Pick & Place MachineTVM920

[thommo]

Hi anfang,

You're doing some great investigation here - thanks for sharing it.

So it seems to become apparant that the drift/variation could be in the nozzle/head movement area. Good that it's getting isolated.

In relation to your comment "Using nozzle as reference is the same, isn't it? And it makes it easy to ask the machine where is the marked location",  I believe the answer is no. The down-looking camera is 'fixed' in its position relative to the head, the nozzles are not. If you only check the nozzles [as you have just discovered] it is not possible to isolate the area where the error is occuring. If you use the down-looking camera it will confirm if the 'head assembly' is repeating accurately, or otherwise.

O Ring
It sounds like the O Ring is slightly too large for the purpose, and is hold the part from seating correctly.

Do you have access to a supplier where you could purchase a smaller O Ring? You could also try putting some silicon on the O Ring as a lubricant. Silicon-based furniture polish would be OK I think.

Juki Nozzle
Are you saying that they vary in length within the grade/size of nozzle, or that they have different lengths for different grade/size nozzles?

If it's the former, then this would be explained by the expectation that larger nozzles are designed for larger [and thicker] components, if it is the latter, then that's probably the price of purchasing non-OEM products.

When nozzle/head height is set, is there not a provision for setting EACH of the 4 nozzles independently? Ideally, this height would be set on a per component basis to match the component characteristics.


Sounds like another item for the SW App wish list if it doesn't already exist.

Cheers - Peter

[ttsthermaltech]

I think once you get the correct nozzle, then your problems will mostly disappear. I don't have any 0402 components on hand, but I am willing to bet that the 503 nozzle is too big. Typical spacing between terminations on 0402 components is 0.7mm, A 503 nozzle is 1mm outside diameter. The terminations on those resistor look rather poor (ie, lots of variation in tinning / solder dimension), so if the nozzle cant seat properly in the flat middle section of the component, it will make a poor vacuum seal, sit cocked on the nozzle, and will likely shift during high speed movements. Just my thoughts....

Thanks though for doing all this testing. I have not had the time to do much work with mine, but I did manage to get things working really well during my initial test, to be confident in its operation, however, I only use 805 and larger comps at the moment, so a 503 nozzle is more than adequate for my needs. I am looking forward to seeing how the 502 nozzle works for your 0402 components.

[anfang]

If you only check the nozzles [as you have just discovered] it is not possible to isolate the area where the error is occuring. If you use the down-looking camera it will confirm if the 'head assembly' is repeating accurately, or otherwise.

Yes, OK understood. The question to be answered is whether the downlooking camera to nozzle relationship is changing (which I've previously assumed is static) OR if the uplooking camera to nozzle relationship is changing.

It sounds like the O Ring is slightly too large for the purpose, and is hold the part from seating correctly.

I got some shim material, but the result is the same. No matter what, the nozzle will never mate metal on metal. In the default case, the nozzle has ~0.2 gap. If I a add a thin washer, then the 0.2mm gap remains. It is the nature of the oring. Yes, maybe smaller would help.

I will over thanksgiving try to turn a new slot on the lathe and install another oring on the Juki nozzle and see what that does for centering. The good news is that the Juki and the receiver both appear extremely round. It's just the mating that needs a little help. If someone else could also note their runout it would be helpful.

Are you saying that they vary in length within the grade/size of nozzle, or that they have different lengths for different grade/size nozzles?

Yes, the length between the small nozzle (Juki 503) and big nozzle (Juki 506) is about 2mm with the bigger nozzle being about 2mm longer. And if the bigger part is also thicker, then this means a pick on the small nozzle at just touches the tape or PCB will barely compress, and then a pick on the bigger nozzle will compress a lot more (2.5mm more roughly). I don't see a place in the SW to specify nozzle length.

I don't have any 0402 components on hand, but I am willing to bet that the 503 nozzle is too big.

Yes, Juki recommends 502 for 0402 and 503 for 0603/0805. When it comes I will report if 502 can be used for 0402 AND 0603. That would be ideal for me.

[IconicPCB]

What does the manufacturer specify for bearing slideway lubrication?

Are your bearings crrectly lubricated since poorly lubricated bearings may introduce a bit of stutter in transport.
If the manufacturer does not specify the grease or oil ensure the lubricant shears smoothly without sudden slip.

[gameguru]

Juki Nozzle for different SMD Parts

[thommo]

Hi Anfang,

Have you received and tested those smaller Juki nozzles yet?

