Semi-automatic passive printable/castable feeder

[wormball]

It's great to see that my design is working for someone else!

1. I made the shaft for rotation a bit smaller as movement wasn't easy at first.

You may try metal bolt as in the alexfeeder. I considered this option but decided not to do so cos:
(a) i do not have M3/M4 bolts with unthreaded shank,
(b) i am not sure the thread will not grind the rotating plastic part,
(c) i do not know how to secure this bolt in the wall. Extra nut is not the best idea cos it eats space on the other side of the feeder. Maybe super glue is ok.
Maybe i will also try bolts as axles, but my nylon axles work fine.

5. Added a narrow lip covering the tape holes only to keep the tape in place (does not cover the removeable tape part or pick up location) (doesn't cover the front screw hole either )
6. Added a full width lip @8mm to guide the tape at the rear. Inserted hole for access to rear screw.

Good point, i think i will do it too. But now your feeder won't eat thick tapes. I think there is still room to increase the tape clearance. Or if you do not need thick tapes, you can delete the big trough that i drawn to accommodate pockets in plastic tapes.

Meanwhile i performed first production run! It consisted of only about 300 parts, but it helped me to figure out some things.



Good news:

1. It works with both paper and plastic tapes.

2. I did not need to adjust the finger length after it was originally determined.

Bad news:

3. I had drawn the feeders 0.2 mm narrower than the exact feeder pitch trying to compensate my printer's inaccuracy, but it had been printed surprisingly precise, and also the castings happened to be shrunk by another 0.1 - 0.2 mm from the master model, and moreover, my screw template was printed with 12.05 mm pitch instead of 12 mm. So i have about 0.5 mm play in the tape position.

4. One feeder is hard to tune (while others required no tuning at all). It either causes parts to jump sometimes, or it advances the tape with some struggle in case i decrease peeling weight or bend the flat spring towards the bed. I think it is due to polyurethane being not as slippery as desired, and other feeder instances do not suffer from this issue cos release agent residues act as lubricant. So it may be a problem when all the release agent will go.

I think i will try to make a design that does not need the flat spring at all so it won't need much force to advance the tape.

[Oaklands]

But now your feeder won't eat thick tapes

Agreed, but the good thing about printing (vs casting) is that if a thicker/deeper tape is needed, we can always reprint to fit .... i.e move the lip up a bit to accommodate the thicker tape (or make the base thinner or provide a deeper channel).
I was indeed wondering what that trough was for & now I know I will leave it in for now as it probably reduces some friction.

Other changes I have made since last post etc:

1. I got 4 springs delivered at the weekend - not stainless steel though, I will have to wait up to 2 months to get the stainless steel ones from China. However, the current springs were not strong enough to advance the tape fully into the next position. So I had noticed that when shipped the springs had merged into each other. This gave me the idea to overlap 2 springs together and try again. This made a stronger spring and this now advances the tape as expected. I expect the stainless steel springs you linked will be stronger, hopefully. If not I will have to reduce the friction a bit more.

2. In the end I also made the pivot hole in the arm a bit bigger with a drill, to free up the movement and this seems to be much better as well. (#2 in image below)

2019-10-30 14_38_40-3D design Feeder_Spring _ Tinkercad.png (282.85 KiB) Viewed 24986 times

3. I fixed the arm to the body with a screw and washer. I also used my soldering iron to heat up the installed screw, which made it more secure as the PLA melted around it like a glue. It may be worthwhile designing in a hole to accept a screw to retain the arm. The screw/washer stops the arm from sliding out from the body.

4. Just as I was writing this I was manually testing one of the 3 LiteFeeder units I printed. Unfortunately, the arm broke off at the thinnest point at the rear. (#1 in image above) I then checked another unit and it also seem to have a stress fracture in the same position.(I suspect the lifespan is about 1000+ components based on the manual testing I did. So I guess 3D printed PLA won't last long for this design rev. I expect, designing in some sort of hinge or joint in the PLA at this position (held together with thin wire or a pin) for this will solve it. I will revisit this when the springs are delivered from China. (Obviously the resin casts would be much sturdier at this weak point than printed PLA.)

