Author Topic: Pick and place machine kit, with vision  (Read 44210 times)

0 Members and 1 Guest are viewing this topic.

Offline JuKuTopic starter

  • Frequent Contributor
  • **
  • Posts: 566
  • Country: fi
    • LitePlacer - The Low Cost DIY Pick and Place Machine
Re: Pick and place machine kit, with vision
« Reply #25 on: February 06, 2015, 01:27:48 pm »
Excellent effort, and I don't want to be overly critical, but I agree with Mike and Dave - I really don't see the point.

Hand placing would be significantly faster (especially when you include setup time), and since the machine doesn't support feeders or auto loading/unloading of boards, it's not like you can let it run all day on it's own.  The placement accuracy also seems quite poor... seems there is 0.010" and/or several degrees of error on many of the placements. 

I don't think it comes close to buying a used machine (I always recommend a Quad 4C) for the $5-7k range which will likely come with a couple dozen feeders, be orders of magnitude faster and more accurate and have factory support if necessary.  I know Mike has another brand of machine he's used that also seems quite capable, and perhaps even less $$.

Picking and placing is not a difficult task.  What makes it hard is performing the task quickly and accurately, as well as the feeding of components.  This device doesn't address any of the difficult aspects of PnP'ing.

But still, a significant achievement in difficulty and you should be proud of your work.
You said you don't see the point, so let me explain: The point of the machine is to help the prototype phase. Feeders, auto loading of boards and other stuff you mentioned belong to production environment. I'm the first to say this is not a production machine - for daily production, there are better options.

But I do claim this is faster than hand for a prototype, unless the board is very simple. Building a board for the first or second time, you are looking up each component and its place on the board, and that is tedious and error-prone. If you are hand-placement for production, you would know the position of the components by heart, for a new board you don't. Not a big problem if your board has 20 passives and half of them are bypass caps and the other half pull-up resistors, but when you have a new board with 100+ passives (nowadays often 0603) with various values, you really don't want to do that by hand. For my designs, that would still be on the simple side. This is the problem I built the machine for. Whether the problem is universally big enough that it pays off to finish the design and software properly so others can use it too, prepare the kits, build the website and documentation and so on, we'll see. It will be interesting and fun to find out!

The inherent accuracy of the machine is somewhere in 0.1 -0.15mm range and a small fraction of a degree. The machine does not look at the part itself, it uses the tape holes for location. The parts sit loosely in their pockets, and that is significant error source for placement and most of the rotational error. Still, the parts would reflow and self-align just fine. Paradoxically, 0402s are placed more accurately than 0603 - the small parts have tighter pockets.

For setup: You must get parts from part storage and place them somewhere, no way around that. When you have done that, the setup is minimal. The video shows all that you have to do: turn on the machine, start the software, one click to home it, another click to select placement tab, three click to load the pick and place file (unmodified, directly from CAD system), three clicks to select all components for placement, two to exclude fiducials (you don't want to place anything there), two clicks on each component type to tell the machine where you put the tapes on and one click to place all; few seconds total. I don't know how this could be simpler?
http://www.liteplacer.com - The Low Cost DIY Pick and Place Machine
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3644
  • Country: us
  • If you want more money, be more valuable.
Re: Pick and place machine kit, with vision
« Reply #26 on: February 06, 2015, 08:55:21 pm »
I am also in the position where I need to quickly assemble new/unfamiliar PCB's with well over 200 components. The plan is to go with a Quad 4C from PPM for production use. The random used ones are somewhat cheaper but could also turn into a project that cost too much time - so PPM wins on Quads.

With that satisfied, I like the idea of having one of these to help with prototypes and pilot runs (maybe 20 PCB's) and not disturbing the production flow. Our layouts have become so dense that silkscreens are almost useless so they are hard to assemble. If I need to make 5, 10, or 15 pcs the speed is largely kmited by my ability to focus at that level of detail for many hours. It is so easy to swap a .1uf cap for a 1uf cap and the PCB may not work.

