A PnP Feeder for $1.34
Вставка
- Опубліковано 3 чер 2022
- Powered feeders are undergoing a tuning pass to increase reliability, but in the meantime many folks in the community have been designing and building passive feeders!
You can get a LumenPnP kit here!: opulo.io/
A huge thanks to the folks making these mods:
-atanisoft
-@curlytalegames2609
-InfiniteNESLives - www.infiniteneslives.com
-jazzychad - / jazzychad
-nicmoly - github.com/SupaCoder/0816-Lum...
-qwertymodo
-------------------------------------------------------------------
Check out my Patreon:
/ stephenhawes
Join the STR Discord Server:
discordapp.com/invite/TCwy6De
LumenPnP Github Repo:
github.com/index-machines/index
-- FOLLOW --
Twitch bit.ly/stephen-twitch
Instagram bit.ly/stephen-instagram
Twitter bit.ly/stephen-twitter
-- GEAR --
These are affiliate links so if you buy anything through them I get a small percentage which helps out the channel. Thank you!
PCB Holder (Omnivice): amzn.to/344qhVD
Microscope: amzn.to/348aSnw
Soldering Iron: amzn.to/2KlHDGs
Wire Strippers: amzn.to/2IESgDV
Cutting Mat: amzn.to/3oNCwOz
Flush Cutters: amzn.to/3ncFwnc
Digital Calipers: amzn.to/3gIzZSL
Leatherman Sidekick: amzn.to/3a4sQuY
Part Organizer: amzn.to/3a9uDyV
Soldering Iron Tip Cleaner: amzn.to/3a3wZiD
Oscilloscope: amzn.to/2W4zrNt
Multimeter: amzn.to/2JVp6Ba
Crimping Tool: amzn.to/37WymwP
SpaceMouse: amzn.to/3gCWR64
Tweezers: amzn.to/3a1auL5
Solder Wick: amzn.to/347kUVI
Files: amzn.to/37bMXp0
Dremel: amzn.to/3gIADj9
Headlamp: amzn.to/3gE5VYw
Pelican Case: amzn.to/3ncFpbg - Наука та технологія
![Prusa MK3s Pick and Place Extension [3D printed] 👌](http://i.ytimg.com/vi/Mx4swtSQmE0/mqdefault.jpg)
To compensate for different film sizes, you need to adjust the depthing between the gears. You're currently using rigid mounts, make the upper gear mount flexible so it can automatically compensate. Bias the upper gear so that it is preloaded against the lower when assembled.
Spring load the gear mesh via the top gear.
@@nicholasthon973 That is what I am thinking too. The good thing about 3D printed plastic is you can draw the bending parts into the plastic frame saving you the trouble of inserting a spring and adjustment knobs
AHHHH that's a cheeky idea, thanks for the suggestion! I'll try that!
@@nicholasthon973 Spring loading could also reduce variations in pull strength between different devices caused by manufacturing differences.
@@stephen_hawes where are the 3d files for this feeder? I want to combine it with my feeder to make the ultimate one.
Feeder's cheap enough that you could just keep it on the reel at all times and swap out both at once, for even more convenience
The sheer size of the community, and all the work they are doing is really shown in this video!
Another great video, Stephen - it must be so exciting having lots of people using the machine for real now! Amazing to see what cool ideas the community are coming up with too!
Never seen any of your videos, and didn't know about your PnP project. Video was just in my feed so I watched it. I must say, you have done some amazing work. As a maker myself, it really makes me happy to see your energy and love for the community you have built. I will be following along for more amazingness from you. Definitely earned a sub from me.
Superb! What a great community you've got around the machine. I'm working on a hardware product, and when we ever get to production, I'll definitely consider getting one.
That is so dope! Love the way you present topics and give shout outs to contributors. It's been super educational watching your videos.
For controlling servos I highly recommend PCA9685 16 channel driver. it can be chained together and control hundreds of servos with 2 I2C cables (up to 1024 I think). It also has external power input for servos themselves, and a capacitor to smooth the voltage. Furthermore it is well documented online. Although Arduino mega has a lot of pins, I am not sure if it is the best choice for controlling 5+ servos.
Can recommend PCA9685 for anythyng where you need many PWM channels, not only servos, amazing cheap IC.
I wondered when looking at your prev vids if using PNP head energy to advance parts had been explored, great to see an implementation!
The coolest thing about this design is the modularity.
Put some kind of index indentations in the profile it's mounted to to place it in the same spot every time, then just keep your reels on the feeders and swap them.
You could possibly also extend OpenPnP (or is that a function already?) that then tells you "ok, put 10k 0805 in slot 1, 100nF 0603 in slot 2" and so on - quick changes!
