One of the projects that has been on [Peter Jansen]’s build list for a long time – besides a fully functioning tricorder, of course – is a pick and place machine. It’s a project born out of necessity; each tricorder takes four days to assemble, and assembling the motherboard takes eight hours with a soldering iron and hot air gun. The pick and place machine isn’t complete yet, but one vital component – the vacuum head for picking up components – is getting there with the help of some odd components.
A few months ago, [Peter] saw a post on Hackaday about repurposing a tiny piezo micro blower for use as an extremely small vacuum pen. The original build was extremely simple – just a few pieces of foam board and a power supply, but the potential was there. A tiny electric air pump that’s able to pick up large chips and modules along with tiny resistors without having to run a hose through the mechanics of a CNC gantry is a godsend.
[Peter] got his hands on one of these micro blowers and started work on a proper tool head for a pick and place machine. A port on the micro blower was covered so it would suck instead of blow, the vacuum port was threaded through a stepper motor with a hollow shaft, and a fine tip was attached to the end.
What can this vacuum head pick up? 0604 size resistors aren’t a problem, but larger modules are simply too heavy. It looks like this micro blower would only be able to pick up small components. There are other options, though: [Grant Trebbin] has had some luck with a larger pump from Sparkfun, but this requires a vacuum line to run through a CNC gantry. There’s still some work to do before a small vacuum head shows up on the tool head of a pick and place machine, but given how long it takes [Peter] to put together a single tricorder, it’s well worth investing the time to do this right.
31 thoughts on “Towards A Tiny Pick And Place Head”
“One of the most challenging things to accept as I’ve been getting a little older is that there’s only one of me, and only so many hours in the day — especially when trying to have a work/life balance.”
This. This times a hundred.
Running tubes on your gantry should not be a problem when using a dragchain (can be found cheaply on ebay).
That it is currently only able to pick of 0604 is a sad tho. I asume traveling distances with 0604 will be to much too then; And that is a shame. But I like his effort! Keep going!
You can get silicone tubing from various places like cole-parmer that is both tough and super flexible. I’d go with weaving that through a machine before messing with silly little vacuum pumps that are never going to have enough suction or reliability to work for something like this. Sometimes K.I.S.S. isn’t the most elegant way of doing things, but it is the smarter way.
Piezo elements are extremely reliable, so I’m not sure where you’re pulling that info from.
Sorry, I just realized you were referring to the suction being unreliable, not the pump.
Yeah, I guess what I meant was a small piezo pump like that isn’t going to have enough suction to hold onto larger parts reliably when they have to be picked up at X location and transported to location Y and then placed with any sort of accuracy.
I think you’d need a good solid grab so the part doesn’t get dropped in transport or vibrated out of position on the way there. Most stepper driven CNC gantry stuff like printers and the like seem kind of jerky at the end of motion, unless you have some sort of ramping routine going on.
I don’t know much about piezo pumps, but I assume they could be stacked to get more out of them maybe?
most pnp machines change heads out, for larger tips or smaller depending on the part about to get picked up, time to work on the head changing system!
Yup, more surface area on the tip would allow larger items to be picked up with the same level of vacuum.
Something similar to this (tips only) would be appropriate:
I had hoped that larger tips with more surface area, or suction cups attached to the head, would be the saving hand for the microblower. Unfortunately even with a tip that’s nearly a third of the width of the small magnetometer it was still just barely able to be picked up, and attaching tiny suction cups from a small penlike pick-and-place tool that I found from China didn’t appear help.
In contrast, the suction from the Sparkfun vacuum pump is so strong that it’s able to pick up everything from an 0603 resistor to a 100-pin TQFP and entire bluetooth module with a single tip, without a suction cup, and running the pump all the way down at 8.5V (instead of 12V), where it’s a little quieter. I’d love to have the elegant microblower solution work, but the simplicity and reliability of the larger pump on the business end seems to be a better choice right now, even at the expense of running tubing around the machine.
Well, it was worth a shot! If you’re interested in some other options, I have what they call the MeMax pump that I’ve used for pumping water. It’s a tiny diaphragm pump that I’ve ran at roughly 50% pwm at only 5 volts in the past; usually near silent at that voltage. Running at rated voltage it is pretty noisy and vibrates a lot, however.
Still, it’s hardly larger than a 9 volt battery.
Thanks — how have you found it to work as a vacuum pump? In the post I mention trying something a diaphram pump in a similar form factor (from http://www.clarksol.com/html/prodspecsKPMSPump.cfm ), and unfortunately it didn’t end up having enough suction. At the time I also thought it was quite noisy, but my threshold for noisy pumps has been increased somewhat after using the Sparkfun pump. It’s loud, but it does a great job.
I’ll admit I have not tested it as a vacuum pump and their documentation makes no indication that it is suitable for air… despite the eBay auction I purchased it on (from their eBay store) saying otherwise. I’ll leave myself a reminder to try and test it for vacuum tonight. Hopefully you’ll see a reply from me tomorrow!