As for alignment, I would not worry about what I saw. I have hand assembled a lot of PCBs and have a feel for what will be ok in the re-flow. Even placing .5mm QFN's by free-hand vac pen placement is better than 99% successful for me and I am not a machine. The alignment I saw was better than I do manually. PCB's at this volume are 100% inspected as well, so if something is a bit off it should not matter that much.

My goal is to offload the ultra tedious nature of finding hundreds of pads on a PCB with little to no silkscreen. Even the the overall time is exactly the same as manual assmebly - I would save my sanity and the machine would be worth it.
Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13971
  • Country: gb
    • Mike's Electric Stuff
Re: Pick and place machine kit, with vision
« Reply #27 on: February 06, 2015, 09:26:37 pm »
Placement accuracy is in most cases not an issue for 1-offs, and certainly not justification to spend this much money. Chip parts will usually either self-align or be on enough of the pads to make a joint even if they don't self-align.
The exception is where cosmetics are important - e.g. LED arrays.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1546
  • Country: au
Re: Pick and place machine kit, with vision
« Reply #28 on: February 06, 2015, 09:42:03 pm »
The machine is a Polish made Mechatronika M10V.

I am looking to make a bunch of cut tape feeder sections of various tape widths to facilitate tighter packing density and more appropriate feeder load.
 

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6572
  • Country: nl
Re: Pick and place machine kit, with vision
« Reply #29 on: February 06, 2015, 09:44:19 pm »
The machine is a Polish made Mechatronika M10V.
Don't get your remark, with a starting price of over  € 15.500,- that is a whole other ballgame  :-//
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1546
  • Country: au
Re: Pick and place machine kit, with vision
« Reply #30 on: February 06, 2015, 10:04:12 pm »
Precisely Kjelt

It is ten times dearer and maybe four times quicker.

The resources Juku's machine offers are great value for money and certainly ought not attract the dark clouds of poo-pooing.( technical term for critical analysis)
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3644
  • Country: us
  • If you want more money, be more valuable.
Re: Pick and place machine kit, with vision
« Reply #31 on: February 06, 2015, 10:19:02 pm »
Placement accuracy is in most cases not an issue for 1-offs, and certainly not justification to spend this much money.

$1,350 plus a day or 2 of assembly and fiddling is below the noise floor. If it saves me from 10 hours of focused pad searching , I will consider it a success story. My little business is in a tough middle-ground where everything we do is too big for manual work and too small for real production. Taking our PCB's to an assembly house is very slow and expensive because the primary cost is setup. Not just the machine setup, but the organization and prep we have to do just to get the job to assembly. After all that, we make a handful of PCB's and move on the next revision.

My long term goal (which Corporate666 was a part of the conversation if I remember) is to have a Quad 4c maxed with feeders so we can run small qty's of a variety of PCB's. With that machine all setup and ready to go, it will be cheaper to have another little helper machine in the lab to assist in prototypes.

We had the same challenge with our CNC machines. Once they are setup and dialed in for production, it was painful to do a prototype. If a few things are moved around, we would have to re-set the machine which takes a lot of time. It was faster to have a hardly used less capable machine dedicate to prototypes and modifications.
Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline JuKuTopic starter

  • Frequent Contributor
  • **
  • Posts: 566
  • Country: fi
    • LitePlacer - The Low Cost DIY Pick and Place Machine
Re: Pick and place machine kit, with vision
« Reply #32 on: February 07, 2015, 08:22:20 am »
My goal is to offload the ultra tedious nature of finding hundreds of pads on a PCB with little to no silkscreen. Even the the overall time is exactly the same as manual assmebly - I would save my sanity and the machine would be worth it.
Exactly why I built the machine! My boards don't have silkscreen either.
http://www.liteplacer.com - The Low Cost DIY Pick and Place Machine
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3644
  • Country: us
  • If you want more money, be more valuable.
Re: Pick and place machine kit, with vision
« Reply #33 on: February 07, 2015, 11:56:43 pm »
Have you attempted any fine pitch IC's yet? QFN's? BGA?
Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline donotdespisethesnake

  • Super Contributor
  • ***
  • Posts: 1093
  • Country: gb
  • Embedded stuff
Re: Pick and place machine kit, with vision
« Reply #34 on: February 08, 2015, 09:50:25 am »
I think it is a great project, and nicely done so far. I have been working on my own DIY machine, but not got very far. For me, speed is less of an issue than convenience. Sure I can spend 5 hours doing manual place, but I have other things to do! For mainly hobby use, I can't justify much more than $1500.