5. I also printed out a push rod to attach to the M3 screw (20% infill). I don't have long M3 screws so I just made this taller to make up the gap. I have not installed it on the LP yet.There is a slot included to take an M3 nut for height adjustment and another nut on top can be used to lock it in place after adjustment.

2019-10-30 14_26_11-3D design PnP_LiteFeederPushRod _ Tinkercad.png (229.66 KiB) Viewed 24986 times

I think i will try to make a design that does not need the flat spring at all so it won't need much force to advance the tape.

You will still need something to pull the cover tape over. I was thinking about a loose rod going the full length of all feeders thru a hole in the feeder body, just where the flat spring touches the tape. It could act a s a roller and keep the tape in place.

[wormball]

Unfortunately, the arm broke off at the thinnest point at the rear. (#1 in image above) I then checked another unit and it also seem to have a stress fracture in the same position.

Maybe it is cos you pushed the handle more than it will be pushed during normal operation. Also it is pretensioned in the file, but it is actually not needed so much cos when the handle is pushed, the plastic spring is being additionally bent receiving the tension needed to push the tooth into the hole, and while it is moving forward, it is held in position by the tooth. So you can print the dog in the conformation close to the actual working position like in the attached file (i do not know why i had not uploaded it to the github).

Also you can make the distal end of the plastic spring slightly thinner so it will suffer less (and more uniform) tensile stress.

I expect, designing in some sort of hinge or joint in the PLA at this position (held together with thin wire or a pin) for this will solve it.

I think it will make things more complex cos you need not only the hinge but also some external spring.

I also printed out a push rod to attach to the M3 screw (20% infill).

Why is it so wide? The feeder pitch is 12 mm.

[Oaklands]

Thanks for the file upload.

Maybe it is cos you pushed the handle more than it will be pushed during normal operation....

very possible. The first 2 prints, got a lot of test cycles using my hand to press down rapidly.

Why is it so wide? The feeder pitch is 12 mm.

Indeed I made the lower part 15mm, but cannot remember why...probably just not thinking. That will be easy to fix.

I wasn't originally thinking of a spring for the 'possible' hinge (Hoping gravity would suffice). But you may be right that spring would be required to have it operate as it currently does. A rubber band may be enough? However, I will know better when I design/print it later, when the springs are delivered.

I haven't installed it on the LP yet, as it will be a while before the springs arrive from China.( Based on tracking the flat spring should be here in a few days.)

[ds23man]

Any news on this?

[Swimmer]

Very impressive feeder design. Looking forward to hearing more about your progress.

[wormball]

Any news on this?

I did nothing, but Jed Smith made film peeler for it, so now it is fully automatic! https://groups.google.com/d/msg/openpnp ... kAq5XVAgAJ https://loonatec.com/pnp-semi-automatic-feeder/ https://bitbucket.org/Loonatec/pnp_semi ... ic_feeder/ However i had not tried it yet.




Also i encountered some other feeder with even more simple peeler design. https://groups.google.com/d/msg/openpnp ... HSaWfxEgAJ https://github.com/ploopyco/pnp-feeder/wiki https://www.youtube.com/watch?v=qa37CS5gLkE Maybe i will try this too (however more strong spring may be needed).

[Krzys]

Jak ustawić liteplacer, by pobierał elementy z ploopyco... urządzenia.?
Musi najpierw nacisnąć dźwignię, a później pobrać element.
Czy można to zrobić na Liteplacer?
Krzysztof

[wormball]

How to set up liteplacer to download elements from a ploopyco ... device?
He must first press the lever and then download the element.
Can it be done on Liteplacer?
Christopher

I think it is impossible with the current liteplacer software, and you have to use openpnp instead.

[AnalysIR]

I think it is impossible with the current liteplacer software

Is it not just a pickup at a fixed location every time? (in Current LP s/w). The lever offset ensures the part is advanced every time??

If so, configure the tape for 'Use coordinates directly, with a pitch of 0.

2020-03-14 14_25_01-LitePlacer.png (53.79 KiB) Viewed 24136 times

...or am I missing something?

I presume you could also use the Tape hole recognition, i.e. Place action with 0 Pitch, if you wanted to???
However, the position should be fixed, so no real need to run Vision processiing.

(FYI: I havent tested this)