In the 80s and 90s, building a professional quality PCB was an expensive proposition. Even if you could afford a few panels of your latest board, putting components on it was another expensive process. Now, we have cheap PCBs, toaster-based solder ovens, and everything else to make cheap finished boards except for pick and place machines. ProtoVoltaics’ semifinalist entry for the Hackaday Prize is the answer to this problem. They’re taking a cheap, off-the-shelf CNC machine and turning it into a pick and place machine that would be a welcome addition to any hackerspace or well-equipped garage workshop.
Instead of building their own Cartesian robot, ProtoVoltaics is building their pick and place around an X-Carve, a CNC router that can be built for about $1000 USD. To this platform, ProtoVoltaics is adding all the mechanics and intelligence to turn a few webcams and a CNC machine into a proper pick and place machine.
Among the additions to the X-Carve is a new tool head that is able to suck parts out of a reel and spit them down on a blob of solder paste. The webcams are monitored by software which includes CUDA-accelerated computer vision.
Of course a pick and place machine isn’t that useful without feeders, and for that, ProtoVoltaics built their own open source feeders. Put all of these elements together, and you have a machine that’s capable of placing up to 1000 components per hour; more than enough for any small-scale production, and enough for some fairly large runs of real products.
You can check out some of the videos for the project below.
The 2015 Hackaday Prize is sponsored by:
25 thoughts on “Hackaday Prize Semifinalist: CNC Becomes Pick And Place”
Why is it $5999 for the kit + one feeder on kickstarter, and the video claims 1/10 the cost of a low end PnP?
A neoden 220A is half that price… and the similar priced tm245P already does 10000 pcs/hr…
All engineers should learn how to avoid “The second-system effect”:
But that is an open loop, no checking of anything…
Even a closed loop, low end pick and place doesn’t cost $60K. That’s ridiculous. Madell makes fully closed loop pick and places for $10K.
$6k < $10k: sounds like a good start to me. Have to compare specs for sure.
Also, don't Madell machines require you to use a lead weight to pull the tape off of the parts reel? At least theirs cleans up a little bit after itself.
No, the lead weight version was the earlier feeder type. The newer feeders don’t need it.
Not saying it isn’t a good start, but it’s only a factor of 2 improvement, not a factor of 10. I’d prefer a company get closer to ~$1K before they start claiming a huge improvement.
Madell machines are absolutely horrific. We bought a $10K one, and it never worked right.
The software is full of bugs. It is missing huge pieces of functionality. Documentation is sparse. UI is atrocious. I had to write my own exporter script for Eagle, despite being told that eagle support already existed. Dispense support was based on some hobby piece of software from a guy in New Zealand. They send me the machine with the wrong number of steps per mm, and didn’t really have an answer for me when I asked why it was off. It uses an ancient analog vision system. Their spring-loaded feeders are fickle and are poorly made. They barely have any solutions for actually setting up and running a machine.
We asked them for the newer version of their software, and they want $850/year for it.
A Madell machine may cost $10K, but it will just be a waste of time and money. You may as well save up and go for a low-end MannCorp machine at around $25K.
Depends if you have any skill in programming/scripting/etc. already, I guess. Yes, the software stinks, absolutely. Yes, it takes a bit of a learning curve to get set up. But not $15K of a learning curve. It took us about 2-3 days. My time’s definitely not *that* valuable. Lack of import support of any kind is just a non-issue in my mind – it’s a simple CSV-type input, and it took about half an hour to write one to convert the XYRS file we send to assembly houses to the right format.
And dispense support in a pick and place machine was just silly already. We never even considered that useful right from the start.
I looks like they ship it fully assembled.
Can this machine be used for fisting?
I would assume with proper modification and the upgraded steppers for more torque.
They have an ongoing kickstarter, you could at least give a link to that:
You could have given that link here too, instead of just linking the video….
Hah! My apologies, I didn’t know that kickstarter links plopped the video in here instead of the link with the way the comments section is set up… I learned something new today.
It was a page link. Same results, maybe HaD just embeds the video automatically instead.
If you want to visit the page, you need to click the “K” in the upper left hand corner of the video.
If you want some informed opinion on what’s right and what’s wrong with it, I’d suggest posting about it on the EEVBlog forum. There are a few pick and place projects going on and plenty of people who know a lot about them.
Seems like the liteplacer http://www.liteplacer.com/ will give similar performance for way cheaper (furthermore, the liteplacer is a much more open project). As far as I could tell from talking to them at the makerfaire booth, furthermore the liteplacer does not require reel holders since it is smart enough to just pick parts out of tapes sitting next to the board.
Hello, we do not “require” reels. We also work with cut tapes, waffle trays, and are adding on the ability to work with tube feeders.
Do they claim to be open hardware? Using some open components isn’t much of a selling point if the machine isn’t.
This is not open source, especially the head and its software with the two cameras at an angle. HaD, please use your power to make a project open source, or do not feature it if it is not open source. The HaD price is for Open Source projects, not purely commercial products, right?
The rules for the prize say open source projects will have and advantage, but is not required.
Incredible! Merely all the nice words use to this content! Thank you from the base of my heart and spirit and I will come see once more if you maintain up the excellent job!
I like the helpful info you provide in your articles. I will bookmark your weblog and test once more here regularly. I am relatively sure I’ll learn many new stuff right here! Best of luck for the next!
I somewhat regret they have not succeed. It was really an advanced machine…
Do anyone knows about them some more?
Their blog returns internal server error, the kickstarter was unsuccessful,
no software repository, nothing.
Hey, I was just wondering the same thing. I guess they went bust.
Sadly, there are a lot of projects that start with great ambition but never really get there. It seems a shame all that development effort goes to waste.
Thanks for the heads up on the blog error! An automatic upgrade to WordPress killed the site, but it’s back up now. It’s still terribly out of date though…sorry about that…
We’re not dead. We just had to refocus on projects that were paying the bills. We’re trying to get grant funding to support the release of the pick-and-place machine. It proved too specific a machine for the general kickstarter audience and wasn’t far enough along for professional users. The folks in the middle mostly wanted to make their own machine and so the potential market size wasn’t viable.
We do use it in-house though and our development costs were still less than buying a machine so we don’t see it as a total failure.
We’re a tiny team though, so it’s tough. We’d like to get back to this project, but our team is too small to prioritize this project over other income generating projects.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)