The sort of jobs I am looking at are one or two off prototypes, and in future batches of maybe 10 or more boards. Mostly small boards. For 0603, some 0.5mm QFN and QFP would be really useful, as I have trouble placing those manually.

I feel that supporting reels is useful, even if with manual or drag feed. Failing that, having a tray of strips I can easily swap in or out. I think that the problems with "loose" strips are surmountable. A cheap automatic feeder is like the Holy Grail of DIY PnP, I would settle for something cheap and simple.

I had a look at PP4, which is like a Rolls Royce of DIY machines, but has relatively simple auto feeder which I think I could build.
Bob
"All you said is just a bunch of opinions."
 

Offline briandorey

  • Regular Contributor
  • *
  • Posts: 51
  • Country: gb
    • Project Blog
Re: Pick and place machine kit, with vision
« Reply #35 on: February 08, 2015, 10:02:14 am »
This looks like a very useful machine and much easier than placing components by hand for hours at a time. When we developed our own DIY pick and place we also had the issue with reliable feeders and ended up using used commercial feeders which we purchased from ebay and modified to fit our machine and uses an air ram to activate the feeders.
 

Offline donotdespisethesnake

  • Super Contributor
  • ***
  • Posts: 1093
  • Country: gb
  • Embedded stuff
Re: Pick and place machine kit, with vision
« Reply #36 on: February 08, 2015, 03:37:06 pm »
This looks like a very useful machine and much easier than placing components by hand for hours at a time. When we developed our own DIY pick and place we also had the issue with reliable feeders and ended up using used commercial feeders which we purchased from ebay and modified to fit our machine and uses an air ram to activate the feeders.

Brian, I looked at your project when doing research, and was very impressed!

I think I found the feeders you refer to on ebay, and I did buy one as a sample, although shipping is rather pricey.  http://www.ebay.co.uk/itm/Panasonic-X708PB084C-8x4mm-Paper-Feeders-SMT-Placement/230858305569?rt=nc At the time of writing, there are 7 available, as these are surplus items I'm not sure where I would find more.

There is also note saying "Please note that some of the waste tape front covers are missing" not sure whether that is significant or cosmetic, and the one I received is not in best condition. Either way, £400 for a set of feeders is a bit on the high side, I was hoping to build a machine for around £500-600.

Volker Besmens design for a simple auto feeder is shown here https://www.vbesmens.de/en/pick-and-place/automatic-feeder.html, although it would take some custom construction, I guess it would be around £150 for materials.
Bob
"All you said is just a bunch of opinions."
 

Offline briandorey

  • Regular Contributor
  • *
  • Posts: 51
  • Country: gb
    • Project Blog
Re: Pick and place machine kit, with vision
« Reply #37 on: February 08, 2015, 05:30:49 pm »
This looks like a very useful machine and much easier than placing components by hand for hours at a time. When we developed our own DIY pick and place we also had the issue with reliable feeders and ended up using used commercial feeders which we purchased from ebay and modified to fit our machine and uses an air ram to activate the feeders.


I think I found the feeders you refer to on ebay, and I did buy one as a sample, although shipping is rather pricey.  http://www.ebay.co.uk/itm/Panasonic-X708PB084C-8x4mm-Paper-Feeders-SMT-Placement/230858305569?rt=nc At the time of writing, there are 7 available, as these are surplus items I'm not sure where I would find more.

There is also note saying "Please note that some of the waste tape front covers are missing" not sure whether that is significant or cosmetic, and the one I received is not in best condition. Either way, £400 for a set of feeders is a bit on the high side, I was hoping to build a machine for around £500-600.
That is the same seller who I got my feeders from and they did a very good deal on the postage costs when buying more than one feeder. The waste tape covers are a plastic disk which stops the plastic tape from falling off the spool but it would be easy to make new ones from some thin plastic or card.
 

