way round.
-
Then your tape has been taken from another reel and wound on backwards. Take it off and put it back the right way round.
:palm: of course, thanks :)
-
I have the exact same hum! And it's driving me nuts! and yes it's transformer, i took the side cover off and touched the transformer with a wood stick and it was vibrating like crazy. Odd thing for me is it only just started happening after moving to a new workshop. We have tried power filters but did nothing. I was told it might be because here in Australia our mains are 240V not 220V and if you run it off a special UPS that puts out 220V it might stop it?
Could maybe remount the transformer on some rubber dampers but I think to get to it you need to pull the whole top desk off :-\
Thanks a lot for confirming this! Here in Germany we have 230-240V. A transformer with such an amount of hum is clearly in saturation. Dampening might help, but I guess it would be better to replace it. It is certainly not the 'positive pump' as stated by qihe_kim: My unit does not have one.
-
I have the exact same hum! And it's driving me nuts! and yes it's transformer, i took the side cover off and touched the transformer with a wood stick and it was vibrating like crazy. Odd thing for me is it only just started happening after moving to a new workshop. We have tried power filters but did nothing. I was told it might be because here in Australia our mains are 240V not 220V and if you run it off a special UPS that puts out 220V it might stop it?
Could maybe remount the transformer on some rubber dampers but I think to get to it you need to pull the whole top desk off :-\
Thanks a lot for confirming this! Here in Germany we have 230-240V. A transformer with such an amount of hum is clearly in saturation. Dampening might help, but I guess it would be better to replace it. It is certainly not the 'positive pump' as stated by qihe_kim: My unit does not have one.
Hi ubbut,
Each machine from our company has the positive pump.And it is okay that machine is connected to the 240V power supply.
Could you send your e-mail by message?I can provide you process of checking the screw of the positive pump.
And it is best to send the Serial Number of you machine.
-
Hello everybody!
Please, can anyone help me? I purchased the TVM802BX, but only the program in Chinese is included (v 3.11). Can someone give the latest version of the software in English?
Thank you in advance.
-
The request is canceled, thanks to the manufacturer - sent me version 3.11 in English :)
-
My unit has a new problem. It is incorrectly placing all parts offset by a constant amount (1-2mm in X/Y). It is able to successfully detect the fiducials, pick up parts from the feeders and do optical recognition. I calibrated the nozzles. Can anyone think of a reason why this is happening? I installed a replacement mainboard and re-entered all of the settings. Is there a hidden menu I need to set?
-
I may have corrected the offset problem by adjusting the Nozzle 1 X/Y offset. It seems to affect part placement with or without vision. Very strange.
On a different topic, is there a good systematic way to adjust the feeders?
-
My unit has a new problem. It is incorrectly placing all parts offset by a constant amount (1-2mm in X/Y). It is able to successfully detect the fiducials, pick up parts from the feeders and do optical recognition. I calibrated the nozzles. Can anyone think of a reason why this is happening? I installed a replacement mainboard and re-entered all of the settings. Is there a hidden menu I need to set?
No, it is just the s/w is weird.
-
Been using the machine now for a while and have some feature requests. Please consider implementing these, as this would greatly simplify working with the machine. Thank you (I have software version 3.09)
- Most importantly: instead of a global 'Down to stack' parameter, please introduce a different one for each stack. At least for the IC-stacks this is mandatory. Instead of having to build all trays to the same hight, just define it in software. Please
- It would be absolutely great if we could use camera 2 to calibrate the feeders. Part position and prick offset. This would be much faster and simpler then the current way of using the nozzle and do several test pricks
- When configuring a stack, only nozzle 1 can be used. This is not great, because I would like to be able to test pick-up with nozzle number 2. Please include this feature.
- At least in my software version (3.09) Move mode is restricted to 'nozzle' in the IC stack configuration, 'vision' can not be used. Please make this possible. Also again, we can only use nozzle 1 here, It would be great if nozzle 2 was selectable. if have had situations where I configured a stack with nozzle 1, only to find that when picking with nozzle 2, the machine will run into its end-switches
- Speaking of end-switches. If the machine knows where it is and how large it is - is it really necessary to run it all the way into the end switch. Could it not stop before? (And not create the laud beeping sound)
- At least in my particular machine there is a strange bug: If I use the manual controls to move down nozzle 2, when it reaches the limit it is not stopped by software. Rather (the belt) slips and both nozzle 1 and 2 position will be wrong. Only a complete shutdown of the machine fixes this
Thank you very much !!
-
On a different topic, is there a good systematic way to adjust the feeders?
I would love to know that, too. I start defining the parts positions with the nozzle), then test the prick movement. However often after the prick movement, the part is in a different location and I have to define it again. This is quite confusing.
-
Hi!
instead of a global 'Down to stack' parameter, please introduce a different one for each stack. At least for the IC-stacks this is mandatory.
Parameter 'Down to ... ' for the IC-trays is 'Down To Front' (see in attachment), it can be configured separately from the feeders :)
It would be absolutely great if we could use camera 2 to calibrate the feeders.
But the camera2 can be used to configure feeders. For this, there is a "Vision" switch and a "Vision MoveTo" button (see attachment).
In general, setting up feeders is a fairly simple process when you begin to understand its principle. But it takes quite some time for all positions.
At least in my particular machine there is a strange bug: If I use the manual controls to move down nozzle 2, when it reaches the limit it is not stopped by software. Rather (the belt) slips and both nozzle 1 and 2 position will be wrong. Only a complete shutdown of the machine fixes this
This is a lack of hardware - there is only a sensor for the upper position of both nozzles, but there are no sensors for the lower position. However, it is unclear why this sensor does not work when the nozzle is raised after the nozzle's normal position has been disturbed.
My software version is 3.11, machine is TVM802BX.
-
Parameter 'Down to ... ' for the IC-trays is 'Down To Front' (see in attachment), it can be configured separately from the feeders :)
Yes but not for each IC-Tray separately. If I have several trays, they all need to be the same hight...
But the camera2 can be used to configure feeders. For this, there is a "Vision" switch and a "Vision MoveTo" button (see attachment).
In general, setting up feeders is a fairly simple process when you begin to understand its principle. But it takes quite some time for all positions.
You are correct of course. I think I should have specified that it would be great to also adjust the Prick Correct with the camera.
-
If I have several trays, they all need to be the same hight...
But you can adjust the height of each component, which will be more correct, I think :) This height is reflected both on the installation on the PCB and on the capture from the tray.
I think I should have specified that it would be great to also adjust the Prick Correct with the camera.
Yes, here you are right, it would be more convenient to correct the prick with the camera too. Now we have to move the prick two or three times for a satisfactory correction of the prick. But it's better to check not the nozzle, but the camera :)
From myself, I can add something that I really miss when calibrating feeders - divisions on the camera lines for more accurate alignment of the crosshairs at the center of the component. Now I have pasted on the screen strips of transparent scotch with the marks drawn by the marker (in the attachment).
-
If I have several trays, they all need to be the same hight...
But you can adjust the height of each component, which will be more correct, I think :) This height is reflected both on the installation on the PCB and on the capture from the tray.
This sounds more like a work-around. Also quite sure it does not work. Imagine I have a tray sitting 10mm lower than the other one. I define the component height to be 10mm/11mm. It will be picked up ok, but then dropped in mid-air over the PCB
-
Probably, it means that all trays of the same height :) I print trays on a 3D printer, one for all the necessary components. The bottom is at the same height, but components of different heights and this height I specify in the settings of the components.
-
Discovered the prick solenoid mounting screws were loose. Needless to say, realigning and tightening them has greatly improved the feeder reliability.
-
I just did a run of boards and it appears that parts are shifted toward the left with increasing Y coordinate (looking at the front of the machine). Does this imply the Y-Scaling value (currently set to 1) needs to be increased?
Thanks,
Mike
-
I just did a run of boards and it appears that parts are shifted toward the left with increasing Y coordinate (looking at the front of the machine). Does this imply the Y-Scaling value (currently set to 1) needs to be increased?
Have you put several separate boards? If so, then the reason may be in the not very precise external size of these boards. It is better to check on the ruler instead of several boards or on one big board.
-
No, just one board.
-
No, just one board.
Then yes, you need to increase the value of scaling, but in X, not Y, since the components are shifted to the left. I correctly understood - the more the component is located to the right, the more it is shifted to the left?
-
Yes, you are right. It is the x scaling.
Thanks,
Mike
-
After verifying the X/Y scaling is correct with a ruler, I ended up slightly modifying the fuducial coordinates to fix the placement issue. I don't like the solution very much but is works. Can anyone think of another adjustment to compensate for parts being shifted with increasing X value?
-
Hello everyone,
Was looking into getting my own PnP machine I do small runs (50 boards a year or less) of a heavily populated 11" by 7" PCB, just one sided mostly all 603, SOIC, TSSOP and SOT components. I have about 4000 Components per board, would you guys recommend something like the TVM802B for my use case? Looking for general feedback as this is new territory for me, I pay a lot for a local PnP shop so this could save me money in the long term but if its something that's constantly breaking or that has a lot of quirks I may just back off.
-
Hello everyone,
Was looking into getting my own PnP machine I do small runs (50 boards a year or less) of a heavily populated 11" by 7" PCB, just one sided mostly all 806, SOIC, TSSOP and SOT components. I have about 4000 Components per board, would you guys recommend something like the TVM802B for my use case? Looking for general feedback as this is new territory for me, I pay a lot for a local PnP shop so this could save me money in the long term but if its something that's constantly breaking or that has a lot of quirks I may just back off.
How many different components / board ? Mostly 0603 / 0805? TVM802 runs mostly smooth when set-up correctly, although I like to watch in case something goes wrong. It's not a fast machine by industry standards, but if you can afford to wait a bit it is great. At 100% speed it surprisingly fast IMHO.
-
I have 76 different components per board. The majority are smd 0603 with some 1206.
-
I have 76 different components per board. The majority are smd 0603 with some 1206.
TVm802B has 48 feeders. Even if you manage to use them all and put some components in trays, you'll have to load the machine twice with such a board. Maybe you can get away by placing the less used components by hand. That depends on the number of boards of course. You'll have this problem with every machine and it is probably the reason your assembly house charges so much.
Usually it is possible to reduce the number of components: A pullup does not have to be 100k for example.
-
I do small runs (50 boards a year or less)
Hard to see how it would make sense doing that few in-house
have 76 different components per board. The majority are smd 0603 with some 1206.
The first thing to look at is whether you can reduce that BOM. That would also save costs on subcontract assembly.
-
I have 76 different components per board. The majority are smd 0603 with some 1206.
TVm802B has 48 feeders. Even if you manage to use them all and put some components in trays, you'll have to load the machine twice with such a board. Maybe you can get away by placing the less used components by hand. That depends on the number of boards of course. You'll have this problem with every machine and it is probably the reason your assembly house charges so much.
Usually it is possible to reduce the number of components: A pullup does not have to be 100k for example.
I do small runs (50 boards a year or less)
Hard to see how it would make sense doing that few in-house
have 76 different components per board. The majority are smd 0603 with some 1206.
