Alternative Feeder Design

[Greg Wroblewski]

I studied different homemade feeder designs and it appeared to me that it should be possible to design one that could be 3D printed as a single piece (more or less, not counting the sprocket). Here is my first attempt:

https://www.youtube.com/watch?v=3FEUZaMgOpI

https://github.com/gwroblew/CoolPnP/tree/master/feeder

I used this motor for the tape advance:

https://www.robotdigg.com/product/1160/35BYJ46-12V-pm-stepper-motor

and this one for the cover tape collecting:

https://www.aliexpress.com/item/Hot-Sales-10pcs-Micro-DC-motor-gear-reducer-for-DIY-3V-5V-6V/32655301826.html

(driven by 5V). Both motors are controlled by ULN2003 drivers bought on Amazon. I found a spare Atmel SAME70 dev board with tons of IOs and ethernet port, so that made a nice hacky controller for 10 feeders.

Both the main body of the feeder and the sprocket are 3D printed. The feeders are really meant for my next machine, but I managed to test them on my franken-pnp, although it did not have proper space for tape cover collecting. So far I placed a few hundred capacitors and resistors without a single glitch. The key to stability is matching of the tape guide shape (width and thickness) to the actual tape it is meant to be used for. I have 5 variants, with tape openings varying from 0.8 to 1.6 mm. I even tested them on 0402 resistors and there was no part jumping.

The 35BYJ46 motor is quite nice. It could be a little bit faster and stronger, but it does the job. It clearly needs a feedback mechanism. My franken-pnp has very poor repeatability, so I have to use vision before picking a part anyway, and that gives me information how far the part is from the expected place. I use it and correct the number of feeding steps as needed. It takes about 170 steps to move the tape 4 mm (~0.3 second). With the correction I have seen 160 - 190 steps. The tape cover collecting motor gets enabled for 120 ms.

This is very much work in progress. It could become a very nice, well below $10 feeder.

Here are my ideas for the final design:

1. Use optical sensor for tape driving motor for feedback (so we could rely on the feeder always delivering the part in the same spot).

2. Use a tiny PIC or Atmel with dedicated small control board per feeder.

3. Use I2C or CAN to connect multiple feeders in a bundle.

4. Cut out sprockets from an aluminum sheet on a CNC, attach it with a 3D printed mount.

5. Detect stall on the tape cover collecting motor (not sure how to do it, or whether it is really needed).

6. Separate the feeder body into driving part and tape guiding part. The driving part would stay put, while the tape guiding part could be attached and detached easily.

Let me know if you have any questions or suggestions for the next version.

Greg

[Greg Wroblewski]

Here is a picture showing how the two 8 mm feeders interlock. The width of a single one is 15 mm, and it looks like it could be brought down to 14 mm.

Greg

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/ff2LNvpnTIw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/85c65758-d1a7-49f6-9228-cf312ffb2689%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Maple_Dude]

 looks like a good approach 

[Brynn Rogers]

Can you do a video of the camera focused on the feeder through a couple complete cycles, and don't move the camera around?    This video shows the picking and placing nicely on your machine, but I can't (even after forcing 1080) see the feeders advance or any other detail of the feeders themselves in operation which you are trying to show us.

Brynn

[ma...@makr.zone]

Hi Greg

Its hard to understand how it works, most of the action is hidden...

Can you do a video of the feeder not mounted inside the machine?

And/or can you post CAD screenshots and explain how the parts interact?

I looked at the OpenSCAD files, but I can't figure it out (there seem to be different versions of the same parts, some are "work in progress"?). I think few people will download and try to understand all that, so explaining a bit more might improve the chances of getting good comment. :-)

Thanks,

_Mark

[sebastian....@gmail.com]

As for the tape detection.

How will you detect holes in transparent tape?
Perhaps the transparent plastic is opaque for IR but I would prefer a solution with a solutin with a Panasonic AV4 microswitch.

About electronics per feeder.
There are already feeders with integrated PCBs.. Like this one
Integrating electronics is a great way to a nice and clean Feeder build.
But complex electronics will probably keep people from building your feeder unless someone generates a nice build tutorial.

Greetings

Seb

[Greg Wroblewski]

Here is a more detailed view and a few feeding steps on 0603 1k: https://www.youtube.com/watch?v=fVr174z0mhE

I also added some information to the readme file on github. The tape cover removal in my current setup is somewhat lame, because of the space constraints.

There isn't much to this feeder: the body piece (it should be printed vertically with supports-on-bed only, with the tape entry laying on the bed), sprocket, and motor. The matching of the slot size for a given tape makes it possible. With the slot too tight the motor cannot do the pulling (maybe a stronger one could), with slot too wide (thickness-wise) parts start jumping.

This feeder is very close to what I have in mind for the final design.

I noticed that the commercial feeders are very accurate. It appears that industrial machines rely on them to deliver parts in the same place at least with 0.1mm accuracy.  I would rather not use holes for positioning the tape, because tapes are flexible and mechanical switches are not as accurate as it seems (my own experience + a number of people mentioned it on this group). I could use switch or optical sensor at the bottom or on a side of the sprocket. Not sure which one would be more accurate.

I wonder if the tape advancing motor I use would be sufficient for wider and heavier tapes, up to for an example 44mm wide with a large 5mm high part.

Greg

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/ff2LNvpnTIw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/d25514a3-f9e9-4d1d-9901-9ce7b7830a22%40googlegroups.com.