Offline donotdespisethesnake

  • Super Contributor
  • ***
  • Posts: 1093
  • Country: gb
  • Embedded stuff
Re: Pick and place machine kit, with vision
« Reply #38 on: February 08, 2015, 06:40:14 pm »
@Brian, thanks for the info. I briefly tried the feeder and it seems to basically work so I will see what I can do with it.
Bob
"All you said is just a bunch of opinions."
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13971
  • Country: gb
    • Mike's Electric Stuff
Re: Pick and place machine kit, with vision
« Reply #39 on: February 08, 2015, 08:47:43 pm »
For those in the UK, Versatronics feeders would be a reasonable option, as they are multi-lane and RS232 controlled
e.g.

e.g. http://www.ebay.co.uk/itm/Versatronics-Pick-Place-Feeders-Smt-/291375813777?pt=LH_DefaultDomain_3&hash=item43d75a1491 
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2010
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Pick and place machine kit, with vision
« Reply #40 on: February 09, 2015, 06:09:16 pm »
But I do claim this is faster than hand for a prototype, unless the board is very simple. Building a board for the first or second time, you are looking up each component and its place on the board, and that is tedious and error-prone. If you are hand-placement for production, you would know the position of the components by heart, for a new board you don't. Not a big problem if your board has 20 passives and half of them are bypass caps and the other half pull-up resistors, but when you have a new board with 100+ passives (nowadays often 0603) with various values, you really don't want to do that by hand. For my designs, that would still be on the simple side. This is the problem I built the machine for. Whether the problem is universally big enough that it pays off to finish the design and software properly so others can use it too, prepare the kits, build the website and documentation and so on, we'll see. It will be interesting and fun to find out!

You would still need to look up where each component goes whether it's hand placed or machine placed.  For hand placing, I would have the prototype parts either loose in a anti-static bag or a piece of cut tape.  So for a prototype, I open up my original CAD file and turn on component values and I place the parts.  I don't see that being slower than having to stick the tapes down to the machine bed, tell the machine where they are, what they are, set the Z offset, etc.  Sure, you could say the parts would stay on the machine and you would use the same parts for prototyping multiple boards, but I do that now with tubs... I have tubs of the common parts I use and when putting together a first unit of a new design, I go to my parts storage trays and grab the parts/values I need and lay them out on the desk.

Quote
The inherent accuracy of the machine is somewhere in 0.1 -0.15mm range and a small fraction of a degree. The machine does not look at the part itself, it uses the tape holes for location. The parts sit loosely in their pockets, and that is significant error source for placement and most of the rotational error. Still, the parts would reflow and self-align just fine. Paradoxically, 0402s are placed more accurately than 0603 - the small parts have tighter pockets.

So the vision system doesn't correct part alignment?  IMO that's a big drawback and the used machines I mentioned will use vision to correct part alignment on-route to the placement.  The accuracy of 0.1mm doesn't help if you are relying on the position of the parts in the tape, which can be off by much more than that amount.

Quote
For setup: You must get parts from part storage and place them somewhere, no way around that. When you have done that, the setup is minimal. The video shows all that you have to do: turn on the machine, start the software, one click to home it, another click to select placement tab, three click to load the pick and place file (unmodified, directly from CAD system), three clicks to select all components for placement, two to exclude fiducials (you don't want to place anything there), two clicks on each component type to tell the machine where you put the tapes on and one click to place all; few seconds total. I don't know how this could be simpler?

I would bet anything that, starting from scratch (components still packaged), I can hand assemble a board much more quickly than machine placing it with this setup.  I can hand place the parts much quicker than the machine does - and I don't have the programming step.



Again, I hope I don't sound too critical - I think you have achieved a lot, but I think you really need to make accomodation for board loading/unloading, use of feeders, vision part alignment and changeable nozzles.  Without those, the machine is slower than hand placing - even for a one-off board.  Those other features are what make PnP hard (and the machines expensive).  If you can add those, you would have a great hobbylist/light-production level machine that would sell for 3-4 times what you are charging now.
It's not always the most popular person who gets the job done.
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2010
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Pick and place machine kit, with vision
« Reply #41 on: February 09, 2015, 06:20:52 pm »
Placement accuracy is in most cases not an issue for 1-offs, and certainly not justification to spend this much money.