The first thing to look at is whether you can reduce that BOM. That would also save costs on subcontract assembly.
Thank you both, essentially my BOM is high because my PCB has a 16 channel analog front end. I dont mind taking 2 passes to do the board I'm just hesitant because I have no experience in the assembly side of things, so I don't know if I'd be buying something with tons of issues or is not reliable.
-
I dont mind taking 2 passes to do the board
Replacing almost all tapes in feeders is a rather long and boring procedure.
-
I dont mind taking 2 passes to do the board
Replacing almost all tapes in feeders is a rather long and boring procedure.
It is a profit destroyer......
-
You only say:
I don't mind taking 2 passes to do the board
Because of the following:
I'm just hesitant because I have no experience in the assembly side of things, so I don't know if I'd be buying something with tons of issues or is not reliable.
When you have tried changing the feeders even a little more than is absolutely necessary you will probably think differently. It combines tedium, long hours, and a significant crushing of the soul.
-
Thank you all, I currently pay well over $100 per board so I was hoping this could maybe be a solution for me but I see i have more research to do.
-
Thank you all, I currently pay well over $100 per board so I was hoping this could maybe be a solution for me but I see i have more research to do.
At the cost of these machines, it COULD make sense to buy two of them, and have each one set up for half the parts. That, of course, doubles the expense, but at $100/board X 50 boards, that could pay for itself in a couple years. You are kind of on the dividing line about whether this is financially sensible, though. I do about 15 different boards, some all digital, some largely analog, so just change over the machine for each job. I run anywhere from 20 to 60 boards at one time, and many of these have parts on both sides, mostly decoupling caps on the back. So, that is sort of two setups/board. In my case, the full setup for one of my boards can take an hour or a little more, depending on number of different parts. I have several boxes of passive component reels sorted by value, and just load up the feeders as needed, then set up the special parts (chips and semis). My Philps CSM84 can theoretically hold 84 feeders, but due to a vibratory feeder in front and a large component aligner in back, it can only hold about 70 8mm feeders, and less when larger sizes are used. I've never run out of feeder slots on that machine, but I have had to put some feeders on the back rail for some boards.
Jon
-
People have a very naive/misunderstood idea of what the problems and benefits of a pick and place machine are.
As others said, changing parts from feeders is a HUGE hassle. Think of it like changing the hard drive in your computer. It's not really that big of a deal, but if you had to do it 20 times per day (each time you wanted to run a new program), you would very quickly become tired of it. You would do anything to find a solution which let you not change the hard drive. That's about the same amount of hassle/tedium as changing parts in feeders.
It is most definitely not just sticking the tape in and you are good. It's a mechanism where you need to peel back the cover, feed the tape in, manually move it forward until the component appears, wrap the tape cover on whatever take-up mechanism the feeder has, etc. When removing a tape from a feeder, you inevitably lose anywhere from a dozen to several dozen parts. Then you have a cut tape which will lose the same number of parts again when you re-load that into a feeder. Beyond that, you have to tweak the feeders on the machine... adjust the pick-up height, the placement height, the pickup location, the rotation, the air vacuum and various other parameters. That can take as long (or much longer) than actually changing out the tape. You would be lucky to achieve 20-30 minutes per feeder for new parts, and 5+ minutes for existing parts which are already programmed. And it's not 5 minutes of enjoyable work. It's tedious, repetitive, boring, and error prone.
Anyone who has gone down this road will quickly want to amass a library of feeders with a specific part dedicated to a feeder. But that gets expensive, but even so, the cost of a feeder quickly eclipses the tedium and hassle of changing out a tape. As does the tedium of programming a new part eclipse the hassle/cost of re-using what you already have programmed (serializing/paralleling existing parts rather than loading new ones, etc).
PnP is ideal for long runs of boards with the smallest number of unique parts. As soon as you start PnP'ing, you will VERY quickly start to design your boards to be as much like that as possible, because the hassle of PnP is huge and you need to run a lot of boards or have the right kind of boards/BOM's to make it worthwhile.
-
Amen brother.
-
Thank you all, I currently pay well over $100 per board so I was hoping this could maybe be a solution for me but I see i have more research to do.
I was faced with a similar cost on 3 PCB's I was producing a few years ago when I decided to dive into a P&P system.
Soon, I learned why the cost was so high - the whole process is a laborious pain in the butt with a huge and steep learning curve. The Pick and place machine WILL NOT solve your problem, it is just a hammer in the context of building a house. It is important, but if all you have is a hammer, you will never build a whole house.
However much you think P&P will cost, it will cost more. However long you think it will take to learn the process, it will take longer.
If you stick with it and shed some blood and have the business volume.....it is great. If you believe it will ride in like unicorn to sprinkle magic on your assembly efforts - that unfortunately will not happen. I am personally still shell-shocked with the learning curve that continues a few years into it. I also got a commercial system with a giant pile of feeders so I was able to skip the problems that plague the small/low cost machines.
When anyone thinks of Pick and Place:
IT IS NOT A MACHINE. IT IS A COMPLEX PROCESS CONSISTING OF MACHINES, SKILLS, SOFTWARE, AND ORGANIZATION
-
If you don't limit yourself to the TVM802B, there are some machines in similar or slightly more price bracket that could accommodate your 70 line BOM (depend on how many is on 8mm, 12mm, 16mm, etc tape width). SmallSMT has some model (their DP machine range I believe) that could fit. Others model from QiHe, Charmhigh or Neoden most probably didn't have that high count feeder machine available.
-
I added a few feeders as IC stacks - look at picture ... +15 ;)
If someone need it I can send stl files to print it .
-
Unfortunately, these are actually not feeders, but SMD trays :)
-
Unfortunately, these are actually not feeders, but SMD trays :)
robotdigg have drag feeder parts that you may be able to retrofit.
-
Unfortunately, these are actually not feeders, but SMD trays :)
"SMD Tray" - is a feeder but with a manual drive ))) anyway it helps me ...
-
@qihe_kim I tried emailing sales1@qihekj.com, I've not received a response in 2 days? What is the best way to procure one of these machines? Thanks Paul
-
Offtopic:
Hey guys, good evening.
I'm new here but have been reading through this topic a lot, and some others in the past as well.
I'm owner of the retrocomputer company ByteDelight, and after assembling a couple of thousand add-on devices by hand, it's finally time to have a pick and place machine do it.
I selected the TVM802A because of its speed and the camera.
Since I do a lot of pcb designing myself, I have no problem at all selecting parts for the reels and adapting my designs to use a smaller number of different components, so I'm really looking forward to adding this to my production line.
Ontopic:
I'm currently looking at AliExpress for the TVM802A, but there are so many offers (514!!) at this moment.
I would just order the cheapest one since I thought they are all equal, but just discovered that there are some differences, like the integrated PC in the more expensive ones.
Could you guys help me finding answers to these questions (still investigating some of them myself, hope I'm not too much of a trouble)?
1. I planned on using my own separate W10 machine to control the TVM802A (have a Core2 Duo laptop standing by for that), so is it ok to go with the cheapest model without integrated PC?
2. Could some of the machines offered be older versions, before major updates?
3. Not all descriptions specifically mention a 'high voltage drive' - what is it and do all come with it?
4. Some offers specifically mention a guide rail (for easy pcb positioning it seems), but it seems the cheapest also has that, am I right?
5. The cheapest ones all seem to have the SMD trays, but some offers do not show it on the photos - do some indeed go without?
6. If I read it correctly, the machine has both an Ethernet and a USB port - I guess you need to use the USB port to control the machine, but what is the Ethernet port for?
7. Are there specific sellers not to buy from maybe?
Ben Versteeg
-
Offtopic:
Hey guys, good evening.
I'm new here but have been reading through this topic a lot, and some others in the past as well.
I'm owner of the retrocomputer company ByteDelight, and after assembling a couple of thousand add-on devices by hand, it's finally time to have a pick and place machine do it.
I selected the TVM802A because of its speed and the camera.
Since I do a lot of pcb designing myself, I have no problem at all selecting parts for the reels and adapting my designs to use a smaller number of different components, so I'm really looking forward to adding this to my production line.
Ontopic:
I'm currently looking at AliExpress for the TVM802A, but there are so many offers (514!!) at this moment.
I would just order the cheapest one since I thought they are all equal, but just discovered that there are some differences, like the integrated PC in the more expensive ones.
Could you guys help me finding answers to these questions (still investigating some of them myself, hope I'm not too much of a trouble)?
1. I planned on using my own separate W10 machine to control the TVM802A (have a Core2 Duo laptop standing by for that), so is it ok to go with the cheapest model without integrated PC?
2. Could some of the machines offered be older versions, before major updates?
3. Not all descriptions specifically mention a 'high voltage drive' - what is it and do all come with it?
4. Some offers specifically mention a guide rail (for easy pcb positioning it seems), but it seems the cheapest also has that, am I right?
5. The cheapest ones all seem to have the SMD trays, but some offers do not show it on the photos - do some indeed go without?
6. If I read it correctly, the machine has both an Ethernet and a USB port - I guess you need to use the USB port to control the machine, but what is the Ethernet port for?
7. Are there specific sellers not to buy from maybe?
Ben Versteeg
I recently bought a machine for OpenPnP development, and the TVM was one of the options I considered. Maybe my notes will be helpful:
http://www.charmhigh-tech.com/sale-7971156-economic-model-chmt36va-smt-pick-and-place-machine-top-and-bottom-camera-externel-pc.html (http://www.charmhigh-tech.com/sale-7971156-economic-model-chmt36va-smt-pick-and-place-machine-top-and-bottom-camera-externel-pc.html)
SparkFun's choice. They probably researched it a bit.
$2800 + $640 DHL = $3440
Juki nozzles
29 drag feeders
https://www.robotdigg.com/product/865/RDG802A-pick-and-place-machine (https://www.robotdigg.com/product/865/RDG802A-pick-and-place-machine)
Clone of Qihe TVM802A
26 feeders
dual vision
$3260 shipped
https://www.robotdigg.com/product/864/RDG802B-pick-and-place-machine (https://www.robotdigg.com/product/864/RDG802B-pick-and-place-machine)
Clone of Qihe TVM802B
Same as above, but with 46 feeders
$4190 shipped
https://www.robotdigg.com/product/1173/RDG802A-X-or-RDG802B-X-Embedded-PC-workbench-pick-and-place-machine (https://www.robotdigg.com/product/1173/RDG802A-X-or-RDG802B-X-Embedded-PC-workbench-pick-and-place-machine)
They say same as above, but includes the PC built in.
But also description says "Closed-loop Servo Stepper Motor (Flexible S deceleration, 80V high voltage drive and S curve acceleration control )"
The description on Qihe's site says the same, but doesn't for the two machines above, so maybe the X gets you faster motors.
