A small, desktop pick and place machine has obvious applications for hackerspaces, small companies, and even home labs. However, despite multiple efforts, no one has come up with a solution that’s both better and cheaper than buying a used, obsolete pick and place machine. [Mika]’s brdMaker is yet another attempt at a desktop chipshooter, and while the prototype isn’t done yet, it’s a fantastic build that might soon be found in your local electronics lab.
The easy part of any pick and place machine is a Cartesian frame. This has been done over and over again by the 3D printing and CNC communities, and the brdMaker is no exception. [Mika]’s robot is a 600 by 600 mm CNC frame powered by NEMA 23 motors. So far, so good.
The tricky part of a pick and place machine is working with the fiddly bits. This means feeders and machine vision. There are several different options for feeders including a ‘drag’ feeder that uses the vacuum nozzle tip to move a reel of parts along, and a slightly more complicated but vastly more professional feeder. A machine needs to see the parts it’s putting down, so [Mika] is using two cameras. One of these cameras is mounted on the toolhead and looks surprisingly similar to a USB microscope. The other camera is mounted in the frame of the machine to look at the bottom of a part. This camera uses 96 LEDs to illuminate the component and find its orientation.
[Mika]’s brdMaker still has a long way to go, but there are indications the market is ready for a cheap, easy to use desktop pick and place machine. The Chipsetter, an exquisitely designed pick and place machine revealed at last year’s NY Maker Faire had an unsuccessful Kickstarter, but they’re still chugging along.
I would say that an unusuccessful Kickstarter is a pretty good indication that market *is not ready* for another pick and place machine. There are plenty of professional machines with proven and working feeders available for little money – like the desktop Neoden ones for places that actually need them. And those are market proven and known quantities.
Hint – if you think you need a pick and place you most likely don’t. You would be much better off by contracting the ocassional assembly out to a prototyping service. Both cheaper and easier than running and maintaining your own p&p machine.
Most people will discover that building these machines usually costs more than the TVM802B or TM245P.
There is also the physical space needed for the noisy machines to operate.
I do like the concept of OpenPNP, but the user base with spare money does not have time to tune proprietary hardware.
Recall that people that buy these machines – want them to solve a labor time problem.
Well the Chipsetter is not exactly “chugging along”. Their website says “Chipsetter has ceased daily operations.”
I think it’s difficult to compete with the Chinese “Neoden”, “Charmhigh”, and others…
“available for little money”
What? €7,995.00 is little money?
This is HACK-a-day not professional board-house-a-day!
Yeah, I know those two are far from mutually exclusive but come on, you are completely ignoring a large portion of the readers here!
“Hint – if you think you need a pick and place…”
I think I need water, food, clothes, medicine and a roof over my head. If you think you NEED something more than that…. you are mistaken. But if everyone stopped with just what they need then we would all still sleep in caves, hunt with spears and medicating ourselves with wild berries.
“…you most likely don’t.”
Really? You don’t say. I never would have guessed that from all the awesome things people have built in their home workshops, garages, or even tucked away in a cubby of their parent’s basement.
As with any expensive, complicated piece of manufacturing kit I think diy homebrew maker hacker hobbyists or whatever people choose to call themselves will just be able to make and do that much more once they have cracked that nut. It’s not about somebody babysitting some expensive and finicky machine. It’s about bringing previously high tech gear into the mainstream, making it accessible, making it cheaper and eventually even making it easier to use. Just look at what Reprap did for the home 3d printer market! It already built a large home consumption market where no such thing existed. And I would say that is a process that is still ongoing and will still go much farther too!
So… stop trying to stand in the way of progress!
Yeah!
I HACKED together a cnc that cost around $400 in parts I didn’t have. If I added some pumps (free), nozzles (diy), cameras (70$?), feeders (contact 3d printed for $15 each? Biggest unknown), led ring (already have leds), hoses (aquarium parts from garage sales almost free/already have), it seems I could be a little over another $400.
It seems too good to be true, and I’m trying to figure out if all the YouTube videos of people making them are really legit machines that can populate a few dozen boards unattended pet day. For me, it doesn’t need to be fast, just good quality placement.
It’s a hobby, so I don’t care about the time in building it. But I do want a functioning machine as a result.
I for one, stopped making cnc videos once I built my cnc because I’m now having too much fun USING it! I can’t figure out if the same happens for the pnp folks…. I see a videos of “first placement” and not much afterwards. Can anyone weigh-in?
“Really? You don’t say. I never would have guessed that from all the awesome things people have built in their home workshops, garages, or even tucked away in a cubby of their parent’s basement…”
I meant to say “…without one.” at the end of that
But the best part of this project is it’s use of OpenPNP software, which is gradually getting amazing.
firepick – http://delta.firepick.org/
I followed Firepick closely. It looks 2 years dead to me. Check the last time their Github files were touched.
Yes, it is for the most part dead. There are still a few people working on it, but it isn’t looking very promising at this point due to all of the mechanical problems that the design has. Calibration of the machine has turned out to be a very difficult challenge, with no real solution at this point.
I wouldn’t recommend that anyone try to build one, as a Cartesian based system will actually work properly.