$1,350 plus a day or 2 of assembly and fiddling is below the noise floor. If it saves me from 10 hours of focused pad searching , I will consider it a success story. My little business is in a tough middle-ground where everything we do is too big for manual work and too small for real production. Taking our PCB's to an assembly house is very slow and expensive because the primary cost is setup. Not just the machine setup, but the organization and prep we have to do just to get the job to assembly. After all that, we make a handful of PCB's and move on the next revision.

My long term goal (which Corporate666 was a part of the conversation if I remember) is to have a Quad 4c maxed with feeders so we can run small qty's of a variety of PCB's. With that machine all setup and ready to go, it will be cheaper to have another little helper machine in the lab to assist in prototypes.

We had the same challenge with our CNC machines. Once they are setup and dialed in for production, it was painful to do a prototype. If a few things are moved around, we would have to re-set the machine which takes a lot of time. It was faster to have a hardly used less capable machine dedicate to prototypes and modifications.

Yes, we talked about this quite a bit before.

I've 'been there, done that' and spent a huge amount more money than I wished I had.  I've done everything from home-brew CNC machines to conversions to buying old stuff and getting it working to buying current stuff and paying alot more to have it factory maintained.

I do about 10 or 20 new PCB's per year.  I am sure I do pretty much what everyone else in my situation does... you tend to use the same parts again and again (like standardizing on 10k resistors for pullups where possible and using the same decoupling caps every time).  So it ends up being mostly the IC's and specific value resistors/caps used for regulators or for setting voltage levels.  I order the parts I need from Digikey at the same time I order my proto PCB's.  I've found the guy on here selling the mylar laser-cut stencils (sorry, forgot your business name) is a lifesaver too, so I order those at the same time as well.

When my boards arrive, I silkscreen 2 or 3 test units and grab all the necessary parts and put them all on my desk, then hand assemble with tweezers.  For a board with - say - 50 different parts, it takes me maybe 30 minutes from start to finish.

I honestly don't see any way that the machine in this thread can be set up,  components staged, board inserted and programmed and board placed in anywhere close to that amount of time.  All you are doing is automating the pick and place functions which are not the bottleneck of board placement... and those functions can be done faster by hand than by a slow machine. 

The idea of "just let it run while I'm off doing other stuff" sounds good - I used to think the same thing, but when the machine needs a level of attention such that in the time it takes to program it and lay out all the parts, you could have just hand placed the board... then it's not a time saver.  And if you get to the point where it becomes worth it because you can set it up and place 10 or 20 boards, then you're at the point where a Quad 4C makes sense.


Again, sorry to the OP if it sounds like I am crapping on the product, but well-intentioned feedback that gives you a false sense of utility is not doing you any favors... there is a reason nobody sells PnP's without feeders, vision correction or board auto-loading... they just don't save time over hand placing or semi-automatic placement.
It's not always the most popular person who gets the job done.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13971
  • Country: gb
    • Mike's Electric Stuff
Re: Pick and place machine kit, with vision
« Reply #42 on: February 09, 2015, 07:52:48 pm »
then hand assemble with tweezers.
For any parts there are more than a few of, foot-operated vacuum pen picking from tape is a lot quicker than tweezers.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2010
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Pick and place machine kit, with vision
« Reply #43 on: February 09, 2015, 11:34:24 pm »
then hand assemble with tweezers.
For any parts there are more than a few of, foot-operated vacuum pen picking from tape is a lot quicker than tweezers.

I don't doubt it... but for my prototypes, the parts are loose in small plastic tubs like these:

http://www.acehardware.com/product/index.jsp?productId=2793691&KPID=3877794&kpid=3877794&pla=pla_3877794

I use a 12x12" sheet of 1/2" thick acrylic (flat and smooth and can be back lit or be placed over the PCB printout) and dump out a few hundred of the parts and then just cherry pick the ones that are already in the correct orientation and place on the board.

I just find tape strips hard to handle.  If I cut them from full reels, it means a bunch of parts get lost in loading that reel into the feeder, and a partially opened section of tape is a pain in the ass to store and label.