A $3590 shipped
B $4590 shipped
https://www.robotdigg.com/product/463/Desktop-Pick-and-Place-Machine-CHMT28 (https://www.robotdigg.com/product/463/Desktop-Pick-and-Place-Machine-CHMT28)
No cameras
25 feeders
$2000 before shipping
https://www.robotdigg.com/product/464/Desktop-Pick-n-Place-Machine-CHMT36-or-CHMT36VA (https://www.robotdigg.com/product/464/Desktop-Pick-n-Place-Machine-CHMT36-or-CHMT36VA)
CHMT36 is no vision, with touch screen. Looks like a clone of the TM-240
29 drag feeders
$2200 before shipping
CHMT36VA is the Sparkfun machine
Same as above with vision, no touch screen, PC required
$2980 before shipping
Probably better to buy direct from CharmHigh tho
https://www.robotdigg.com/product/465/Desktop-Pick-and-Place-Machine-CHMT48VA (https://www.robotdigg.com/product/465/Desktop-Pick-and-Place-Machine-CHMT48VA)
48 feeder with vision
$4100
https://www.robotdigg.com/product/1099/Benchtop-CHMT48VB-pick-and-place-machine (https://www.robotdigg.com/product/1099/Benchtop-CHMT48VB-pick-and-place-machine)
56 feeder with vision
$4550
https://www.robotdigg.com/product/730/2-or-4-head-Pick-n-Place-Machine-CHMT528-or-CHMT530P4 (https://www.robotdigg.com/product/730/2-or-4-head-Pick-n-Place-Machine-CHMT528-or-CHMT530P4)
Looks like the TVM920 base
Embedded Linux
Uses Yamaha CL feeders but doesn't include any
T528P 2 nozzles, 28 feeders $3998 before shipping
T530P4 4 nozzles, 30 feeders $4998 before shipping
not a lot of details, but clearly dual vision
Good luck!
Jason
-
1. I planned on using my own separate W10 machine to control the TVM802A (have a Core2 Duo laptop standing by for that), so is it ok to go with the cheapest model without integrated PC?
2. Could some of the machines offered be older versions, before major updates?
3. Not all descriptions specifically mention a 'high voltage drive' - what is it and do all come with it?
4. Some offers specifically mention a guide rail (for easy pcb positioning it seems), but it seems the cheapest also has that, am I right?
5. The cheapest ones all seem to have the SMD trays, but some offers do not show it on the photos - do some indeed go without?
6. If I read it correctly, the machine has both an Ethernet and a USB port - I guess you need to use the USB port to control the machine, but what is the Ethernet port for?
7. Are there specific sellers not to buy from maybe?
From my experience:
1. Should work fine. The integrated computer is very weak (Intel Atom D2550, 2 GB of RAM, 32 GB SSD), but it works successfully.
2. I do not think that there were any significant changes in hardware.
3. A high-voltage driver means a higher voltage supply for motor drivers, this gives higher accelerations and speeds of movement of the axes. It seems that they are only available in X-versions (802AX and 802BX). Additionally, in these versions there is a feedback between the motors and drivers, which also improves the speed and accuracy of work. In fact, there are stepper servomotors there.
4. Yes, guide rails for PCB are available in all versions.
5. SMD trays should be bundled in all versions. But in any case it is best to make your own trays (3D printing or milling).
6. In the 802BX version, the machine controller is connected to the embedded computer via Ethernet. How is the connection in 802A - I'm not sure, but it is possible that two ports at once - Ethernet for controlling the machine and USB for receiving images from cameras.
I myself bought an 802BX machine on the Taobao site from this seller - https://item.taobao.com/item.htm?id=525680982287
-
Thanks for the replies to my questions.
I just ordered the TVM802A from "Desh Woo's store" on AliExpress.
-
Thanks for the replies to my questions.
I just ordered the TVM802A from "Desh Woo's store" on AliExpress.
What version of software comes with your machine ? 2.xx or 3.xx ?
-
hi folks, if in my CSV file I have a rotation of the component 180 degree, so when i load this csv file from computer to software the angle becomes 90 automatically same for every component on PCB, can anybody help me in this,,what should i do.
-
hi folks, if in my CSV file I have a rotation of the component 180 degree, so when i load this csv file from computer to software the angle becomes 90 automatically same for every component on PCB, can anybody help me in this,,what should i do.
Show here the original .csv file.
-
here I am sharing the pic of the .csv file, here you can see the angle of the first component is 270(R46-1), but when I upload this for placement it becomes 180, and same for all the other components. I have checked many aspects but still failed.
-
I meant to attach the file itself so that we can try to open it in the PnP software :)
-
I am unable to upload Csv file here, but let me tell you it is the same file that came with TVM802B, (as an example).
-
csv file
-
In my case, the program did not want to open your file at all. When I try to open it shows an error (screenshot attached).
This was corrected by renaming the column “Footprint Description” to “Footprint”. After this, the program normally opened the file with the correct angles (a screenshot is attached). UPD: Excuse me, the angles still remain wrong. I'll try to tinker with this, find the cause.
If you open a .csv file in Exel, change something in it and save it, then you need to delete the header of the file before the name of the columns.
Regarding the column names from my experiments with .csv files:
The software of the machine requires several required fields:
- "Designator" (can also be "RefDes") - component tag, for example, R12, C6, etc.
- "MidX" (can also be "Center-X" or "Center-X (mm)") - component coordinate along the X axis.
- "MidY" (can also be "Center-Y" or "Center-Y (mm)") - component coordinate along the Y axis.
- "Rotation" - component installation angle.
- "Explanation" or "Footprint" - about them below
Optional field:
- "Layer" - component installation side. If this field is present in the file, then when opening the file, the program will ask for which side of the board to take components from the file. If you specify the bottom side, then the "Reverse X" option will be available for mirroring the coordinates along the X axis.
- "Explanation" - description of the component, it can be any value, for example, the type of housing and the value for resistors and capacitors, or the name for the chips, etc. Serves solely for the convenience of the operator. If this field is not present, then the program inserts the "Footprint" and "Comment" values into the description.
- "Footprint" - the name of the body component.
- "Comment" - any description of the component, for example its face value.
Concerning the "Explanation", "Footprint" and "Comment" fields. If the file does not contain one of the two fields - "Explanation" or "Footprint", the program crashes with an error when trying to open such a file. If there is an "Explanation" field, the program takes a value for the "Comment" column in the table displayed from this field, ignoring the "Footprint" and "Comment" fields. If the "Explanation" field in the file is not present, then for the "Comment" column get the value from the "Footprint" plus "Comment" fields (if any).
Any other fields in the .csv file will simply be ignored by the machine program.
Coordinates when importing are rounded to hundredths of a millimeter.
This all refers to software version 3.11. I did not explore earlier versions.
-
well, you have explained it very well, but from pic still, there is a 90-degree difference in the uploaded file. when I run this program the placement of resistor( is accurate in according to the angles). but the problem comes when it places SOT23. as it is three pins package, so when I add 90-degree manually in the program then the placement becomes accurate.
-
I never paid attention, but now I looked - in all my files the angle also changes. However, the components are put on the board correctly.
With SOT23 - yes, I have to change the angle, but I always believed that it was because I created the component in the library in the wrong position. (By the way, changing the parameters for several components in the software of the machine is quite simple. It is necessary to select lines with these components and change the parameter in any of them. This parameter will automatically change in all other selected lines.)
As I understand it, the components must be created in the library in the position in which they are located in the feeders, standing in front of the machine.
Maybe for this there is some kind of industry standard in PCB design?
And as for a machine 0 degrees is the position from the left feeders, it immediately adds 90 degrees when importing a file. This is like a guess, I don’t understand the real reason :)
-
well, you have explained it very well, but from pic still, there is a 90-degree difference in the uploaded file. when I run this program the placement of resistor( is accurate in according to the angles). but the problem comes when it places SOT23. as it is three pins package, so when I add 90-degree manually in the program then the placement becomes accurate.
Hi maheed,For most resistors and capacitors, there is 90° different between the feeder and the drawing when the components are in the left feeder.So we add 90° in the left feeder when nozzle pick up components.
-
The TVM802A arrived this morning.
Doing a detailed video blog about it, starting with the arrival!
Will update this webpage as well: https://www.bytedelight.com/?page_id=3298 (https://www.bytedelight.com/?page_id=3298)
https://www.youtube.com/watch?v=IL0MDOvFfZA (https://www.youtube.com/watch?v=IL0MDOvFfZA)
-
bverstee, congratulations! :)
-
Uncrating the huge machine:
https://www.youtube.com/watch?v=WfEAC46igDU (https://www.youtube.com/watch?v=WfEAC46igDU)
-
@bverstee
Hello,thank you for choosing our TVM802A machine. The video you made is very good. :-+ We have professional after-sales technical support,if you have any doubt with using machine,can connect us.
We will also work hard to perfect our machines, and finally thank you for choosing our machines again.
-
Hello!
Can I ask? Add to the camera the gradation of the ruler on the camera in order to more accurately position the components and not by eye. I have now drawn a gradation pencil straight on the monitor, but it spoils my monitor.
(https://pic.mysku-st.ru/uploads/pictures/02/88/88/2018/09/22/d23276.jpg)
Thank! I hope they hear me.
-
I join the request to add rulers to the camera screen relative to the center. This is very lacking. The photo above is mine, I pasted an adhesive tape with rulers on the screen, but they interfere in other tasks.
-
I join the request to add rulers to the camera screen relative to the center. This is very lacking. The photo above is mine, I pasted an adhesive tape with rulers on the screen, but they interfere in other tasks.
What are you using this for?
-
What are you using this for?
For example, when calibrating feeders and when searching Fiducial Marks on boards. Rulers allow to more accurately assess the position of the object under the camera.
-
to bad bverstee stopped showing any progress with the machine he purchased. It would have been interesting to see his progress and how the machine performs.
-
I still don't get it ;) Does not matter.. I guess we just have different approaches to calibrating feeders and finding fiducials.
An other thing I don't understand: Why can I not use the down camera to calibrate IC-Stack locations? It's disabled in software (3.09)
Also: Does anyone know how a I can disable the use of both nozzles simultanously? While it looks cool when working, if one of the parts causes problems, it will drop both parts. Then tries to pick up both again, only to fail again and then drop both. This is repeated until you press stop. I often have parts all over the place because of this..
-
After very accurately calibrating the upper chamber with respect to the nozzles, it is very convenient to calibrate the Left Stack, Back Stack and IC Stack with the camera, but without gradation on the camera one has to guess about the positioning accuracy. The gradation function is optional, and is required only for a very accurate Stack positioning, in the extreme case the gradation function can be implemented to be disabled in the settings, I think it is quite simple to implement at the software level.
Thank!
-
and IC Stack with the camera
Can I ask you which software version you have? In my version one can not calibrate the IC stack with the camera for some reason. thanks
-
V3.15, at the moment I also have no vision mode in IC Stack. But the point Vision or Nozzle is present, which means that in the near future it will become available, I believe in it).
Attached image.