I have some little gear pumps like that. I haven’t quantized how much vacuum they’ll make with air, but they will self-prime water through aquarium airline tubing from surprisingly large head heights. Like a couple feet at least. They’re slow though. The pressure might be OK, but the air pulls slowly and may not be enough to hold on to components in a pick n place.
Setting up a water column I managed to pull up to 67″ of water (about 5″ mercury) which was as high as I could test. It was still pulling water at this point and not appearing to have slowed down much.
This was drawn through 0.14″ line and took roughly 30 seconds to reach that height which is a flow rate around 2.04 cubic inches per minute or 0.034 liters per minute. This is quite short of the sparkfun pump which is rated at 12-15 lpm and up to 16inhg.
Two pumps in series and pick-up tool area to match weight of object to be picked up? Two in series sounds easier with no tool changing, and probably pulls down faster. Wondering about a solenoid working a rubber bulb syringe…..
I would check for leaks around the shaft of the stepper before giving up.
The microblower can be stacked for a more powerful vaccuum. The way it is driven also affects wildely its sucking power. Add all that with a rubber cup for picking large components and you can move largish qfp.
How about using a larger tip? Equal pressure times larger area = higher force (think fridge door). For instance, I once used a air-mattress-type foot pump in a little ad-hoc P&P tool, but it never worked until I added a 3D printed nozzle expanding to slightly below the size of the IC.
BTW it is either 0603 or the much smaller 0402 parts in the actual article. The write up states 0604.
If he is using the microblower, he could mount it along with the needle on the other side of the stepper to minimize leaks. Routing power can be pretty easy these days with wireless chargers or the old fashion brushed contacts.
Really there’s no need for a pick and place to be capable of more than one 360 degree rotation, so just a little extra wire length should be enough to solve that.
The only real issue I see is cost. These little blowers are stated to be $45. The whole conventional head/hose/blower can’t be that much, can it? Replacing bits and pieces would be less expensive as well.
HaD effect again (a link so interesting, the author’s page blows up, and you can’t acces it).
Must check later!
Is running tubing through the gantry really such a big deal? I was under the impression you already have stepper motor and sensor wiring going to the gantry. I don’t understand what is so hard about throwing another 4mm OD tube into the cable chain.
In my own first crack at a PnP head, I’ve had reasonable results with small diameter PVC tubing and a right-angle hose barb through a hollow shaft for a leak-free air path: https://www.youtube.com/watch?v=Deq2ZG6usIs
The vacuum here is provided by a cheap reversed aquarium pump, with a solenoid mounted on the head switching the suction. It can pick up fairly large parts using the rubber suckers from a Chinese vacuum pen, but for realistic use I’d probably replace the fishpump with something a bit stronger (or parallel a few fishpumps in a pinch – super quiet and the price is right).
This ties in nicely with [Acidbourbon]s XY drill table from yesterday!
A few years ago I used to work on an SMT line, and it is interesting to see what makers take from commercial machines and what they ignore. As Nashblackcats comment (on the article) mentions, a venturi and compressed air is often used in PnP machines, rather than vacuum pump; it’s cheap to do, easy to service. Some feeder designs advance using pneumatic cylinders and solenoid valves, so the machines need the air anyway.
Noise is not such a concern, you don’t hear much air over whirr of the XY axes.
It’s hard to tell what [Peter] is planning to do for his Z axis – Like others I’d suggest a PVC or Silicon tube and call it a day. One can vary the vacuum with solenoid valves, or crushing the tube (with servo or a stepper) to save cost. And if he cares about his ‘unsprung weight’ (the Z drive moving the R axis stepper) use a telescoping pair of metal tubes: the body of the Rstepper remains stationary relative to the Y head, the outer tube moves in both Z and R, but the inner tube does not, and connects to your vacuum venturi.
As for feeders, have a look at the mydata agilis feeders – the cover tape is peeled right off, just pushed to one side to expose the pocket. I realise for smaller boards a flat set of pockets works well. but ~three hundred mostly 0402s? He’ll want a feeder system surely.
Brilliant. I’m excited to see the coming advancements in DIY pick and place. I’m just wondering how easy it will be for me to take a completed Eagle design and get this thing to plan its travel movements based on that. Also, won’t the components need to be on reels? I’m wondering what the workflow for this whole process will look like.
why run the tube through the centre of the stepper motor? just seems weird and pointless to me
oh wait… so you can rotate the part after picking it up right?
I’m building a Pick and Place machine too. I ordered in 2 of these micro pumps as well.
No, they’re not cheap. You need to locate the exact frequency to maximize their sucking power. A khz more or less and you lose half the suction. Even then it was just strong enough to lift a 0805 part but not an IC. I was going to cascade them but I cracked one when I man handled it. In the end I bought a spark fun pump, and hauled the tubing via cable carriers along with the wiring. However its way to noisy and shakes all over the place. Something better is needed.
Also creating suction is only half the battle. When you turn off the pump the component stays in place because the tube is still evacuated. I’ve used a release valve to let air back in, but a positive pressure “puff” would be better.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)