I often thought that a device that holds reels, untapes the parts and advances to the next component would be great for hand placement with a vaccum pen.  A Quad feeder would do this nicely - it's 12V, self contained, requires no air and automatically advances to the next part when an IR beam over the component is broken.  Just too expensive to buy a dozen for a hand placed board.
It's not always the most popular person who gets the job done.
 

Online sleemanj

  • Super Contributor
  • ***
  • Posts: 3043
  • Country: nz
  • Professional tightwad.
    • The electronics hobby components I sell.
Re: Pick and place machine kit, with vision
« Reply #44 on: February 10, 2015, 12:07:42 am »

I honestly don't see any way that the machine in this thread can be set up,  components staged, board inserted and programmed and board placed in anywhere close to that amount of time.  All you are doing is automating the pick and place functions which are not the bottleneck of board placement... and those functions can be done faster by hand than by a slow machine. 


If you read the page about the reason for creating the machine, it may become clearer to you the specific use case - for the types of prototypes that he produces, it takes him 3 hours of manual placing work with a vac pickup to populate one (already pasted) board, 450 SMD parts on a board.  I'd say that would be enough to drive me to make my own PnP as well. 

Quote
Doesn’t look like much, but there are 450 SMD parts on it. Even with a good vacuum pick-up tool, that took about three hours. I was able to do it, but it was more of a hero story than how this should be done. My hand started shaking after the first hour, with two to go. Do note that after putting down the soldering paste, you only have a few hours to place the parts. I did manage to finish the placement before the paste went bad; it wasn’t that good towards the end, but fortunately, the result was great, except my hurting neck. But no, this is definitely not how this job should be done! No way I wanted to go through that again.
~~~
EEVBlog Members - get yourself 10% discount off all my electronic components for sale just use the Buy Direct links and use Coupon Code "eevblog" during checkout.  Shipping from New Zealand, international orders welcome :-)
 

Offline DerekG

  • Frequent Contributor
  • **
  • Posts: 883
  • Country: nf
Re: Pick and place machine kit, with vision
« Reply #45 on: February 10, 2015, 12:48:08 am »
If you read the page about the reason for creating the machine, it may become clearer to you the specific use case - for the types of prototypes that he produces, it takes him 3 hours of manual placing work with a vac pickup to populate one (already pasted) board, 450 SMD parts on a board.  I'd say that would be enough to drive me to make my own PnP as well.
Yes, if your workshop is air-conditioned (refrigerative), the paste would likely have well dried out in that time.
I also sat between Elvis & Bigfoot on the UFO.
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2010
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Pick and place machine kit, with vision
« Reply #46 on: February 10, 2015, 01:31:12 am »
If you read the page about the reason for creating the machine, it may become clearer to you the specific use case - for the types of prototypes that he produces, it takes him 3 hours of manual placing work with a vac pickup to populate one (already pasted) board, 450 SMD parts on a board.  I'd say that would be enough to drive me to make my own PnP as well. 

I just don't see how it's faster.

The machine seems to have a placement to placement time of about 10 seconds.  That's about 3 times slower than a person can achieve.  In either human or machine placement, you need to prepare the parts, and it will be at least as slow (likely slower) to prep the parts for machine placement.  But then you have the additional task of programming the machine.  Even if you load your BOM and parts locations automatically, you still need to go through and program the machine where the source parts are and where all the destinations are.  If it's a few seconds per part, then in the same amount of time you could have just placed it by hand.

Now if you were going to leave your parts setup on the machine so you could run multiple boards, then you achieve an economy of scale by having to only setup the machine once, but if you are running multiple boards - then you immediately run into issues with lack of feeders and lack of board transport in/out of the machine.   It's nice not to have to do the actual placing one's self, but again - it's going to take a lot longer than just placing it by hand, and an occasional one-off by hand is no biggie and if one is doing enough to warrant buying a machine, it should at least be faster and more capable than hand placement.