-
Just popping in to give a more long-term report on mine, been running it close to two year now I think, and have run a few thousand boards through it. My boards are really nothing complex at all. Only 3-4 unique parts at most, up to about 75 total placements on average. Mainly rows of passives, LED's, things like that. Occasionally some QFP's and various bodied transistors/rectifiers. But basically I just place 1206 resistors, reel after reel, about 50-100 board sets per month.
First things is, the feeder situation is a serious pain in the neck. Paper reels aren't too bad now, since I've come up with a 3D printed insert into the feeder bank, but the springs that come from the factory can be annoying at best. They were the biggest hurdle I've had to overcome. But I still struggle with embossed plastic carrier tape. Since I don't use that many parts with the plastic tape I haven't worked too hard at a solution yet. I'll get around to another insert design for plastic carrier tape eventually.
The pin-prick for tape advance, what a nightmare. I've had issues with this since the machine was new. Busted about a half dozen pins as it gets caught on stuff, mainly due to the plastic carrier tape. Since I started using those inserts for paper tape I haven't broken a pin. I'm still having regular issues with the pin sensor, it basically doesn't work at all anymore. Machine still runs but extremely slowly as the software adds a delay of about 1.5 seconds when it advances the tape. Just e-mailed Qihe about this one as I really need to get my cycle times back to where they were. My most common board went from a cycle time of about 2 minutes to around 5.3 minutes. I also find using the 1/2 nozzle setting causes more feeder issues than its worth. So I just use a single advance now all the time. Reliability to me is more important than speed.
Trays, trays are easy. I've done some 3D printed ones to fit specific parts. I'm thinking about some kind of clamp to hold full sized trays but I don't have to use those very often. So the original metal tray, or my printed ones, tend to be sufficient. Since it avoids the issues with the pin-prick advance system I tend to never have issues with tray parts.
Otherwise the machine seems to be running very solid for me now. Sure, it's slowed down due to that pin sensor, but it's finally at the point I can just put in a board, press the button, and walk away for a bit. If I can get that pin sensor solved, and get plastic carrier tape to feed reliably, I'd be pretty happy with it. However I do look forward to when finances justify something with more normal feeders.
-
But I still struggle with embossed plastic carrier tape. Since I don't use that many parts with the plastic tape I haven't worked too hard at a solution yet. I'll get around to another insert design for plastic carrier tape eventually.
Do you mean similar plastic tapes like black in the photo? I have no problems with such tapes.
In general, I had some minor problems with the feeders - from time to time the driving pin did not come back up, remained in the tape. Sometimes the weakening of the pressure springs helped to solve that problem. But fundamentally, I decided this problem by hanging all the reels on the aluminum tube. Without it, the reels are quite hard to rotate, especially when the tape comes to an end. And on this axis they rotate very easily. And now the pin has ceased to get stuck in the tape :)
-
Not a bad idea with using a rod/tube to hold the reels, I may try that.
But yes, those are the tape style I have trouble with. They don't advance properly, get mangled by the pin, and parts go flying all over the place.
-Hans
-
I then do not know what to say. I have such tapes from the very beginning work fine, and 8 mm and 12 mm with different components.
Maybe you have one of the screws loosened and the edge of the tape falls between the base and the washer and gets stuck there? I drew a sketchy tape in the holder (in section).
-
Not a bad idea with using a rod/tube to hold the reels, I may try that.
With the reels propped up using a rod my sticky prick needle issues are a thing of the past, so I highly recommend it. Weird that they don't ship this with the machine. Note that in my case just a rigid rod did not help: I have to prop the rod up with some cardboard scraps so the rod actually touches the ceiling of center holes of my reels. All my reels seem to have different center hole dimensions.
But yes, those are the tape style I have trouble with. They don't advance properly, get mangled by the pin, and parts go flying all over the place.
I had the same issue but in my case it was simply because the prick needle was not aligned perfectly with the hole pattern. It was off by like 0.2mm or so. The default setting is set up for paper tapes, not plastic ones. Play around with the "PrickCorrect" setting for your reels, Y axis for the left stack and X axis for the back stack. If it catches correctly once it should not mangle the tape at all anymore. The manual talks about this in length in chapter 9.6, though it does not read well...
-
I ordered one of these from QiHe a while ago, finally getting it up and running. What a mistake this has been.
1. Getting the order originally was a pain. They mis-declared pricing on the invoice so I had to spend more in duties to get this imported. When I contacted them they sent me an itemized, but still incorrect invoice.
2. My left and right head rotations seem to be just barely out compared to one another, so parts are always crooked. There doesn't seem to be any way to adjust this. With vision it improves but it's still off which is strange. Most of my designs are art / presentation type designs so if I am doing one of these I need to be very careful.
3. Which brings me to documentation. I fought for several hours trying to get QFPs to work, turns out there's a setting that the manual tells me to reference a section that doesn't exist that affects recognition for vision. As for the manual, there isn't anything digital, only an outdated stapled together paper book that is very poorly translated and has outright omissions, inaccuracies and references to non-existant sections.
4. Homing isn't accurate at all. Need to readjust a bunch of things every time I turn this unit on. I understood there would be some time waste in a unit like this but this is beyond frustrating.
5. I ordered some other stuff through them, figured it was OK and has been in the box. Well I go to open the box and it's not OK. Sure, my mistake but it was safer in the box. Having missing parts that we expressly discussed they would send is disappointing, and is now stopping me from going into production until I spend more elsewhere to get this resolved.
6. Emails to them only result in a response of "send us pictures and video" - the questions are is there updated documentation, and what is the latest software?
Does anyone have updated documentation for this unit? Anything digital? What about the latest software, does it fix anything notable? I'm on 3.13 right now.
I haven't experienced a broken advance pin yet, where would I find a replacement? QiHe seems to be worthless to contact.
-
to bad bverstee stopped showing any progress with the machine he purchased. It would have been interesting to see his progress and how the machine performs.
I haven't, I only needed more time because of all the work during the last months.
Finally I decided to close my webshop for just over a week to catch up on some things, and started working on the machine again.
Videos are coming!
-
Hi,
what is the latest software version of 2.XX currently ?
thanks.
-
Just uploaded the first run of my TVM802A!
https://www.youtube.com/watch?v=fD7WkfirgBA (https://www.youtube.com/watch?v=fD7WkfirgBA)
-
Did a full run after lots of tweaking and tuning today.
Very satisfied with the results, but also got frustrated a couple of times during the process, especially when the prick is stuck and the machine crashes when tearing some tapes while moving |O |O
https://www.youtube.com/watch?v=qc10XJ5Lr60 (https://www.youtube.com/watch?v=qc10XJ5Lr60)
-
Did a full run after lots of tweaking and tuning today.
It looks like besides the larger component trays there is a bottom up camera but it is not used at all?
Shouldn't the machine pick the large part and check position and rotation on that camera and then correct for any off center pickups before it places the component?
-
Did a full run after lots of tweaking and tuning today.
It looks like besides the larger component trays there is a bottom up camera but it is not used at all?
Shouldn't the machine pick the large part and check position and rotation on that camera and then correct for any off center pickups before it places the component?
You'd think so wouldn't you. Looking on Youtube at other videos it would seem using the camera is an option you choose to use or not, doing so would presumably remove most of those alignment issues. I would suggest using it for all components, if it works the way you would expect it should then auto correct the pick position for the next attempt. It will be interesting to hear how much easier this process becomes when you are more familiar with the machine and its software.
-
Did a full run after lots of tweaking and tuning today.
It looks like besides the larger component trays there is a bottom up camera but it is not used at all?
Shouldn't the machine pick the large part and check position and rotation on that camera and then correct for any off center pickups before it places the component?
The 'vision' (IC detection algoritm) didn't work at all with my tests (it didn't properly detect the parts dimensions) so I disabled it for now.
Will look into it later.
-
A video showing the whole process from stencilling until using the completed product:
https://www.youtube.com/watch?v=uOouhUPhANo (https://www.youtube.com/watch?v=uOouhUPhANo)
-
Pretty nice product you made. :-+
The placement of the p&p is still off but the camera should fix that.
Your squeegee is a bit small, at the Action with the paint supplies dept. they have a set of four or five blue plastick for €0,89 , they work fine ;)
-
Yeah, thanks, I should get a bigger one.
As said, the vision software is not able to detect the dimensions of the component at this moment, I need to look into it.
Hope it's easy to fix.
-
Did a full run after lots of tweaking and tuning today.
It looks like besides the larger component trays there is a bottom up camera but it is not used at all?
Shouldn't the machine pick the large part and check position and rotation on that camera and then correct for any off center pickups before it places the component?
The 'vision' (IC detection algoritm) didn't work at all with my tests (it didn't properly detect the parts dimensions) so I disabled it for now.
Will look into it later.
Hi, bverstee.
What do you mean - "Vision is not working at all" ?
Funny video - to place IC without vision ))). did you check the calibration of vision systems ? what comp threshold did you set, why you do not use the marks ... etc
-
Did a full run after lots of tweaking and tuning today.
It looks like besides the larger component trays there is a bottom up camera but it is not used at all?
Shouldn't the machine pick the large part and check position and rotation on that camera and then correct for any off center pickups before it places the component?
The 'vision' (IC detection algoritm) didn't work at all with my tests (it didn't properly detect the parts dimensions) so I disabled it for now.
Will look into it later.
Hi, bverstee.
What do you mean - "Vision is not working at all" ?
Funny video - to place IC without vision ))). did you check the calibration of vision systems ? what comp threshold did you set, why you do not use the marks ... etc
I just added Vision and Mark.
The problem was that with default settings and whatever Comp Treshold I tried, the vision system didn't work - the part dimensions weren't detected (all kinds of rectangle boundaries on the screen).
I had to change the Angle Offset to make it working, but after that it wasn't set to the correct angle.
The Angle Offset was -0.2 by default, but the Vision system only started working when I set it to -0.1 or higher.
I also measured that the Angle Offset actually should have been somewhere in the neighbourhood of -0.8, but that certainly made the Vision system unusable.
But it works with -0.1, and with a slight angle adjustment to the parts that need it, it's perfect.
Watch the video:
https://www.youtube.com/watch?v=5gl4fyimEDg (https://www.youtube.com/watch?v=5gl4fyimEDg)
-
Let's break the silence: time for another 'instructional' TVM802A video.
https://www.youtube.com/watch?v=wkhYdwlRGKk (https://www.youtube.com/watch?v=wkhYdwlRGKk)
-
And two more videos with some examples:
https://www.youtube.com/watch?v=FSOqoW6JPEk (https://www.youtube.com/watch?v=FSOqoW6JPEk)
https://www.youtube.com/watch?v=5nmLLkywIVM (https://www.youtube.com/watch?v=5nmLLkywIVM)
-
Hey bverstee,
thank you very much for all your nice and candid information and videos! I have a TVM802BX and just recently started to seriously tinker with it. It also took me quite some time to figure out how to calibrate it properly and get most things working - at some point I want to write up the calibration process and my findings.
BTW - I asked the manufacturer for their latest software and this is now at V3.22 which finally adds vision aiming for the IC trays - very handy!