Did any of those suggestions improve the output result? I think that @ttsthermaltech's comments opened up the issue most pointedly. If you can't get good vacuum contact with the component, it's unlikely it'll stay in place until it is placed on the paste.

Please let's know your outcome - Peter

[glenenglish]

I received my TVM920 on Saturday. The quality is very very good ! The finish is at least as good as any machine I have seen.

Packing and accessories,  kits etc all excellent. 1st class. A few screws a little loose.. They  need some thread-locker / spring-washers.

But in general, excellent. I can recommend this machine it is excellent value for money.

Small problem , now resolved : My genuine Yamaha feeders will not attach to the feeder plates. The feeder plates are too thick..
The Chinese copy feeders will attach but not securely - the attachment handle can be pushed home but only with some deforming of the mechanism. The genuine feeders are heavy and do not bend...

The feeder plates measure 17.17mm thick, at the inner  edge.
Yamaha Feeder plates must be 16.0 mm thick at least at the inside edge, 15mm deep.
I have measured my own Yamaha YVL88-ii  for  'the truth"

QiHe have said they will replace the out of spec feeder plates no cost .  Very good QiHe !

photos: attached

[RobK_NL]

Things are going fast all of a sudden. I received mine on Friday.

Will measure feeder plate on Tuesday.

[thommo]

I have contacted the QiHe guys and they are investigating it now, and assure us that they will replace immediately at their cost if they have made a mistake.

Other than perfection straight up [which is impossible with any new product - we've made enough to know], I don't think we could ask for anything more from these guys right now.

Peter

[anfang]

The feeder plates measure 17.17mm thick, at the inner  edge.

Yes, same with mine: 17.19mm. I have a few feeders where the handle looks like it won't stay put, but so far they have. I looked for an adjustment when I found the first feeder where the handle wouldn't go down, but couldn't see a way to adjust it. I ordered more feeders from KingSun (including nozzles) and will monitor those.

If they send a new plate, does it include the air hose fittings so that the quick release hoses just have to be swapped? Or does it require that you take all the old quick release fittings out and swap them? The former seems easy, the latter seems like a long afternoon.

Have you received and tested those smaller Juki nozzles yet?

Not yet, in the air from China but I'm backed up on other things just now so it'll take some time to verify.

BTW, how does everyone adjust the offset between the nozzle and the downlooking camera? Adjusting uplooking camera is easy of course. But the biggest challenge with 0402 is adjusting the offset between the DL camera and the nozzle.

[glenenglish]

Hi Anfang

I suggested to them they just ship blank plates

It is fairly easy to change over the air system interface,- very quick actually,
4mm hex to remove 6 bolts that hold plate in
10mm spanner to loose air sockets so can be removed
Phillips #1 screwdriver to remove the rest....
Probably 20 to 30 min per plate.

Or you can mill down your own plates-inside underneath 15mm deep (the genuine feeders have long length grab) and 16mm total height  (so about 1.2mm comes off the bottom in a step)
16.0 + /- 0.05. probably +/- 0.1 would do it.

[Smallsmt]

I think 16mm is not ok the feeder will not fix well.
We used 16.75mm and it works fine.
Has anyone tried to fix the feeder cassette using 16mm plates?

[mrpackethead]

Yamaha thought 16.0mm was a good size.     With 16.75mm, you won't be able to fit genuine yamaha CL feeders and clones that have bene made to spec.  Some of the CL feeder Clones are not made to spec, are a lot thinker and will fit on thicker base plates.

[RobK_NL]

Same here, 17.15mm

Tried one 8mm feeder, but locking it in place doesn't quite feel like I feel it should feel 

Or you can mill down your own plates-inside underneath

Wutt? You're kidding, right? I didn't buy this machine as a hobby project to get me through the winter, I bought it to do serious work, which is precisely what QiHe sells it for. So I fully expect them to come with a serious answer and a serious solution. I'll wait for Peter to report back on this.

[Smallsmt]

Same here, 17.15mm

Tried one 8mm feeder, but locking it in place doesn't quite feel like I feel it should feel 

That means you can't put the handle down?
But the feeder cassette is fixed no shaking?

[RobK_NL]

Yes, the feeder is fixed, but it feels like I have to force something to get there. Never worked with Yamaha feeders before, so I don't know how it should actually feel.

[Smallsmt]

It should only need a little force to get fixed you feel a click.
But most important is the feeder need to be fixed not shaking!