I really love this level of open source enthusiasm normally only reserved for software now expanding to hardware
Before it's just the reprap project now there's openPnP and the community is ever so alive
This proyect has such a great community, congrats to you all guys!
I like how you improve the product and give it back to the people who bought it. I want to go with a similar route with my company. More companies should do it. And eventually you come up with a new model, or other services.
Thanks for the shout out! Atanisoft, Infiniteneslives and I have some really cool stuff in the works!
Hi Stephen. About 25 years ago at the beginning of SMD craze I made up a hand PnP with feeders which automatically fed the tape forward when the pickup head was approaching the feeder. And it did this without a motor or fancy mechanism, simply using the downward or sideways force of the pickup head. Very simple and will remove the need for the head to do the dance around every time a component needs to be fed forwards. The mechanism consisted of 3 parts in its basic form. The head will simply need to go to where the component is and pick it up, nothing more. I also developed a simple means to take off the tape also only a few parts. I once for fun put a motor with a cam against the mechanism and I could spit out 18 (yep you read it right eighteen) components a second from this feeder. I can send some short videos if you can give me an email to link up with. I think this added to your feeder would be the ultimate as it would speed up the machine overall placing time.
Yay, the superhappy Stephen is back. In the last few videos you seemed a little less upbeat. I guess starting a company and shipping all those machines took its toll. But it is so cool to see a lot of people using it.
This! This is the first video I’ve seen from him, but his happiness and excitement is so contagious and genuine
Awesome design, thanks for the video!
Love the simple and effective design! Is it a problem for the head to now have lateral force on it? Have you done a repeatability test with it pulling the tape then placing?
This is all pretty new, no long term reliability testing yet, but it's definitely something we're working on.
I LOVE what you do and the way you do it. I never comment this on youTube, but you are genuinely THE channel I wish the most success upon. Will definitely get a machine once it's going well!!!
That is a damnably clever refinement of the passive feeder concept especially resolving the tape peel back. These are so cheap that with a little more design you could leave the heads attached to the tape/reel so that few if any parts are lost during setup.
Very cool concept! There's quite a bit speed lost there on extra movement, BUT if to advance tape by 3+ parts per single pull then it wouldn't be as pronounced.
I am really excited to see more open source products like this, I really hope more and more businesses follow suit as it will push technological progression even further than we are currently
This is so phenomenally cool. Much love to the maker community ❤
fitting all different sized tapes, you could use a curved slide for the bottom gear with the interface, and spring load it downward, that way it remains the same interface with the upper gear to keep the tape peeled back properly, but also fits any size being fed through, and self corrects no matter what is loaded
Hey Stephan great videos. I recently found your chanel and have watched pretty much all of them. I'm really keen on building my own pick and place machine and getting into the building and design of PCBs . As someone who is well versed and has alot of experience. What would be the best place to start?
Best regards
Matthew
With the passive feeders being pulled through either by the head itself or a new peg......is there a chance the tape can spring backwards slightly when the head/peg releases due to the small amount of tension it's under from the film?
Yes, it can happen. There is a "backoff" setting in openpnp to slowly release the grip on the tape hole before lifting the head.
Subscribed as soon as I saw this =) Is this a closed design? I want to try to print it =) I couldn't find the project files on GitHub. Thank you for your ideas and your work)
I love your mission! Let's get USA more proficient in electronics!!!
Have some fun with this idea. Get a piece of our 0.25 mm thick x 8mm wide spring steel say 20mm long. Bend it into a curved shape of a 90 degrees circle (1/4 circle). Put the 2 ends on a flat surface. Measure the length on the surface. Then press down the high middle curved part. Measure the new length. You will see that the length on the surface gets longer. If you get the right length then you can use this as a means to advance the tape forward an exact amount. Simply put this metal in a place on the feeder where a small ball bearing on the head runs over this metal on it's way to get the component. Much quicker way to move the tape forwards and there is no moving of the head/pin to move the tape forwards.
This is super cool, didn't even know this community existed.
Love it, maybe next version add a little ramp on the back side so it gets kicked down and out and doesn't fall on neighboring feeders and risk a foul?
Amazing. Your calculations for pulling just enough cover slip (film) for tape works great. Next what to do with the spent film. It just hangs in the air??? So I will stick to the idea of having dangling weights for the films - it works and is tidy.
I just watched the new video on passive feeders and the first idea that popped into my head was a ratchetting rack-and-pinion setup, where the head descending toward the part presses down on a sliding rack which rotates the gears to move the tape forward. By repositioning the initial height of the rack, you can set how far the tape advances. That will allow the feeder to work with any part spacing and no extra head movement is necessary; it only needs to move high enough before descending to clear the highest possible rack position. That would also completely remove any lateral pressure against the vacuum tip.