I think by now I have most things figured out and am in the process of creating a Python program to create CSV files for the machine from KiCad projects. I have figured out most data fields from the TVM802 CSV files and also found that most of the non-component fields can actually be omitted, which is nice! Then you do not have to replicate the stack configuration all the time. Now I need to think about how to best handle stack configuration and some kind of stock keeping and map this to pick&place projects. Will see...
I also found that the PCB mark bitmaps are actually raw 8 bit gray scale base64 encoded bitmaps, cool! What I have not figured out yet though is how to calculate the resolution, which seems to be a function of the mark size which is selected in mm in the software. Since the raw bitmaps do not seem to have any embedded information about their pixel resolution it must be calculated from this size setting? Oh well, some time someone or me will figure it out... ATM I use in KiCad fiducial "components" in the board design and the Python program will automatically assign the coordinates to Mark1 and Mark2 if it finds components FID01 and FID02 in the component list. Makes alignment of the PCB lot easier.
But more importantly I still have a major issue with the up facing camera. The brightness / contrast of the picture is very very low, you can barely see the pins of the chips and thus the vision system has a hard time to detect and measure the part. By playing with the component threshold I can get it to accept the part but the result is more like a lottery, sometimes it's marvelous, sometimes it's just a bit off and sometimes it's totally wrong. It is not usable in this way.
I got myself the TVM802BX machine, i.e. the version with two reel stacks and the embedded PC built into it. Since I usually only use Linux I did want to have to bother to get and setup a Windows machine. So now everything is contained nicely in the TVM802. The PC they built into it is based on an Intel N2800 CPU, which is just enough to drive the machine - exactly what I wanted. Why I write this is because I found a couple of JPEGs on the SSD of this PC:
[attach=1]
[attach=2]
These look like some kind of picture quality setting from some device built into the machine? You can also see the up facing camera being seen from the down facing camera. If I could change these settings, like brightness, contrast, saturation etc. then I think the up facing camera could become useful. Or some other trick? I already tried to search for a tool for the STK1160 grabber built into the machine to maybe adjust the picture, but did not find anything useful. The machine also came with installers for some DVD/TV grabber software. These of course allow to change settings but these get reset once the SurfaceMount application starts.
Apart from these problem I am pretty happy with the machine, exactly what I was looking for and totally worth the buck :-) Only construction flaw is the back side reel feeder. The excess tape of the left side feeders is fed into a slid in the machine table and turned around so that it comes out from under the feeders again. You can run a hole reel without having to mess with the excess tape. But the back side feeder pushes the tape below the PCB holder and over the hole table out to the front. If you have an additional tray and such mounted there it will get messy. I have seen similar machines which attach the second feeder stack to the right with the same slit and tape turn around. Oh well, a minor annoyance.
Just the other day I also finally ordered a T961 small reflow oven completing my workshop :-)
Oh, BTW, this here is my (start) of my KiCad TVM802 project:
https://github.com/nica-f/tvm802-mdgen
I need to update it soon once I made up my mind how to handle the component<->stack assignment.
Looking forward to sharing more experiences and wisdom about the TVM802 :-)
Cheers
nicole
-
BTW - I asked the manufacturer for their latest software and this is now at V3.22 which finally adds vision aiming for the IC trays - very handy!
Omg finally. had reqested this ages ago!
But more importantly I still have a major issue with the up facing camera. The brightness / contrast of the picture is very very low, you can barely see the pins of the chips and thus the vision system has a hard time to detect and measure the part. By playing with the component threshold I can get it to accept the part but the result is more like a lottery, sometimes it's marvelous, sometimes it's just a bit off and sometimes it's totally wrong. It is not usable in this way.
Strange. I have basically all thresholds on 50 and it works. Reliability is improved when turning off the lights in my workshop. If you have overhead lights it also helps to build a small cardboard 'shield'.
Apart from these problem I am pretty happy with the machine, exactly what I was looking for and totally worth the buck :-) Only construction flaw is the back side reel feeder. The excess tape of the left side feeders is fed into a slid in the machine table and turned around so that it comes out from under the feeders again. You can run a hole reel without having to mess with the excess tape. But the back side feeder pushes the tape below the PCB holder and over the hole table out to the front. If you have an additional tray and such mounted there it will get messy. I have seen similar machines which attach the second feeder stack to the right with the same slit and tape turn around. Oh well, a minor annoyance.
Agreed, but it's manageable. My main issue here is access to the back feeders (I have the machine against a wall due to space constrictions)
-
BTW - I asked the manufacturer for their latest software and this is now at V3.22 which finally adds vision aiming for the IC trays - very handy!
Omg finally. had reqested this ages ago!
Sadly they do not have a public download for it, so if you still have a contact to them I'd recommend asking them. They are very responsive via Alibaba, which is where I got the machine from, directly from QiHe.
But more importantly I still have a major issue with the up facing camera. The brightness / contrast of the picture is very very low, you can barely see the pins of the chips and thus the vision system has a hard time to detect and measure the part. By playing with the component threshold I can get it to accept the part but the result is more like a lottery, sometimes it's marvelous, sometimes it's just a bit off and sometimes it's totally wrong. It is not usable in this way.
Strange. I have basically all thresholds on 50 and it works. Reliability is improved when turning off the lights in my workshop. If you have overhead lights it also helps to build a small cardboard 'shield'.
Yeah well, I guess they also use different cameras... my machine came with a defect LED light for the down camera, some LEDs did not work. Instead of just the ring PCB with the LEDs they sent me a completely new camera assembly! And this new camera is much brighter and has more contrast than the original one.
Anyway, yes, I started out with the default of 50 and it kind of worked. It worked a bit better for a LQFP100 but also there every once and again I have misalignments. With a QFP32 it gets even worse which is, I think, due to the construction of the lighting, which has most illumination towards the outside and less towards the center. I am in contact with QiHe about this problem, let's see what they come up with. And yes, I have switched off all light in the workshop room :) I think I'll also get myself some matte black paint and paint the bottom side metal / aluminum parts of the head to avoid reflections.
BTW - the cameras in my machine, as far as I can see it, are regular video cameras outputting a composite video signal which is then fed into the controller PCB (the one with the Ethernet) and get switched there to the USB framegrabber, which in my machine is a STK1160 (or STK1120, not sure) based grabber USB stick - so not the XBOX cameras anymore which they used before. I don't know if this is an enhancement or regression :) A proper digital camera with more resolution than just a PAL or NTSC picture should be better?
Apart from these problem I am pretty happy with the machine, exactly what I was looking for and totally worth the buck :-) Only construction flaw is the back side reel feeder. The excess tape of the left side feeders is fed into a slid in the machine table and turned around so that it comes out from under the feeders again. You can run a hole reel without having to mess with the excess tape. But the back side feeder pushes the tape below the PCB holder and over the hole table out to the front. If you have an additional tray and such mounted there it will get messy. I have seen similar machines which attach the second feeder stack to the right with the same slit and tape turn around. Oh well, a minor annoyance.
Agreed, but it's manageable. My main issue here is access to the back feeders (I have the machine against a wall due to space constrictions)
Yes, the feeders in the back are a bit hard to access. I also do not think I will use them often but I thought, if I buy such a machine I'll go for the fully loaded one, you never know :-)
Would be agonizing finding out later that for a super cool project you would need just these two more feeders!
Tonight I took off the main table and took a couple of pictures of the inside of my machine, will upload these in a bit... pretty interesting. The controller PCB uses a STM32F107 :)
Cheers
nicole
-
And here come the pictures of the inside of a 2019 TVM902BX - sorry for the partially not so great quality, I had to hold the table (which is REALLY heavy) with one arm while placing the light and taking pictures with my phone on the other arm :-)
[attach=1]
This is the small embedded mini ITX PC - Intel N2800 based, 2GB DDR3 RAM, small 32GB M.2 2244 SATA SSD. It is attache with the blue ethernet cable to the machine controller and the red ethernet is my connection to my LAN. Beneath the blue ethernet cable you can see the black USB frame grabber for the cameras. Its input cable is attached to the machine controller PCB.
[attach=2]
A bit better view of the embedded PC and the machine controller board.
[attach=3]
And a bit closer...
[attach=4]
First try to get a picture of the machine controller *meh*
[attach=5]
Ah, here we are, a bit better :-) What you can not read is the chip marking on the controller's CPU - STM32F107
[attach=6]
A set of three power supplies.
[attach=7]
An overview...
[attach=8]
Slightly better overview...
[attach=9]
Again the PC mainboard.
[attach=10]
Close up of some of the PCs connections and the frame grabber.
[attach=11]
Another shot of the machine controller board.
[attach=12]
Another try...
[attach=13]
And again... :-)
[attach=14]
The pumps - there are two, I think one is pressure and one is vacuum.
Oh well, not much to learn from here. But I am pretty happy how this machine is built, pretty solid and clean, all wires are properly connected and kept into place.
Cheers
nicole
-
BTW - the cameras in my machine, as far as I can see it, are regular video cameras outputting a composite video signal which is then fed into the controller PCB (the one with the Ethernet) and get switched there to the USB framegrabber, which in my machine is a STK1160 (or STK1120, not sure) based grabber USB stick - so not the XBOX cameras anymore which they used before. I don't know if this is an enhancement or regression :) A proper digital camera with more resolution than just a PAL or NTSC picture should be better?
Yes they started doing this a few years back all machines with firmware 3.xx use the new cameras with STK1160. Seems to be sufficient though, I can place 0.5mm pitch well (If I don't forget to pray to the right gods and perform a voodoo dance)
Overall the reliability is borderline sufficient. One day I can place boards just fine, the next day it's never ending hiccups and faults.I still ( after 2 years) have not figured out how to calibrate the part and pick positions correctly. It just does not make sense and does not work.
-
This is my machine by the way. You can see the tapes of the back feeders are not really an issue..
I added a joystick for better manual mode control
The vibration feeder works well, to. You just define an 'IC Stack' with 1*1 and it will always pick the same location.
-
Oh, very cool!
I wasn't aware that QiHe also offered a vibration feeder, I am missing that clearly :-)
Meanwhile today I partially solved the camera mystery. The cameras used are, according to their markings, 1080p (how do they output 1080p to composite video!?) Sony cameras. The more interesting part is that the small maybe 4x4cm PCB not only has a 3 pin connector for GND, +12VDC and the composite video signal, but there is a second header with 6 pins - hmm! On the PCB next to the header are markings with "G U D L R M" - what could that be? Well, turns out, this is GND, Up, Down, Left, Right and Menu! These small cameras have an OSD (On Screen Displlay) menu integrated which, you may have guessed it by now, allows to control the image taken! Like brightness, contrast, exposure mode, white balance and a couple more.
So, if your back camera is too dark and delivers too little contrast to properly detect parts just connect a button pad with these five buttons (each pulling to GND when pressed) to this connector and fine tune your camera picture! The problem is just that the camera is a bit tough to access. At least in my machine the connector is exposed and accessible for the up facing camera - not so much for the down facing camera where the connector is removed completely. But my down facing camera isn't a problem anyway, so that's OK for me.