[xdave]

Hopefully this is a simple question for those of you who already have the TVM920?  I see from the QiHe website that the footprint is L 1080 * W 805 * H 750mm.  Can it be installed on a bench that is 900 deep without fouling the feeders?  Not having used the Yamaha-style feeders I don't know if they come down further than the base of the machine.  Are the feeders the same for 13" reels as with the smaller ones?

[thommo]

Well, on this subject all I can offer is my knowledge that an original Yamaha PnP machine has a 16.0mm thick feeder plate at its clamping surface.

It should only need a little force to get fixed you feel a click.
But most important is the feeder need to be fixed not shaking!

[thommo]

Hi xdave,

The answer you are looking for is YES.
You can sit a TVM920 on a flat surface and not have concern that the feeders will interfere with it.
The feeders sit in the space above the table surface plane, and remain well clear of it.

Cheers- Peter

Hopefully this is a simple question for those of you who already have the TVM920?  I see from the QiHe website that the footprint is L 1080 * W 805 * H 750mm.  Can it be installed on a bench that is 900 deep without fouling the feeders?  Not having used the Yamaha-style feeders I don't know if they come down further than the base of the machine.  Are the feeders the same for 13" reels as with the smaller ones?

[glenenglish]

QiHe have been most helpful and gracious with this problem, and  are modifying their feeder plates and will provide new blank feeder plates (and new spacers)  for those who ask for them. They will be supplied blank so you have to changeover the air system interface which is quite easy and faster than you think.

You can also modify the existing plate yourself, contact me directly, I have done a technical drawing of the thickness  mod.

Both MrPacketHead and myself own Genuine Yamaha machines, YVL88ii and YV100
These are CL feeder machines.
So we should know what we're talking about.

If you do not own a genuine Yahama machine and/or do not have experience with these CL feeders on genuine Yamaha machines , please do not contribute with idle speculation that confuses those who do not have experience.

POST SCRIPT
Couple of minor issues like feeder plate O ring retention plate is a bit too high (1.6mm) and needs to be recessed (the one with 20 little screws) , that will be fixed, also. It only (in a minor way) affects people trying to use old Yamaha CL feeder (like me) . But it will be the perfect article. I will get CAD dwg for approval. Qihe want it perfect which is good. 
I have a large variety of CL feeders and they have  variations between them and I am confident we have covered all bases by being 100% compliant.
 
Glen.

[glenenglish]

Hi Xdave
I can confirm no problem with large reels (13" etc) and large feeders (24 etc)

My 920 is currently sitting a couple of inches off the floor and even with the large feeders 24mm or small and large reels, there is still plenty of clearance.
So to answer your question, no problem even with machine sitting on the floor.

My suggestion is that in order to get feeders in and out easily, front clearance needs to be at least 450 and rear clearance 300  (the feeder plate is recessed in the rear) . I have my monitor and kbrd above the machine so I dont need any working room left and right. But access to covers is useful allow min 150.   large feeders 24+ wont fit on the rear feeder bank due to the recessed feeder plate - minor issue.

cheers

[RobK_NL]

So, now that we can securely place those feeders, how do you go about loading them?

I mean, do you just lay it down flat on the bench or is there some kind of jig available in which you can set it upright? I had a jig like that for my ancient DIMA, so I can imagine there's something
available for Yamaha feeders as well.

[glenenglish]

Hi Rob
Flat on the bench is fine for loading, but you can get jigs. Or you can do it with a 115 x 16mm thick slab of anything with a couple of holes in the right spot in a vice...

For storage , I have got feeder Trolleys, roughly up 80 on each trolley. QiHe sells Trolleys, or you can buy them  . About $175USD seems to be the right number for 80 feeder CL trolleys.  I got latest from Qihe.

http://www.wickonsmt.com/sale-7925610-smt-feeder-cart-fuji-machine-feeder-cart-feeder-storage-cart.html
Others like  SHenzhen J-wide, goodlucksmt  ,
They make them into whatever lengths you need

You can buy simple and complex complete calibration jigs ( for post repair or periodic maintenance) . I have no experience using them.

g

[Spikee]

Has anyone assembled boards yet with this machine ? if so how does it handle 0402 and those kind of things ?
Also can anyone comment on the software; is it usable as is (with possible minor tweaks to pnp file) or is there thing that just can't be fixed at this moment ?

Neoden 4 does not use "real" feeders but it seems to be working well (as in vision) for many people.

I'm just looking at options.