Don't get me wrong - I am not saying the machine is junk... I think it is a good first crack at the problem.  But the problem is well defined - tons of people have tried to make DIY PnP machines over the years, and the hard-to-solve problems are feeding of components, vision (board skew adjustment and part misalignment) and board loading/unloading.  We had a big thread about this a while back.  If the OP can solve the feeder, on-the-fly vision and board loading/unloading issues, he will have a very valuable product on his hand if it can be <$5,000 (the price of a TM-240), or even better if <$3,600 (the price of a TM-220A).  Just picking and placing from taped down sections of reels is something you can do with Mach8 and a cheap Chinese CNC machine or even an old plotter or motorized XY table.  Solving the vision and feeding problem would be revolutionary.
It's not always the most popular person who gets the job done.
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2010
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Pick and place machine kit, with vision
« Reply #47 on: February 10, 2015, 01:33:30 am »
If you read the page about the reason for creating the machine, it may become clearer to you the specific use case - for the types of prototypes that he produces, it takes him 3 hours of manual placing work with a vac pickup to populate one (already pasted) board, 450 SMD parts on a board.  I'd say that would be enough to drive me to make my own PnP as well.
Yes, if your workshop is air-conditioned (refrigerative), the paste would likely have well dried out in that time.

Kester R276 has a rated tack time of 8+ hours at room temperature.  We all know solder paste specs are overly pessimistic and in real world use, I've placed boards 4 or 5 days after pasting without a single issue. 

If the paste is losing tack at 3 hours, the paste is junk. 
It's not always the most popular person who gets the job done.
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3644
  • Country: us
  • If you want more money, be more valuable.
Re: Pick and place machine kit, with vision
« Reply #48 on: February 10, 2015, 04:59:31 am »
I just don't see how it's faster.

The machine seems to have a placement to placement time of about 10 seconds.  That's about 3 times slower than a person can achieve. 

I don't really agree. Humans can place in bursts only. When there is a an all new PCB it takes a lot longer than 10 seconds to find the pad, find the part, place it, and then confirm that you didn't just place R23 where C56 goes. it is very true that manual and automatic placement still require planning and setup to get the parts in order, but placing a large pad count PCB is way slower than 10 seconds per part - unless you ccan make a lot of mistakes and your circuit still works.

As I said earlier, my PCB's are generally too dense for a workable silkscreen AND they have hundreds of parts. The mind-numbing nature of it takes its toll on speed, accuracy, and repeat-ability. For PCBs that I have memorized, I can burst place 0805 passives at about 200CPH (measured) which is 18 seconds per placement. I can keep that up for maybe 15 minutes at a time. At the end, I am fried and hate life. If this little helps could do a part every 10 seconds with no fatigue and not mixup components - it will be heroic in book. A new PCB that I have not created my little color-coded layout references for take WAY longer than 18 sec per part.

I see it like an auto pilot in an airplane. the pilot can certainly fly the whole time, but the auto-pilot is used to take the tedious part and let the pilot monitor it with a clear mind. Then, when it is time to land and the pilot must be in direct control, they are ready to go. I will say again, even if this machine saved no time at all, it will save my mind. Although I suspect it will save a bunch of time and help me push through prototype and pilot production PCB's.

Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline JuKuTopic starter

  • Frequent Contributor
  • **
  • Posts: 566
  • Country: fi
    • LitePlacer - The Low Cost DIY Pick and Place Machine
Re: Pick and place machine kit, with vision
« Reply #49 on: February 10, 2015, 05:46:46 am »
If you read the page about the reason for creating the machine, it may become clearer to you the specific use case - for the types of prototypes that he produces, it takes him 3 hours of manual placing work with a vac pickup to populate one (already pasted) board, 450 SMD parts on a board.  I'd say that would be enough to drive me to make my own PnP as well.
Yes, if your workshop is air-conditioned (refrigerative), the paste would likely have well dried out in that time.

Kester R276 has a rated tack time of 8+ hours at room temperature.  We all know solder paste specs are overly pessimistic and in real world use, I've placed boards 4 or 5 days after pasting without a single issue. 

If the paste is losing tack at 3 hours, the paste is junk.
The paste was good. The bright (and warm) lights of the microscope were the biggest culprit, I think. (I can't place 0603s manually without good vision help.)
http://www.liteplacer.com - The Low Cost DIY Pick and Place Machine
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf