If you happen to do a lot of SMD work, a pick and place machine is an incredible time saver. The problem is that most automated pick and place solutions are well outside of the “small outfit” price range, let alone the budget of a hobbyist.
We have seen some great DIY pick and place implementations around here, though most are lacking professional features or the sort of documentation that would make it easy for others to replicate. The OpenPnP project is looking change things, with a completely open source hardware and software solution with a price target of under $1,000.
Things are already well under way, with plenty of details available in the project’s wiki. According to the development page, a prototype should go into construction in the near future, and development of the pick and place’s control software is coming along nicely.
While things are looking great for the OpenPnP project, they can always use some help to keep things moving. Be sure to check out the project page if you are interested in lending a hand.
To see some of the progress being made, stick around to see a short demo video of the control software and camera in action.
[via Make]
[youtube=http://www.youtube.com/watch?v=sTWdujEdT1k&w=470]
I wonder if it would be possible to use a syringe on the head to apply solder paste first, then place the components? I’m looking forward to seeing how this project turns out.
yes, they even make pretty small tubes of paste for them
“Overall state of the project is pre-alpha” I suppose every project has to have that status early on, and the earlier an idea an open project is made such, the more minds get on board. How advantages that ultimately turns out to be, who knows? I’m not sure at one point some of these CNC tools start making sense for small production shops, or a home shop either. As long as the home shop owner admits it’s a hobby, I suppose it doesn’t have to make sense to anyone else. Anyway I’ll rely on the Hackaday staff to keeps tabs on the project, and report it.
Please educate me as to why an individual (hobbiest, DIYer, hacker, etc.) would need a PnP machine? I mean, I get the concept, but I can only imagine industrial/commercial applications.
It probably doesn’t make sense for someone who only occasionally wants to build something. Or can build it using through-hole technology. However there are many of us who often build more than one board using SMT parts. It would be a great convenience. It would speed up the process. It would allow us to more easily use small geometry parts.
While I will hand-build a prototype board using 0805 sized parts, I currently either send kits of boards off to a turn-key manufacturer (larger quantities) or hire a tech who works out of her home. It would be productive for me to have a small pick&place machine.
Devices like this and some of the low-end 3d manufacturing devices will give a guy like me the ability to design and build a complete device — even a product in prototype or low-production volumes. Actually pretty exciting times.
So, what happens after a component is placed? Do you have to solder it by hand, or is it actually a “Pick-n-Place-n-Solder”?
If you intend to build more than a few of a certain board, you generally get a solder stencil to put down solder paste. This can be simple as a sheet of laser cut mylar. You use the stencil to place all the paste, then you place all the parts (by hand or with pick and place) and then you throw the board in a toaster oven or on a hot plate. 1-2 minutes later you have a fully soldered, ready to go board.
Jason,
Have you ever taken a loaded board and accidentally bumped it putting onto your hot plate or reflow oven? It’s a big bummer to see all those dislodged parts.
This request is made in jest, but I think your machine should have one more “axis”. The board should sit on conveyor belt that can feed a toaster oven…
I’ve never bumped a board. I’m absurdly careful, but I can see it happening. A conveyor system would be great! That will have to be part of OpenReflow :)
@Jason: Thanks for the education. Good luck with the project!
These are really cool, but the small shops that will do this for you are pretty solid are they not?
If I were to design a board and I needed quite a few populated it seems like it makes sense to use those shops rather than take time away from my next project by getting something like this running.
Maybe it’s a case of: I want to see if I can..
Don’t get me wrong if I had every other gadget already built in my home shop this could easily become my next project, but as it it sits it isn’t going to make my project list any time soon.
They are pretty solid and generally filled with people who know what they are doing. Doing it yourself opens you up to all kinds of quality-control issues (for example soldering issues). Although shops make mistakes too. I had a bunch of boards mis-loaded once by a shop. That was a pain for me and of course it immediately became a blame game.
I think you’re a bit right with the “I want to see if I can” comment. Getting a process like this figured out in your own shop would give you slightly more control and potentially quicker turn-around. You get to trade off the value of your time for the cost of someone else doing work for you (which may vary depending on a lot of things).
LOL…plus bragging rights…
I think it’s the same reason why people like to have laser cutter, cnc mills, etc. you can send it off and it might as well be cheaper.
i think the real advantage is that you can do faster iterations of your designs. Make a change and do the production right away. test, make changes again, produce, and so on. Of course with pick and place you need to produce the boards as well.
PnP’s are awesome for hackerspaces, ours is almost up and running. outsourcing pnp can get expensive really fast if you use it a lot.
but they’re great for just the experience of doing it, mechanical aspects, firmware, control software.
though we do build a lot of smd stuff.
You can also use the control software and gantry designs for other things..
I wouldn’t go with a pnp for solder dispensing though, beyond the its cool to watch it. Its really slow. Stencils are much faster/easier.
Hi all, I am the creator of OpenPnP. The question as to “Why?” is a FAQ and I should probably put it up on the site! :)
The why really comes from personal experience. I have made a few open source PCBs that have attained moderate success and it quickly becomes arduous to place them by hand. For one, I was selling enough that it was a useful amount of income but not nearly enough to justify buying a $25k+ pick and place, or sending off to a board house.
If you have not spent many evenings placing hundreds of parts, it’s hard to understand why it’s such a big deal, but it is. One of my boards which was somewhat popular had about 40 unique parts. I would get orders for 4 or 5 of these at a time through my small web store and it would mean 4-5 hours of painstakingly placing parts each time I got an order. I liked the money, but I hated getting orders :)
Sending off to a board house has it’s own problems. It’s generally not cost effective to order less than a hundred or so boards and that can mean thousands of dollars in parts and boards. For a hobbyist or someone looking to sell a few boards that’s a lot of money.
So, really what it comes down to is if we can make the machine affordable and easy to use, it enables just in time manufacturing and very rapid prototyping. At the price point I am designing it would have saved me time and money.
have you thought about integrating this with a reprap?
and how are you going to store components? directly from the reel or something else?
I’m not sure I follow with regards to integrating with RepRap?
The system currently supports two types of part feeders. Tape and reel using an indexing pin to move the tape, and tray feeders which supports a lot of very simple, DIY configurations.
people think of repraps to place solder paste on the boards. don’t think thats practical compared to solder masks.
Hi Jason,
You’ve done a great job with this project. It’s been exciting to see things take shape and progress. We can’t wait to begin using it for our operations.
Do you know is the resolution with these machines? How accurate are these placements?
Machine resolution is up to the builder. People have built OpenPnP machines with super high end linear servos and glass scales all the way down to converted 3D printers and eBay routers. In general, a median machine using decent linear guides, steppers, Smoothie, etc. will be able to place 0.5mm pitch ICs and 0402 passives reliably.