Does that sound reasonable?
Exactly what I was wondering about while watching! The sliding rack contact point could be adjustable to vary feed length so that the top doesn't have to vary its travel, and maybe some sort of cam type system so that the motion is complete before the vacuum tip reaches the component. The 2 movements it's doing here will be halving its possible production rate
Have you tried adapting this passive design with an additional powered driver gear (perhaps mounted underneath) to get a really simple powered version? You could even potentially design them in such a way that the driver motor is optional and could be added later, without needing to make a new set of feeders
Also, what about adding a dedicated pin right next to the suction head so you can advance a passive feeder while simultaneously picking up a part? Might not be super practical compared to a powered feeder, but it sure would look cool to watch!
Hi grate work, when will the files be online :-)
Neat concept , slight mod and could fit that to a Yamaha YM85S loading machine.
some ideas for handling the tape size issue:
- use a spring (maybe from like an old broken pen or smth, or just get a bulk order of similar cheap mini steel springs) and use those to tension it
- rubber band based tensioner
- set up the mounting system so it attaches through the gear so it uses its own weight to tension the gears
- add 1 or more gears that the tape has to go through to increase friction
- use a higher friction filament to print the gears, but use PET-G for the frame
man, i see in like 5-10 years there being equivilents to the speed boat race with people building rediculous high speed pnp machines. i want one
Just add a spring to the gears to tension them together, this will help grip the tape better
I would probably just add a single plate spring below the gears to press the tape against the gears
2:22 my feeder design at the upper right corner :ь
This is absolutely genius!
hiii I'm just so interested by what you do, it makes want to do similar awesome machines ! but one question ... how do you assemble and debug code from your Mac to your MCU ????? I just don't get it I've tried so much things I just can't figure it out !
Instead of having to re-adjust the gears spacing for every component tape, why not switch to friction wheels ? (A grooved cylinder with one or two o-rings to ensure a high friction coefficient) as such the system may be a little bit more resistant to being pulled by the picking tip, but would passively accommodate for any component tape of the right width.
For designing aspects, sorry I won't be able to participate in the project and develop it myself, I am currently preparing the competitive entry exams for the french engineering schools.
I would give the following recommendations :
Make the cylinder's groove inner diameter half or a full mm wider than the o-ring's inner diameter to ensure a tensionned fit, thus avoiding the o-ring slipping in his groove and reduce the spacing of the wheels (from the sum of their expected radius : groove radius +o-ring thickness) down a quarter or a half mm, you will turn a dual contact point (o-ring1/tape and tape/o-ring2) into a dual surface contact by squeezing localy the o-rings, ensuring a far better pulling
would love a reel system that has different same type parts ( say 100 different 0805's) in the reel in sequence to populate a board for my diy synthesizer pcbs
-one after the other instead of 1 reel per pcb component
If the tape gets pulled more than the reel, how does it not bind? After many parts are used, it seems like the length difference between the reel and tape would cause the gears to bind or something to foul up.
So, where I can find models to print those? Thanks
I can't find the version of the 0816 from infinitesneslives anywhere. Can you help?
please share the modified feeder model with reel holder! In original
jazzychad doesn't have a reel bracket as I understand Stephen modeled it.
Can you make a playlist just for feeders. Love you channel
Well you could make a gear that grips and moves the tape, same as your 'intelligent' design, but has a larger gear exposed where the head has a fixed gear to engage, that couples to the 'spooler' gear that sits on the same shaft. This way the basic principle is retained, with the addition of a third mechanical part that can be adapted to the spacing/tape dimension.
Why not use a tighter gear fit with a compliant mount for the top gear?
How about adding a leaf spring under the SMD tape that adjusts itself to the thickness of the tape.
So, people often forget the simplest solution is usually going to be the most reliable. Leave the gears fixed, introduce a multi-finger or single finger with steps in it compliant spring to accommodate different width/thickness of tapes. Now if the holes are different for different tapes, then thats a different issue and a different "series" of the multi-finger would need to be developed.
You should keep in mind that the reels won't always roll off perfectly. They sometimes start tugging or even seize up completely. This is a huge problem for the PNPs at my workplace, and it will only be exaggerated with passive feeders.
For different film-thickness, I would just print two smooth wheels in TPU which squeeze the tape, rather than gears
idea to keep the tape from skipping. have a pully with matchi g holes under the big gear that sandwiches the tape. that way it can't flex under the big gear spur teeth
I made a comment earlier about an open source passive feeder called push pull feeder. It was deleted or maybe automatically filtered for possible SPAM.
Will next version made on laser-cutter from acryl? :)
Have you tested to see if the nozzle gets bent over time?