No I have to find a matching connector and build myself a cable - I made a keypad today and tested it with the old down camera and this worked super nicely! And the cameras are actually pretty good, they deliver a nice and sharp picture, when you unleash them from the brutal settings they came with. The OSD can BTW also be set to English (by default it's Chinese, of course).
Getting there!
So I think there is serious hope to get the up camera to behave.
Cheers
nicole
-
Cool info about the cameras. Hope I won't need it though ;)
The vibration feeder is not from QiHe and I don't know if they even sell them. Just a random one from aliexpress
-
Yesterday I finally received the camera keypad PCB from QiHe and after some playing around I was able to increase the contrast a lot! The camera now produces a pretty clear vision of the chip pins/pads which helps to make the vision system much more robust! The jitter in the part detection (the red rectangle) is less and I could increase the component vision threshold again which helps a lot with precision.
What I also had to do was to adjust the "distance per pixel" in the sys settings since the measurements of the camera were a bit off all the time. Now I get readings which are just a few 1/00mm off, which of course helps a lot with the positioning.
What I still find a bit annoying is the speed of the up vision system. It takes ages for it to adjust the chip which is IMHO caused by the system to try to adjust the part position while hovering over the camera. With the latency of the vision system this ends up in a lot of back and forth with hysteresis efffects and so forth. Takes ages and I think this is the wrong way to do it. What one should do instead is to move the part over the camera taking the nozzle as the center and then to just measure the part as is and not trying to correct its position already - only measure size, position and angle. Then take these values as correction values for placing the part. This should be a lot faster and this is AFAIK also the way how bigger machines do it. If you watch videos from other machines these just speed over the camera quickly snapping just one picture and then move on to placement.
So I went ahead and suggested this to QiHe and they responded that I should try the "fast vision" mode - WTH!? Cool! I have not seen such an option yet anywhere in the software. They will tell me tomorrow *meh* how to enable it and I am frankly pretty excited! I'll let you know once I know how to enable it.
PS: If you are also struggling with setting up your machine and all the calibration necessary, one quick tip: Make sure you have some adhesive on your PCB when trying to place parts. I once made the mistake not having it and lost almost half a day with total frustration because regardless of what I adjusted the outcome would be just like a lottery. After applying some adhesive to the PCB surface - tadah! As adhesive I use a 3M double sided tape which uses the same glue as on PostIt notes, i.e. easily removable and reusable. Not cheap that tape but very handy for this purpose and you won't need much of it anyway.
Cheers
nicole
-
*doh* ! The "fast vision mode" they mentioned is the "Quick" setting instead of "Accurate" when choosing the per component vision mode. Oh dear, I need more coffee. And this mode exactly behaves as I described and wanted it :-) Now I just need to check if it is just fast, which would be great, or also indeed less accurate and if, how much less accurate.
BTW: I am playing around heavily with the vision parameters. There is a general vision component threshold in the systems setting but you can also set a component threshold per part, which I do not fully understand. Which one gets applied when? This parameter is pretty important since it determines the required "contrast" in the vision picture for the outline detection. Setting it too high will result in the part not being detected at all, setting it too low may take anything (even a reflection or the bends of the pins) as an edge.
Getting there!
(I think at some point I will write a posting with the calibration steps... this learning curve can be avoided a lot by some documentation...)
Cheers
nicole
-
Hi
Has anybody found an *easy* way to get software / firmware / whatever updates for these machines? The whole "send them an email and see what happens" process has been very hit and miss over the years .....
Thanks
Bob
-
No, that's also what I did the last couple of times, asked via Alibaba messages (I do not even have an email). On Ali they are pretty responsive and I usually got a download link within a couple of days - latest version I have now is V3.22 (this is for the machine version with just one USB video grabber).
BTW - I have started to document some (hopefully) helpful things in WiKi:
https://source.dpin.de/nica/tvm802-documentation/-/wikis/home (https://source.dpin.de/nica/tvm802-documentation/-/wikis/home)
I spent quite some time figuring out the calibration and wanted to share it since it really makes a difference - once done right it just works like a charm :-)
Cheers
nicole
-
Hi
Ok, I don't even remember when the last time I actually got a reply was. I'll try the messages thing on Alibaba this time.
Thanks !!
Bob
-
Hello everyone!
Please provide a link to the latest software for TVM802b with one video input.
At the moment I only own version v3.15, I see some have version 3.22.
Thank!
I gathered here several software versions 2.37, 3.05, 3.06, 3.15 dixom.ru/Software/TVM802B.rar (http://dixom.ru/Software/TVM802B.rar)
-
·············· Notice The Red Box！···············
( The new version software is modified on version 2.32 )
V3.23 ----- 2020/4/22
1、 Modify the BUG about “an error occurred when setting the coordinate parameter to IC stack”;
2、 Modify the BUG about “ the descending height of Z1 and Z2 Axis are inconsistent”;
3、 Add the maximum allowable range of position offset value during vision measurement;
4、 Optimize the speed of the reel.
V3.22 ------ 2019/11/26
1、Modify the function that the coordinate of IC Stack in the SysConfig can be set and moved with down vision.
...
-
Hi all,
I have a TVM 802A, which was working fine so far, with all limitations of this low cost PnP. By the way, the worse part of it is the feeder, which is a pain in the ass. Some days ago, while it was working on a small production batch, it started having a very strange behavior. It's impossible to home the head and while positive and negative limits are detected, I cannot even manual move the head in some directions. Please watch the video https://youtu.be/CucdKt0oZNE. I contacted the support of the company and I got this reply:"Hi sir, It seems the wire of the motor is broken.It needs to be replaced." It's well known that in this machine there is only one motor and only one wire. When I asked again of which motor, I've been told that it's the X axis motor, which by the way, as it can be seen in the video, if the head reaches the negative limit, it works fine. I've checked all wires with a multimeter and everything is well connected. It's well known that the price to pay for a cheap Chinese machine, is the absence of after-sales support.
All motors and limit switches, under certain conditions work fine and thus I suppose that the problem is related to firmware corruption or fault. Unless someone else was having the same problem and a solution, I suppose that the only alternative is to buy another PnP, staying away from Chinese no supported products. After all, in the end, you end up paying much more for the "low cost" products.
-
I have a TVM 802A, which was working fine so far, with all limitations of this low cost PnP. By the way, the worse part of it is the feeder, which is a pain in the ass. Some days ago, while it was working on a small production batch, it started having a very strange behavior. It's impossible to home the head and while positive and negative limits are detected, I cannot even manual move the head in some directions. Please watch the video https://youtu.be/CucdKt0oZNE. I contacted the support of the company and I got this reply:"Hi sir, It seems the wire of the motor is broken.It needs to be replaced." It's well known that in this machine there is only one motor and only one wire. When I asked again of which motor, I've been told that it's the X axis motor, which by the way, as it can be seen in the video, if the head reaches the negative limit, it works fine.
What happens at 00:46 looks like a broken limit switch, you are pretty much crashing the machine hard into the side railing. Since the machine is not properly homed it most likely stops on soft limits after that, so you can't move it anymore. Check the if the limit switches still work with a multimeter, and check the solder joints on the limit switches.
-
Thanks ttt for your reply. I've checked all limit switches and all are ok. For example, when I manually move the head, near the X left or right limit, it crashes. If it's a bit away from the limit and I insert something in the limit switch while manually moving, it immediately stops and a warning about limit reached pops up.
As I wrote in my first message, all motors and limit switches are working manually under certain conditions. The only way it doesn't work is when pressing the up arrow to move towards the normal homing position. Also on X axis it moves on both directions only when the negative Y limit has been reached.
-
As I wrote in my first message, all motors and limit switches are working manually under certain conditions. The only way it doesn't work is when pressing the up arrow to move towards the normal homing position. Also on X axis it moves on both directions only when the negative Y limit has been reached.
IMO it'll be very hard for anyone to understand your problem unless they've had the exact same issue. :-(
To me it sounds like it could be several issues:
- Configuration/Setup issue, for instance something in "System Config->Other" might be wrong.
- Cold solder joint or broken wire, either in motor wiring or limit switches. If the problem is intermittent this is most likely your problem.
- Corrupted firmware
Personally I had an issue with one of the limit switches on the head for which I had to completely dissemble the head. It turned out to be a cracked solder joint (it was the upper pin in the attached picture).
This is the downside with cheap machines from China: no local support. So you either spend the big bucks or reverse engineer a machine to fix it...
-
In my machine all limit switches are light switches. The problem started while it was working and it's consistent. I opt for the firmware
corruption, but without support from the manufacturer, there is nothing to do, other than buying another one , of course, no qihe made.
Is there any member of the forum with some experience of a manufacturer with better after-sales support?THX
-
No experience wth this specific machine, bu experience with XYZ plane CNC machines.
What normally happens is that you home your machine, the axis move the correct position towards the home sensors and when they are reached the software sets the machine coordinates on 0,0,0 or another predefined value.
From that point on the software limits (maximum limits the axse can reach before crashing into the sides) are valid and they are software limits so can be corrupted.
So you should check if the machine homes correcctly, meaning:
- axis move towards the homing sensors
- after reaching the homing sensors set the position value to the correct value *(check!)
Then check te software limits by manually slowly runnin the axis to their max position, write down the position and check in the software if these are correcct.
If after this the machine still crashes now and then, then you should check:
- #steps per mm travel , are these still correct
- your belt tension if the tension is too low it might happen that there are steps lost.
Good luck.
-
Thanks for your suggestions Kjelt. Unfortunately, I've checked everything, connections, motors, limit switches etc. I'm convinced that the problem is corrupted firmware, but without the manufacturer's collaboration and support, I can do nothing. I've got an idea though, without the upper cover and without all electronics and filled with good quality soil, could be used as flower pot. A bit expensive, but that's the price to pay for unsupported Chinese products.
-
Plan B replace the controls, keep drivers and motors and try to make openpnp work
-
Thanks for your suggestions Kjelt. Unfortunately, I've checked everything, connections, motors, limit switches etc. I'm convinced that the problem is corrupted firmware, but without the manufacturer's collaboration and support, I can do nothing. I've got an idea though, without the upper cover and without all electronics and filled with good quality soil, could be used as flower pot. A bit expensive, but that's the price to pay for unsupported Chinese products.
Did you try to ask them to check the mainboard and if you can send it to them?
We've had a defect microscope PCB two years ago, the item burned after around 8 months and we had no problem to get a replacement (even for free), no problems with it anymore since then (it was one of those Eakins microscopes - really good customer handling).
We didn't even ship the damaged PCB back to them just took some photos and videos what didn't work.
-
Machine is bought from QiHe or through the XYing? Just ask nicely from QiHe, I'm sure they will at least reply.
-
I recently got a new main board from QiHe, because my TVM802A was disconnecting the ethernet.
I think they would be better off with selling the machine to be used with OpenPnP, because the hardware is ok, but their software suck.
-
The machine has been bought from XYing and I've asked nicely for support. They replied, but the reply "change the X motor wire" isn't support. Anyway, I replaced the cable of the X-axis but the homing problem still exists. It seems to me that the machine has lost coordinates. Is there any way to reset it? I've noticed that every time I restart the machine, it starts showing X=0, Y=0 regardless of the head's position. I don't remember if this is normal, but I think it's not correct. In settings, the correct coordinates, 326,348 are set.
And a general comment. I was using version 2.37. Entering this excellent forum, I realized that the latest update is 3.15. It's incredible that I have to send email to this company and ask if there is any software update. All companies inform their customers, directly or in their web site, when there is a new update. Has anyone managed to use TVM802 with OpenPNP?
-
Yes, i had one with override switch (XOR) and grbl controller 6 axes. Its not difficult,
I have sold it and bought chm-T36 because of juki nozzles.
-
All companies inform their customers, directly or in their web site, when there is a new update. Has anyone managed to use TVM802 with OpenPNP?
Consumer companies do. Its a much more complicated picture with SMT equipment, some charge significant fees, some might upgrade during an on-site service, some make almost no updates anyway, some make it freely available and others charge token sums because you get new features.
-
Hey guys, anyone got any TVM802A (or B) images of the guts and connectors inside the unit?????
RichardS
-
These are part of the guts. Hope this helps.
-
One word
Speghetti
:-)
-
As usual with many Chinese products. |O
-
As usual with many Chinese products. |O
Flowchart installer:
Does it work ?
No: keep adding solder blobs to the joints and pull wires to see if you missed one solderjoint somewhere
Yes:
1) Put on huge blobs thermal glue or other sticky fasteners so wires won't come loose during transport
2) SHIP IT!
-
I'm still struggling to repair my TVM802A. As soon as I finish, I'll tell the long story which started 4 months ago :palm:. Now I have a software question. My machine is running v2.37 and cameras are ok. If I try version 3 and up, cameras are not working and there is no way to select them. Does anybody know how to enable the cameras?
-
I'm still struggling to repair my TVM802A. As soon as I finish, I'll tell the long story which started 4 months ago :palm:. Now I have a software question. My machine is running v2.37 and cameras are ok. If I try version 3 and up, cameras are not working and there is no way to select them. Does anybody know how to enable the cameras?
You cannot run ver. 3 on ver. 2 hardware - v.3 uses another camera and controller ...
-
End of Qihe story. I asked to buy a new driver board and they replied: Sorry, because your machine version is too old, we no longer have the old driver board.
Qihe's PnPs die young. My machine was only 3 years old. If every 3 years I need to buy a new Qihe, it turns out to be veeeery expensive. |O
-
Can't you repair it ?
-
I have already lost a lot of time and money trying to repair it. I decided to buy a new one. Most probably I"ll sell it. It doesn't rotate nozzle one and accurate
placement, most probably because nozzle one doesn't rotate. I could try another driver board but no time available.
-
The same thing happened to me with 2 boards, nozzle 1 did not rotate, the rest was fine, luckily they did send me another CPU the last one was 6 months ago.
In three years I have changed 2 motor drivers and 2 CPU boards, the machine is of very poor quality, I also had other failures that I fixed, such as a broken cable and electronic failures in the head above.
I decided 6 months ago to change and update p & p, an SMT550, it has nothing to do with the TVM802B, the SMT550 looks like a solid, fast, accurate machine, I am very happy with the change.
-
I'm about to order a ZB3245TSS. It seems a good quality PnP, they are having a good customer service and it's within my budget. They sent me the user's manual before ordering the machine. From QiHe I've got the manual 2 years after purchasing the TVM802. Has somebody used this PnP? Any comment would be greatly appreciated.
-
Hi,
Have you considered to repair the TVM802 by replacing the control board with a smoothieboard, add new higher quality cameras and use OpenPnP?
Bjørn
-
Hi Guys,
Those are using TVM802A/B, please help me to fix two issues I'm facing with TVM802B machine which I just recently bought. Their after sales technical support seem to be not that great.
(1) TVM802B Fail to Drop the Chib
https://www.youtube.com/watch?v=DAK_gxsW2bE (https://www.youtube.com/watch?v=DAK_gxsW2bE)
https://www.youtube.com/watch?v=-ImbZVneufI (https://www.youtube.com/watch?v=-ImbZVneufI)
I used 0805 chib, it is not being dropped on the board even with solder paste on the pad. I believe the blow is not happening. When the nozzle drops the chib, and when the vacuum is disabled, there is small pressure is holding the chib on the nozzle around Pressure=6, this causes the nozzle not able to release the chib.
With using strong sticky pad (double sided tape), it seem it works.
(2) TVM802B Prick Intermittent Fail
https://www.youtube.com/watch?v=TluhnMNQ1-0 (https://www.youtube.com/watch?v=TluhnMNQ1-0)
https://www.youtube.com/watch?v=dXB8YPLOYDU (https://www.youtube.com/watch?v=dXB8YPLOYDU)
Suddenly the prick is fail to function. I can hear small sound whenever press the prick button on the SW but the notification indicator on the SW is not lighted up.
To temporary fix this, I need to power cycle the machine and SW for few times.
I really appreciate all of your advice.
-
It looks like a CPU failure, they should send you a replacement CPU to test.
This machine is a disaster, I do not recommend its purchase to anyone, the construction quality is terrible and it is full of bugs, I have failed in 3 years the wiring (a cable broke from the movement), a connection of the motor to the shaft, 2 motor driver boards, 3 CPU
The board in the head also burned and I had to repair it and change the capacitors.
Now I have it repaired and functional, but I bought a superior model and I no longer use it.
The experience with TVM802B has been terrible.
-
help me!
Please provide a link to the latest software for TVM802b.
At the moment I only own version v3.15, I see some have version 3.22.
Thank!
-
help me!
Please provide a link to the latest software for TVM802b.
At the moment I only own version v3.15, I see some have version 3.22.
Thank!
i have 3.26 , why you did not ask directly Qihe ?
-
Hi,
I own a tvm802b.
I run the software on a laptop under w10w.
I got a lot of problem for configuring the stacks,
mostly prick problems: it drag too much, or it can’t drag(plastic tapes)
what a nightmare. I also own a Charmhigh 36va ans finally
despite the fact that the sw is even worst than the 802 I finally
got less problem with feeders on the 36.
Do some people report that when a drag issue occured(the head drag
and the prick stay stuck) lead to a shift in X of pcb fiducials(real pos) ans also
feeder stack pos?
At a moment I used git to store screnshot of configured panel(panel is 10pcb with 2 fid)
but it’s worst, git doesn’t mix well with the software.
Should I switch to win7? >> I just switch on W7 and have a try.
-
On the back side of head there is a screw to regulate presure, please test if it is correct
Best regards, Luk
-
Hello guys, can someone give me software for TVM802B? Im email to vendor, but it still in silence for some days.
-
I buy one and made modifications, as on pictures.
Best regards, Luk
-
Another one...
-
And another one
-
Before final
-
And the last one :)
-
Hey guys,
recently we bought a TVM802AX, it would be arriving soon.
I want to know about its good and bad. What should be the first thing i should change or be worried about.
Anything that you can share would be great
TIA :-+
-
I have TVM802B pick and place machine. And i have a problem. Machine gets component from left stack, than goes to place it on board, but nozzle doesn't going down to PCB (it goes half way down) and doesn't place the component. On next step machine goes to discard position and drops the element there. Nozzle positions are calibrated and there is no change.
Please advice me
-
Place a movie, what's going on, and place foto with parameters.
Best regards, Lukasz
-
FYI I just received software V3.31 for my TVM802BX from QiHe... I did not see major functionality changes, a changelog was also missing. They did add a user interface language switch button to the user interface and also started to implement what seems to be some kind of user management, you can switch between "Operator" and "Manager" but funny enough you can not configure these users - tse...
Will test this further later.
Some weeks ago I did a job with the machine and had quite some issues with plastic tapes in the feeders, these black ones, made from thin plastic. The pressure of the spring metal things in the feeder base were way too strong and held the tape so tight that the head was not able to pull / feed them. So I took the spring metal out and bent it to loosen up the tension, after that it worked better, but still not ideal. The thin plastic tape often tilt and twisted a little which made it getting stuck. Need to test this more when I get more of this type of tape.
BTW did anyone try the TVM802xx with other software than the one from QiHe? Like OpenPNP or such? Last time I checked OpenPNP support was discussed but not implemented.
Cheers
nicole
-
hello
I want to place the ICs precisely as in the video. but I couldn't make it. How can I do this fine tuning? can you help me. What adjustment could have been made in the video?
The ic is constantly inserted incorrectly.
skewed or not fully on target.
The chips are checked with the down camera but the problem persists.
https://www.youtube.com/watch?v=TwvnQStELvU (https://www.youtube.com/watch?v=TwvnQStELvU)
-
So, first then let me ask a couple of questions...
How do you try this? Do you use a bare PCB and "dry" place the chip? If you do then this will not work, I tried that too and failed. You need something on the PCB that the chip will stick to or the nozzle will shift it in a random direction while retracting. I am using a pretty transparent double sided sticky tape on a test PCB, that works well for placement testing.
If you have something sticky on the PCB and still have false placement results, are the errors you see consistent? I.e. is it always the same error for the same component on the same place?
If yes, then most likely your nozzle and/or camera calibration is off - and this is the really painful part. Quite naturally the machine needs to know where the camera positions are in relation to the nozzles and eventually also a rotation angle correction. The calibration for this can take a while and you need some tools for that.
First of I would recommend either finding some PCB or printing something with a printer that looks like a bullseye. Then attach the two smallest nozzles you have, put the bullseye on a fixed spot on the machine (so it can not move) and then aim the center with the down camera. Now enter machine settings and there nozzle settings, select "measure" (IIRC, I am writing this from memory). Aim the bullseye with the camera, set this as camera reference and then aim the center with the first nozzle. This is now very painful since you can not move the head while the nozzle is down so you have to aim, lower nozzle, up nozzle, correct, lower nozzle etc. until the nozzle lands perfectly in the middle. Now do that for both nozzles and save the results.
Now your machine knows where the down camera and nozzles are relative to each other. For parts placements without using the up facing camera this should now work in regards to position. You can try it out with a known layout (e.g. the test PCB that came with the machine), select a component and right click "aim with nozzle" or "aim with camera". Nozzle and down camera should land in the center of the component.
Now for the up facing camera. There is also a calibration menu for that. What you basically have to tell the machine is where the center of each nozzle is in relation to the center of the up camera. In my machine the nozzles are black so you actually do not see much of them in the camera view. But what I found to be useful are the values in the camera view, i.e. the offsets for X and Y while lowering the nozzle all the way down to the camera until the nozzle touches the glass. Now the camera logic will try to detect a "component" and draws a rectangle around the nozzle. The offset values now tell you how far the detected center is away from the camera center. Bingo! Exactly what you want. Now up the nozzle again, move the head and down the nozzle until these values are as small as possible.
Save and you are done!