Since the ratio of components used is fixed for each board, couldn't a single motor be used to advance enough components for each board at the same time between boards? A motor could advance the reel with the most used component, then gear reductions could advance the lesser used reels accordingly. A central rotating shaft could run along the feeder area, and gear boxes could be used to tap into this shaft and facilitate the correct number of turns for each component. If a ratio is higher than, say, 4:1, then multiple spools could be used rather than making every possible gear box ratio. This would be faster and reduce wear on the nozzle.
Just a thought. You're a big inspiration of mine!
And when you want to make a different board you have to change 20 gears?
Doesn't sound like a good idea.
@@Jehty_ make a selectable transmission. Nah... just advance each reel, one at a time, while the new board gets fed into position.
Maybe make the gear with two kinds of theeth, side by side. One conventional like now for rotation and one very spiky for more bite in the thinner tapes.
Or just replacable gears, color-coded for tape thickness
@@dancollins1012 From the video i got the sense that with 3D printed gears the tolerances aren't really 100% repeatable. If that is not the case just making the tightest gear mesh should work for any tape.
maybe you could suck the part and move the tape at the same time with a needle next to the nozzle, so it doesn't take more time
Hey, Stephen! How can I send you some design ideas?
Now do that at 16 components a second !
Amazing!
гениальное изобретение !
Awesome!
Just add a spring tensioner Like a 3d printer feeder for the thickness problem
Since the cost of the feeders is so low, why not just have different feeder tape width/thickness for different tape styles? And just label it in the 3dprint direct. I guess keeping an external width of the feeder to be consistent would be helpful so they can be easily swapped.
I am worried that the tip will wear a lot faster when used for this load it iss not designed for
Could use a little solenoid to poke a metal prog down to advance tapes. 6:40
@@TheRainHarvester I’ve seen designs that do this. Using the nozzle itself seems like a cheapness too far. Not that I’m an expert.
@@lawrencemanning one might get away without the solenoid\additional weight by making some kind of "click lever to advance" on the feeder. Put it where the nozzle doesn't touch anything where a shorter,fixed, protrusion could still reach the lever from the head.
I love it so much
How does one stay up to date on this project when one doesn't do Discord?
Follow the videos and GitHub I guess.? Consider doing discord if you really want to see the daily progress of the community.
@@InfiniteNesLives Every time I try discord I get a series of wave hi entries no information on hardware design.
is there a particular advantage that powered feeders have over a design like this? otherwise, it seems like this is as much as you would need!
For commercial operations, speed is everything. The extra time it takes to use the placement head to advance the tape would never fly.
@@MLSgeek there are many crappy commercial machines out there (aside from the big companies which charge much more).
Especially the feeders are really a key thing, if something's wrong components will be missing on the PCB and having PCBs with >1000 components it's just crap to check that with a naked eye.. even worse having to expect that something's missing is bad.
To me it seems Stephen has nothing on the feeder side at the moment.
@@MLSgeek You could cut down on the time waste if you pulled like 10 through at a time.
Is new design?
This system is normal for many many years from the label industry.
How do you grab a 3 mm x 2 mm or a 0402 part? vacuum? how does the machine find the part in the tape?
than how you align that part with the PCB footprint?
there is a problem, if you switch parts you have a whole bunch of loose components, lost?. that action should be place much shorter to start of the tape.
cool development.
It's pretty normal to need to feed quite a lot of tape when setting up a feeder, and manufacturers know this -- reels of parts always come with long empty leaders of just empty tape before the actual parts start.
@@AJMansfield1 i haven't bought a entire reel, yet. yeah, true, i heard to a reel starts "empty" tape, but what if you have cut tapes, with just 10 ic's or 100 100nF by pass capacitors?
All your questions were answered in his other videos.
So I would recommend that you watch them.
@@AnalogDude_ then you do not need automatic feeders
That thing is NOT pulling more film than tape... /multi facepalm/
WHAT IS PNP? i know only Peer Networking Protocol
This slows down the work. How about making a lever placed in front of the feeder that when pushed, advances the tape. The nozzle moving into position, where it takes the part will push the lever "on it's way" and advance the tape just before picking the part up.
Mother who?
believe PnP is already reserved for televisions buddy. maybe pNp
Because abbreviations can only mean one thing. 🤣👍👍
Зачем такие сложности? Делай проще! Грузик (прищепку, канцелярскую скрепку) на свободный конец ленты, и он сам будет убираться из рабочей зоны. $1 за тысячу.
Yeah but once the weight touches the ground you have to be there moving it up again. If you have 20 feeders that will be a lot of work.
всё уже украдено до вас ua-cam.com/video/ol07YgZOKgs/v-deo.html и каждые 10 минут надо гири подтягивать