Next you need to look at the IC tray and feeder settings as well as the component settings. You can have rotation offsets in there as well as vision thresholds. If the rotation is wrong your part will be assumed rotated and lands wrong. If the vision thresholds are wrong the margin of error for positioning the component can be either too high (aka "anything goes") or too low and vision will never complete correcting the position (and eventually abort and discard).
The calibration could IMHO be done a lot smarter and much more automated by the machine. But well... it can be made working with some patience - take your favorite hot or cold beverage, sit down, relax and be patient while teaching your poor little machine, it will be very grateful by placing your parts correctly next time.
Hope this helps?
Cheers
nicole
-
There is an easier way to get the offset cameras, I have used this on some other pick and place, the principle works.
I put a little bit of blutak or some other soft material on the board, plasticine for example ... anything soft where the nozzle will make a mark.
Lower the nozzle until it touches the board and makes a mark in the blutak ... raise the nozzle (you might have to clean it!) and now you can switch to the camera and just centre it on the mark. Very quick, very easy.
-
Oh, indeed, that is actually pretty clever and avoids parallax effect errors. Just the question which kind of material to make the mark in? I am also a bit worried about the residue on the nozzle.
Hmm... maybe create $something with a hole in it which pretty much exactly fits the nozzle, move down the nozzle into that hole, fix the material in its location and then move the camera to the hole's position? But anyway, yes, doing it the way around seems pretty clever, indeed :)
Cheers
nicole
-
I have used a similar scheme to find the nozzle offsets on my Quad machine. I have "spotting dye" here for use in the machine shop. It is similar to artist's Prussian blue oil paint, but washes off more easily. I just put a tiny smear of the dye on my finger and wipe on the nozzle. Then, press the nozzle down to the board and raise, and it leaves a round print like the letter "O". The dye wipes off easily from the nozzle.
The actual dye I use is made by Canode, but is not easy to get. The dye leaves a much more visible mark on shiny steel parts of the machine, but will work on a PC board, too, with less contrast.
The guy who taught me how to use the machine prefers to put "Silly Putty" (a children's toy item) on the board, and the nozzle will leave a round ring in the putty. But, then you will have to clear the nozzle for sure.
Jon
-
Thanks ttt for your reply. I've checked all limit switches and all are ok. For example, when I manually move the head, near the X left or right limit, it crashes. If it's a bit away from the limit and I insert something in the limit switch while manually moving, it immediately stops and a warning about limit reached pops up.
As I wrote in my first message, all motors and limit switches are working manually under certain conditions. The only way it doesn't work is when pressing the up arrow to move towards the normal homing position. Also on X axis it moves on both directions only when the negative Y limit has been reached.
Hi,
I know your post ist over a year old, but I have the same machine with almost the same problem:
X-Axis does the same yerky movement when homing and doesn't succeed.
I got a replacement Mainboard from the manufacturer. Didn't help.
Today I found the Problem: One of the X-Axis homing lever was slightly bend and not releasing the button underneath. The indicator was always active on the homing screen, but i didn't notice (really dumb on my part).
So I bend it back and now its working again.
Maybe this helps
-
Update on the below message. I was able to get the English version of the software from the manufacturer.
Now I am looking for an updated user manual. Let me know if someone has a newer one, please!
Hello all, I know this thread has some age on it. I purchased this machine in 2016. Tried it out, put it away. Now I have time to get back to it.
If anyone has a link to an English version of the latest V2.xx I would be most grateful.
Thank you in advance!
-
Link
https://www.qhsmt.com/support (https://www.qhsmt.com/support)
Hello, all users of Qihe SMT :)
Qihe team have updated support on the official website to support you to quickly find the TVM802 or other models
and please contact customer service to update the software .
Also add customer case sharing section for friends who interested in the machines .
Welcome to subscribe qihe team via social media .
-
Hello, I've been working with a TVM802 for 6 months and a problem started a day ago.
Nozzle 1 or A, is not turning properly, seems to be because the stepper motor is worn out and does not have enough torque to hold the position after turning against the kink of the suction hose (it has already been replaced with a more flexible one) .
I thought about requesting the purchase of a replacement stepper, but I fear the problem will recur in a short time.
Is there any indication of an equivalent stepper motor model that has a higher torque and is compatible with this machine?
-
Olá Joao
Did the engineer help you solve the problem?
:)
-
Hey guys
Good day :-+
Qihe team joined in 2022 Ningbo International Lighting Exhibition.
We showed new 10 heads machine models such as QM10 machine ，etc.
If you are interested in learning more about us
Welcome to click the official website link or follow us via social media .
https://www.qhsmt.com/news/2022-ningbo-cnll-exhibition-qihe-july-show-59003021.html (https://www.qhsmt.com/news/2022-ningbo-cnll-exhibition-qihe-july-show-59003021.html)
-
Hello qihe_kim, we received a new engine sent from the factory, the engine was changed but the problem still happens, it seems to have something to do with the software, the version used by us is 3.35. I emailed support asking for help.
-
Hello everyone. I became the owner of the TVM802B machine and everything would be fine, but there is a problem immediately after receiving it, there was no software for working with the machine on the flash card and there is nowhere to download it either, QIHE technical support does not respond to my letters. Please share the current software, I want to start mastering the machine already.
-
I am using a TVM802B-X machine. About 3000 components installed. I noticed that when setting up the machine,
the needle for pulling the ribbon with the components does not always work. I have found that when a magnetic coil is heated,
its magnetic properties decrease.
I applied thermal grease to the coil to remove heat to the machine body, it helped.
Where can I buy TVM802B-X repair parts? I wrote to the mail of the plant and on Skype, but they do not answer.
-
PM @qihe_kim here in the forum. He is the owner of QiHe.
-
Hello everyone. I became the owner of the TVM802B machine and everything would be fine, but there is a problem immediately after receiving it, there was no software for working with the machine on the flash card and there is nowhere to download it either, QIHE technical support does not respond to my letters. Please share the current software, I want to start mastering the machine already.
I have the TVM802C and can get you the latest software (v2.39 in English).
PM @qihe_kim here in the forum. He is the owner of QiHe.
Are you sure about the owner? ;-) Seem they are more concerned about selling new machines rather than troubleshoot.
-
Hello everyone. I became the owner of the TVM802B machine and everything would be fine, but there is a problem immediately after receiving it, there was no software for working with the machine on the flash card and there is nowhere to download it either, QIHE technical support does not respond to my letters. Please share the current software, I want to start mastering the machine already.
I have the TVM802C and can get you the latest software (v2.39 in English).
PM @qihe_kim here in the forum. He is the owner of QiHe.
Are you sure about the owner? ;-) Seem they are more concerned about selling new machines rather than troubleshoot.
Yes, QiHe's owner's name is Kim. It's a relatively small company.
-
I have a very old TVM802 machine.
My software is V1.18
Can you share your version 1.22 with me ?
Or maybe do you have a version from v1.19 to v1.25
Please help me get them. My email avtogorelka@gmail.com
-
They answered me and I already received all the parts that were needed.
Now can work more confidently ))
-
Attention please
All friends
Maybe you are a customer of Qihe SMT machine, or maybe you will consider buying in the future
I hope these sharing can be useful to you, please feel free to contact us
Qhsmt software update link https://www.qhsmt.com/support/ (https://www.qhsmt.com/support/)
Qihe Product Center https://www.qhsmt.com/porducts/ (https://www.qhsmt.com/porducts/)
The news of Qihe https://www.qhsmt.com/blog/ (https://www.qhsmt.com/blog/)
In addition, share some customer cases about our latest exhibitions and latest models
QM81 model
https://www.qhsmt.com/2022/09/11/smt-pick-and-place-machine-qm81-fully-automatic-smt-line-applied-to-smart-wearable-device-prototyping-pcbs- (https://www.qhsmt.com/2022/09/11/smt-pick-and-place-machine-qm81-fully-automatic-smt-line-applied-to-smart-wearable-device-prototyping-pcbs-) and-mass-production/
QM81 model
https://www.qhsmt.com/2022/09/03/smt-pick-and-place-machine-customer-sharing-qm81-fully-automatic-smt-line-applied-to-prototyping-pcbs-and- (https://www.qhsmt.com/2022/09/03/smt-pick-and-place-machine-customer-sharing-qm81-fully-automatic-smt-line-applied-to-prototyping-pcbs-and-) led-lamps/
QM62 model
https://www.qhsmt.com/2022/12/25/qm62-automatic-pick-and-place-machine-smt-setup-pnp-automatic-production-line-operation-video-smt-pick-and- (https://www.qhsmt.com/2022/12/25/qm62-automatic-pick-and-place-machine-smt-setup-pnp-automatic-production-line-operation-video-smt-pick-and-) place-machine/
QM61 model
https://www.qhsmt.com/2022/11/10/smt-pick-and-place-machine-qm61-smt-line-video-sharing/ (https://www.qhsmt.com/2022/11/10/smt-pick-and-place-machine-qm61-smt-line-video-sharing/)
TVM926S model
https://www.qhsmt.com/2023/05/03/ball-lamp-bead-led-producing-case-by-fully-automatic-led-line-tvm926s-4-heads-pick-and-place- (https://www.qhsmt.com/2023/05/03/ball-lamp-bead-led-producing-case-by-fully-automatic-led-line-tvm926s-4-heads-pick-and-place-) machine/
-
:)
desktop pick and place machine new arrivals
TVM802B Plus SMT pick and place machine 2heads 58slots desktop pick&place TVM802B+ deluxe edition
https://www.qhsmt.com/product/tvm802b-plus-smt-pick-and-place-machine-2heads-58slots-desktop-pickplace-deluxe-edition-desktop-pick-and-place-machine/ (https://www.qhsmt.com/product/tvm802b-plus-smt-pick-and-place-machine-2heads-58slots-desktop-pickplace-deluxe-edition-desktop-pick-and-place-machine/)
-
Hello,
Using TVM802A. I've upgraded the software to v3.38 because I was having trouble configuring a PCB array.
After the update, vision is not working anymore.
Now I tried to use the old version (v3.23) and it's also not working.
The software is looking for the camera in the wrong position. When I try to set the correct nozzle camera position in the software, it tries to use vision in the correct place but can't orient the component. It just goes further to the right with each vision adjustment.
I tried to contact support but didn't get any reply.
Can anyone help?
-
...The software is looking for the camera in the wrong position. When I try to set the correct nozzle camera position in the software, it tries to use vision in the correct place but can't orient the component. It just goes further to the right with each vision adjustment.
This sounds like two different things to me, can you clarify? Camera Position is set in System Config (I believe). Nozzle to camera offset is also set there, but is a different thing.
When you "Move To" Nozzle 1 at Camera Position -do you see Nozzle 1?
I have a TVM802B but am also running 3.38.
-
Hello.
Thank you for your reply.
After several attempts, I have concluded that something went wrong during the upgrade, and config values were lost.
I have attached screenshots of my config.
Can anyone with a TVM802A please verify those values?
Thanks,
Joao
-
Also the camera config
-
Hello Kim! My friend send request for firmware update on TVM802, but still didn't get answer :-//
P.S. Problem solved, mail server supposed it is spam ???