Feeder sprockets

[Brad Parcels]

Howdy guys and gals,

  Correct me if I'm wrong, but by way of a quick search or two, it looks as though the feeder sprocket "problem" is still sort of open?

 

  I was just wondering if anybody has tried using our old friend FR4 as the sprocket material? I was thinking that if you round out the base of the teeth so that it can use standard routing (i.e. at a cheap PCB Fab), you could just slap that between two discs that provide the base (either printed, or FR4 as well), and screw/glue it all together:

 

 

 

For a 30mm diameter x 1mm, it would only be about $50 for 100(!) of them delivered from PCBWay – and half of that is the DHL charge. (Not to suggest that’s the place to use – I just picked a fab at random)…  This is assuming they don’t surcharge for the routing, but I have seen at least some intricate routing in their sample pictures, so I’m optimistic.

 

Given that it’s a PCB, you could also get a bunch of customization and complexity for free – i.e. a keyway, holes to screw together, holes for an opto-sensor, tracks for some kind of drag contact or capacitive sensing, routing a side disc as a 2GT belt pulley, etc. etc.

 

At that price, I’m tempted to give it a shot the next time I put a PCB order in – unless somebody has already gone down this path and can advise against it? Or has a good/cheap/quick source been found for the real deal that makes this all for naught? Barring that - with all the expertise here, is there any suggestions on the profile, diameter, width, and the like? Has there been any sort of standardization amongst what people have been using?

 

Cheers,

Brad

[Anthony Webb]

dang PCB is a GREAT IDEA! @peter

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@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/f9844215aa2344dea245bf9b75db96b9%40viopix.com.
For more options, visit https://groups.google.com/d/optout.

[Jason von Nieda]

That is a genius idea, and first I've heard of it! Let us know how it turns out!

Jason

On Mon, Dec 4, 2017 at 7:12 PM Anthony Webb <anthon...@gmail.com> wrote:

dang PCB is a GREAT IDEA! @peter

On Mon, Dec 4, 2017 at 6:09 PM, Brad Parcels <bpar...@viopix.com> wrote:

Howdy guys and gals,

  Correct me if I'm wrong, but by way of a quick search or two, it looks as though the feeder sprocket "problem" is still sort of open?

 

  I was just wondering if anybody has tried using our old friend FR4 as the sprocket material? I was thinking that if you round out the base of the teeth so that it can use standard routing (i.e. at a cheap PCB Fab), you could just slap that between two discs that provide the base (either printed, or FR4 as well), and screw/glue it all together:

 

 

 

For a 30mm diameter x 1mm, it would only be about $50 for 100(!) of them delivered from PCBWay – and half of that is the DHL charge. (Not to suggest that’s the place to use – I just picked a fab at random)…  This is assuming they don’t surcharge for the routing, but I have seen at least some intricate routing in their sample pictures, so I’m optimistic.

 

Given that it’s a PCB, you could also get a bunch of customization and complexity for free – i.e. a keyway, holes to screw together, holes for an opto-sensor, tracks for some kind of drag contact or capacitive sensing, routing a side disc as a 2GT belt pulley, etc. etc.

 

At that price, I’m tempted to give it a shot the next time I put a PCB order in – unless somebody has already gone down this path and can advise against it? Or has a good/cheap/quick source been found for the real deal that makes this all for naught? Barring that - with all the expertise here, is there any suggestions on the profile, diameter, width, and the like? Has there been any sort of standardization amongst what people have been using?

 

Cheers,

Brad

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, 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/f9844215aa2344dea245bf9b75db96b9%40viopix.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, 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/CALsNZy3zQxu4zPc%3DeocvkdK4%3D36yrTRVP8tKqML0hkXfXnvS7Q%40mail.gmail.com.

[Bernd Walter]

On Tuesday, December 5, 2017 at 2:09:15 AM UTC+1, Bradley Parcels wrote:

Howdy guys and gals,

  Correct me if I'm wrong, but by way of a quick search or two, it looks as though the feeder sprocket "problem" is still sort of open?

 

  I was just wondering if anybody has tried using our old friend FR4 as the sprocket material? I was thinking that if you round out the base of the teeth so that it can use standard routing (i.e. at a cheap PCB Fab), you could just slap that between two discs that provide the base (either printed, or FR4 as well), and screw/glue it all together:

 

 

 

For a 30mm diameter x 1mm, it would only be about $50 for 100(!) of them delivered from PCBWay – and half of that is the DHL charge. (Not to suggest that’s the place to use – I just picked a fab at random)…  This is assuming they don’t surcharge for the routing, but I have seen at least some intricate routing in their sample pictures, so I’m optimistic.

 

Given that it’s a PCB, you could also get a bunch of customization and complexity for free – i.e. a keyway, holes to screw together, holes for an opto-sensor, tracks for some kind of drag contact or capacitive sensing, routing a side disc as a 2GT belt pulley, etc. etc.

 

At that price, I’m tempted to give it a shot the next time I put a PCB order in – unless somebody has already gone down this path and can advise against it? Or has a good/cheap/quick source been found for the real deal that makes this all for naught? Barring that - with all the expertise here, is there any suggestions on the profile, diameter, width, and the like? Has there been any sort of standardization amongst what people have been using?

Using a PCB manufactorer is not a completely new idea, but rethink using FR4.
On a website I've seen aluminium with reflective pattern to trigger position feedback.
You could use an alucore with black soldermask for a reflective light sensor.
PCBway also offers alucore in different thicknesses for a good price.
I once asked them on twitter if they could build this using DXF template, but got no feedback, but I think my question just got lost.
I didn't follow that idea any further because I've got hold on a bunch of industrial feeders and learned that a non mechanical
endstop won't work very well and milling aluminium sheet is something I could do myself.
Probably a ratchet mechanism with a sandwiched alu wheel would be a better idea.

Your diameter looks like it is on the small side.
The siplace feeder I have at home has a sprocket wheel diameter of about 55mm (not exactly measured).
My assumption is that going much smaller is asking for trouble.

Btw:
The older siplace feeders don't use a sprocket wheel at all, they use drag system with a set of pins.
The dragging is controlled by a cam wheel, which has a range on non movement to allow the DC motor to stop after being triggered by a microswitch.
The most amazing thing is that they run with a simple transistor/relais statemachine and a single DC motor for 2 lanes.
Clever design, but not very DIY friendly I think and the drag design can only transport full 4mm.

[Bradley Parcels]

Hey Bernd,

   Thanks for the input... Good call on the aluminum core - roughly doubles the price, but in this case, that's still under a buck a piece delivered... I guess it will all boil down to how much they care about the complex milling (of either material).

   I had seen somewhere that the minimum bend radius for most tape was in the ballpark of 25mm (thus 50+mm diameter) - but what I had in mind wouldn't actually bend the tape around the sprocket - just engage tangentially to the circumference... Anyone know any reason such a setup would still require the larger diameter? I realize that it would require a different tooth profile than the larger, but save that, I couldn't really think why it would matter? I'm guessing that on the siplace feeders, the tape wraps that sprocket to some extent?

    In any case - if anybody happens to have easy access to a proper sprocket (i.e. have a feeder disassembled already - please don't sweat it if you'd have to tear things apart!), and could be convinced to throw it on a flatbed scanner or something - I'd likely go with whatever diameter we can get a "known good" profile from... I was thinking of the smaller diameter purely for torque and cost reasons... but I'd almost certainly go with a known good profile (at whatever diameter) over having to potentially iterate a smaller custom design!

Cheers,

Brad

   

   

[Jason von Nieda]

I have some Yamaha CL feeders and an Neoden 4 feeder that I'll be happy to measure. It doesn't look like the Yamaha's sprocket comes out easily, but I should be able to measure it's diameter at least. The Neoden 4 one comes apart easily, so I'll get good dimensions on that one. Can't do it tonight, but will keep it on this list for this weekend.

Jason

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, 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/1adc496f-f3f4-4c7f-855c-d0567f605de0%40googlegroups.com.

[Brad Parcels]

Fantastic – thanks so much Jason! A few high-res pictures above/beside some graph paper for scale (or a banana – your call) could even save you some of the measuring (if you happen to have any on hand)…

 

Cheers,

Brad

 

Yes, that was a 12 year old meme joke… My sincere apologies!

[Bernd Walter]

On Tuesday, December 5, 2017 at 4:27:54 AM UTC+1, Bradley Parcels wrote:

Hey Bernd,

   Thanks for the input... Good call on the aluminum core - roughly doubles the price, but in this case, that's still under a buck a piece delivered... I guess it will all boil down to how much they care about the complex milling (of either material).

That's true.
With FR4 they can use a diamond shaped endmill, which puts almost no load on the mil.
With alu they stresss the mill and endmill way more.

   I had seen somewhere that the minimum bend radius for most tape was in the ballpark of 25mm (thus 50+mm diameter) - but what I had in mind wouldn't actually bend the tape around the sprocket - just engage tangentially to the circumference... Anyone know any reason such a setup would still require the larger diameter? I realize that it would require a different tooth profile than the larger, but save that, I couldn't really think why it would matter? I'm guessing that on the siplace feeders, the tape wraps that sprocket to some extent?

No - the tape runs straight over the sprocket wheel - same for the Yamaha pneumatic, which is the main point here.
Only a 1-2 sprockets actually handle the tape, the others are at least so far out that they won't handle a load.
Also the pick position is exactly on top of the sprocket wheel.
The tape gets slightly bend on the way to the sprocket wheel and then bend down after the wheel on the way out.
The sprockets have to go in and out of the straight tape, which is easier with a big diameter.
I've added some detail pictures, some have the shutter removed to get better vision to the sprocket.

[Bernd Walter]

On Tuesday, December 5, 2017 at 4:57:58 AM UTC+1, Bernd Walter wrote:

On Tuesday, December 5, 2017 at 4:27:54 AM UTC+1, Bradley Parcels wrote:

Hey Bernd,

   Thanks for the input... Good call on the aluminum core - roughly doubles the price, but in this case, that's still under a buck a piece delivered... I guess it will all boil down to how much they care about the complex milling (of either material).

That's true.
With FR4 they can use a diamond shaped endmill, which puts almost no load on the mil.
With alu they stresss the mill and endmill way more.

   I had seen somewhere that the minimum bend radius for most tape was in the ballpark of 25mm (thus 50+mm diameter) - but what I had in mind wouldn't actually bend the tape around the sprocket - just engage tangentially to the circumference... Anyone know any reason such a setup would still require the larger diameter? I realize that it would require a different tooth profile than the larger, but save that, I couldn't really think why it would matter? I'm guessing that on the siplace feeders, the tape wraps that sprocket to some extent?

No - the tape runs straight over the sprocket wheel - same for the Yamaha pneumatic, which is the main point here.
Only a 1-2 sprockets actually handle the tape, the others are at least so far out that they won't handle a load.
Also the pick position is exactly on top of the sprocket wheel.
The tape gets slightly bend on the way to the sprocket wheel and then bend down after the wheel on the way out.
The sprockets have to go in and out of the straight tape, which is easier with a big diameter.
I've added some detail pictures, some have the shutter removed to get better vision to the sprocket.

Just added another picture showing the tape path over the sprocket from the side.
Normally the tape would be pressed against the wheel, but I had to remove the shutter mask for the picture.
IIRC the Yamaha are similar on their sprocket wheel usage, but I don't have one at home.

   

[Bernd Walter]

On Tuesday, December 5, 2017 at 4:27:54 AM UTC+1, Bradley Parcels wrote:

Hey Bernd,

   Thanks for the input... Good call on the aluminum core - roughly doubles the price, but in this case, that's still under a buck a piece delivered... I guess it will all boil down to how much they care about the complex milling (of either material).

   I had seen somewhere that the minimum bend radius for most tape was in the ballpark of 25mm (thus 50+mm diameter) - but what I had in mind wouldn't actually bend the tape around the sprocket - just engage tangentially to the circumference... Anyone know any reason such a setup would still require the larger diameter? I realize that it would require a different tooth profile than the larger, but save that, I couldn't really think why it would matter? I'm guessing that on the siplace feeders, the tape wraps that sprocket to some extent?

Yes - I think the tooth profile is the issue here.
The siplace sprockets have a (almost?) straight side profile in the tape and only the top, which goes through the tape, is narrowed down.
Well - you have to know that the siplace feeders have no reel holder, the reels are stored in slots and the feeder needs to be able to pull hard on the tape.
So maybe the Yamaha sprocket wheel is smaller because they can run with a different tooth profile - I may have a chance to check that tomorrow.

[Brad Parcels]

Awesome – thanks Bernd – very much appreciate all the closeups and the extra details!

 

Beyond the sprocket (which between your comments, and Jason’s ability to provide some good dimensions - I’m more than happy with being 50+mm) - I’m also quite curious as to exactly where the tape contacts the feeder (other than the sprocket obviously), and where (if at all) spring pressure is applied?

 

To clarify: I’ve been experimenting with simple “rounded rail” profiles on either side of the tape, to guide it in (while leaving room for parts in plastic tape):

 

 

Versus pulleys/bearings/etc to guide the tape in… I’m just curious as to how it’s done “properly”?

 

Also – to accommodate both the thicker paper tape, as well as the really thin plastic tape, I’ve always just assumed that they must have a spring pressing it into the sprocket? Is it just pressing the tape down onto the sprocket, or vice versa? Is the tape just pressing into a flat plate or something with features on the underside? I haven’t managed a really robust/reliable mechanism that accommodates all different types of tape yet, so I’ve resorted to an interchangeable baseplate that has grooves of different depths for paper vs. plastic tapes, and for each width:

 

 

Finally, if I can trouble you (or anybody) to show some good closeups of the exact point where the cover-tape is peeled back (and how it’s pulled), that would help a ton too… Some of my attempts to replicate the “finger feeder” (or various other names – the type that leaves the cover attached to the tape on the sprocket side, and just lifts it enough to access the parts) are also visible above… I can get it working great for any specific tape, but it seems to need tweaking for each particular case (even between the same size tape/part between different manufacturers)… It also needed a lot more torque than simple peeling… All in all, it seemed like a bigger headache than it was worth, so I’ve pretty much resigned myself to the more typical peeling (be it slipping spools on a common shaft, or otherwise… at the moment, I’m kind of leaning towards keeping each feeder “compartmentalized”)… Either way - I’d love to see as many examples of how it’s done as possible (and take the best of each)!

 

Cheers,

Brad

--
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/B7-6KBnYbVI/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/f28f6d9d-ed21-4e82-b8f2-fbe9d1115f6f%40googlegroups.com.

[Michael Anton]

I never thought of using FR4, and I think it could work ok.  Most fabs seem to use a 1mm dia router bit, for routing the perimeter, or at least you should be able to rely on that.  So that means you can work with 0.5mm radii, which isn't too bad.

I've attached a step file of a sprocket I came up with, and had laser cut out of 1mm thick stainless steel.  It looks like it should work ok, but I haven't completed the feeder, so I've never really tried it.  Probably laser cutting these out of regular steel, and then nickle plating it would be pretty inexpensive as well.  Stainless is ideal, but it is quite a bit more expensive to cut, since it requires far more power.  This sprocket was modeled after the ones used on the feeders on my ECM machine, at least as far as the diameter goes, though I think this design is a tad bigger in diameter.

You may not be able to sandwich the PCB sprocket between two other pieces if you plan on using plastic tape (unless the material used is thin), as there is only about 0.8mm from the sprocket side to the tape pocket.

[Michael Anton]

Here is a close up of how the ECM feeders do cover tape peeling.  The tape is held down with a spring loaded tape peeler on the left (it slides behind the plastic piece that the tape sits on), the tape is wrapped around the sprocket a bit, and there is a spring that holds the now empty tape to the sprocket on the right.  The cover tape is wound up on the spool on the left, which is driven by the same ratchet mechanism that drives the sprocketKeeping tension on the tape like this is important, or parts tend to jump all over the place, when picking or feeding.

I have a design for strip feeders that is spring loaded, and it works far better than any non-sprung strip feeders that I've used in the past.  Someday I should share this one.

On Monday, December 4, 2017 at 10:00:46 PM UTC-7, Bradley Parcels wrote:

Awesome – thanks Bernd – very much appreciate all the closeups and the extra details!

 

Beyond the sprocket (which between your comments, and Jason’s ability to provide some good dimensions - I’m more than happy with being 50+mm) - I’m also quite curious as to exactly where the tape contacts the feeder (other than the sprocket obviously), and where (if at all) spring pressure is applied?

 

To clarify: I’ve been experimenting with simple “rounded rail” profiles on either side of the tape, to guide it in (while leaving room for parts in plastic tape):

 

 

Versus pulleys/bearings/etc to guide the tape in… I’m just curious as to how it’s done “properly”?

 

Also – to accommodate both the thicker paper tape, as well as the really thin plastic tape, I’ve always just assumed that they must have a spring pressing it into the sprocket? Is it just pressing the tape down onto the sprocket, or vice versa? Is the tape just pressing into a flat plate or something with features on the underside? I haven’t managed a really robust/reliable mechanism that accommodates all different types of tape yet, so I’ve resorted to an interchangeable baseplate that has grooves of different depths for paper vs. plastic tapes, and for each width:

 

 

Finally, if I can trouble you (or anybody) to show some good closeups of the exact point where the cover-tape is peeled back (and how it’s pulled), that would help a ton too… Some of my attempts to replicate the “finger feeder” (or various other names – the type that leaves the cover attached to the tape on the sprocket side, and just lifts it enough to access the parts) are also visible above… I can get it working great for any specific tape, but it seems to need tweaking for each particular case (even between the same size tape/part between different manufacturers)… It also needed a lot more torque than simple peeling… All in all, it seemed like a bigger headache than it was worth, so I’ve pretty much resigned myself to the more typical peeling (be it slipping spools on a common shaft, or otherwise… at the moment, I’m kind of leaning towards keeping each feeder “compartmentalized”)… Either way - I’d love to see as many examples of how it’s done as possible (and take the best of each)!

[Bernd Walter]

On Tuesday, December 5, 2017 at 6:00:46 AM UTC+1, Bradley Parcels wrote:

Awesome – thanks Bernd – very much appreciate all the closeups and the extra details!

 

Beyond the sprocket (which between your comments, and Jason’s ability to provide some good dimensions - I’m more than happy with being 50+mm) - I’m also quite curious as to exactly where the tape contacts the feeder (other than the sprocket obviously), and where (if at all) spring pressure is applied?

I hope the pictures explain it.
This time a tape is fully loaded with cover tape inserted into transport mechanism.
The exact mechanics are specific for this siplace design, but the principle is the same for the older siplace and Yamaha.
You have a metal part over which the tape gets removed, then it goes to the pick mask, on some feeders the pick mask have a shutter, like this feeder.
Everything is hold down to set an upper barrier for the tape - in this case with an extra part from the front side with a string spring, but the idea is just that it is pushed down.
The older siplace don't have a spring, just a locking mechanism to hold the mask down.
The whole reason for the mask to be removeable is to ease tape insertion.
The mask ensures that the tape sits on the sprockets and defines the pick height.
In some feeders there is a springy metal bar (not in this case) to push the tape up against the mask.
E.g. my Yamaha CL8-2 (for 0402 parts) has such a bar, but my CL8-4 do not, although the body frame has the tapered mounting hole for such a thing.
The tape has some loose side alignment, but I've noticed that in the Yamaha, as well as the older siplace feeders it can move a bit sideways, which adds to the pick offset.
I assume for industrial machines it is mostly important to pick the part and if required align per camera instead of ensuring a fixed pick position.
The modern machines are designed to have the camera in the path anyway.
On the siplace star head, which is a carussel system with AFAIK up to 20 nozzles, the nozzles runs over the camera while picking parts, so they get the camera vision without additional time.
Depending on the tape the parts sits more or less loose in the tape anyways and the is nothing you can do about it.

That said:
I have run my older siplace and Yamaha feeders with OpenPnP without bottom camera.
0402 from Yamaha needed some finetuning at start but stayed the same over the 20 PCBs at least.
The PCBs only had a single 0402 cap on them, most of the passives had been 0603.
Just a few 0603 had been a bit off to solder properly (there had been many 0603), one of the 4x0603 arrays had been too off as well - IIRC it failed on 4/20 PCBs.
1206 and tantal A and D had been perfect, even when picked from plastic tape (which have more room for the parts to move in the tape).
Even SOT-23 went fine.
MFP optocouplers were off.
So all in all it seems that 4x0603 arrays need bottom vision, the MFP4 need bottom vision too, plus 0402, even when everything went fine with the 0402 for me.
And maybe 0603 too, but it wasn't a big issue.
I didn't place any bigger ICs by machine and left them for hand placing because of missing feeders and proper vision setup for all the parts.
I'm not running the newer siplace yet, of which I took the pictures.

[Brad Parcels]

Thanks Michael - much appreciated! Geeze this group is awesome!

 

I agree – laser cut metal is pretty much the ideal answer for this (in which case you may not need to layer things to make the base either - depending on the kerf)… But my thinking was that until we’ve got the “perfect” feeder design (or at the very least – a “perfect” sprocket) nailed down, the PCB Fab would offer a great functional prototype opportunity, while being easily accessible to everybody… and at a low cost to boot! My suspicion (or maybe my hope) is that at least for those not doing serious production work - FR4 might even be all that’s ever required…

 

You are absolutely right about the “inner” disc for the plastic tapes though… Although you can theoretically get away with 0.8mm with most of the plastic tapes I’ve played with – but my experience has been that any wobble or misalignment makes for rubbing etc… Fortunately, even a “one-sided-sandwich” seems to have been working pretty well:

 

 

Cheers,

Brad

[Brad Parcels]

 Awesome – this exactly the kind of thing I was looking for – thanks Michael!

 

I can’t really tell with the shadow – is it just a flat plate pushing the tape up into another flat plate at the top?

 

Thanks again,

Brad

 

From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Michael Anton
Sent: December 4, 2017 10:27 PM
To: OpenPnP <ope...@googlegroups.com>
Subject: Re: [OpenPnP] Re: Feeder sprockets

 

Here is a close up of how the ECM feeders do cover tape peeling.  The tape is held down with a spring loaded tape peeler on the left (it slides behind the plastic piece that the tape sits on), the tape is wrapped around the sprocket a bit, and there is a spring that holds the now empty tape to the sprocket on the right.  The cover tape is wound up on the spool on the left, which is driven by the same ratchet mechanism that drives the sprocketKeeping tension on the tape like this is important, or parts tend to jump all over the place, when picking or feeding.

 

I have a design for strip feeders that is spring loaded, and it works far better than any non-sprung strip feeders that I've used in the past.  Someday I should share this one.

 

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/a997a89e-19c4-4b42-a6c2-db8a26767824%40googlegroups.com.

[Brad Parcels]

Awesome – thanks once again Bernd… You guys have given me much more in an hour than a week or so of google image searching!

 

Cheers,

Brad

 

From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Bernd Walter

Sent: December 4, 2017 10:43 PM
To: OpenPnP <ope...@googlegroups.com>

Subject: Re: [OpenPnP] Re: Feeder sprockets

--

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/B7-6KBnYbVI/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/dcef66ef-4113-4b8f-aa51-bef7415ff653%40googlegroups.com.

[Michael Anton]

The flat plate at the top pulls down on the tape.  Actually, it is not flat, as the metal piece below the black plastic part is part of the same piece, i.e. it is U shaped.  The black plastic part below the tape (which is U shaped to accommodate plastic tape) is fixed, and doesn't move.  If you see my post here: https://groups.google.com/d/msg/openpnp/wbr4wcgolxg/j_Rz88YICAAJ, you can download a 3D pdf of a feeder that I've been working on that handles the tape peeling in a similar way.

[Michael Anton]

I got the stainless sprocket from http://lasergist.com/ in Greece.  I think the pair of them were $10 USD including shipping, which is not too bad for prototypes.  I've attached a picture of my sprocket assembly and the parts used to create it (except for the gear and screws).  I think there is actually a problem with the laser system they used to do the cutting, as there are all sorts of wobbles in the cut edge, mostly in one axis, so I suspect the machine had a backlash problem.

[mike]

Ime 1.6mm is typically the smallest router before any costs increase, but different fabs will have different standards. It may be feasible to use a combinatiom of drills and routing to get a smaller radius.

An issue with FR4 is as it's made with fibres laid up at right angles, a sprocket may have different friction and wear characteristics around the circumference.

If it's not smooth, there may be issues with scraping on plastic tapes that could bounce parts out.

An alternative may be to use a few layers of stainless stencil to make up a metal sprocket.

(Sent from phone)

-------- Original message --------

From: Michael Anton <3d.m...@gmail.com>

Date: 05/12/2017 06:00 (GMT+01:00)

To: OpenPnP <ope...@googlegroups.com>

Subject: [OpenPnP] Re: Feeder sprockets

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, 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/b46e9c91-3402-4516-a24d-14a9da55940e%40googlegroups.com.

[mike]

Problem with.lasercutting from final thickness sheet is you probably don't want a square tooth profile. A laminated 3d profile made from multiple sheets of stencil material might be worth looking at.

(Sent from phone)

-------- Original message --------

From: Michael Anton <3d.m...@gmail.com>

Date: 05/12/2017 07:59 (GMT+01:00)

To: OpenPnP <ope...@googlegroups.com>

Cheers,

Brad

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, 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/d08a8f28-b577-4841-8df0-516e5726528f%40googlegroups.com.

[Michael Anton]

If you size the square tooth properly, it should work fine.  For mine, I figured out the width of tooth that would fit in the tape holes if I made it out of 1mm plate.  That's how the ECM feeders are made.  Just flat plate, that looks to have been punched out of sheet.  The only appreciable radius is from one side, which is what leads me to believe that they were punched.  I had one of the sprockets put through the "shine up bath" by lasergist, which is probably a chemical etch, which did a nice job of taking off the burrs, and slightly rounding the corners, though I'm not sure I would do that in the future, as it was a little bit expensive.

I suppose the advantage of using stencil material, is that you might be able to get it cut quite cheaply if you worked out a deal with a PCB house.

[SMdude]

What about edge plating of the pcb gears? 2oz with  enig?

[Trampas Stern]

I have done various designs and tried various things on the feeders the conclusions I came to are:

1. I can not build anything cheaper or better than used Yamaha CL feeders (less than $40 each on ebay). 

2. If you want a lot of feeders, consider drag feeders using head  pin to drag, it will be cheaper but slower.  The problem here is the take up of the tape cover, I have not found a solution I like. 

I have tweaked by brain for the last year, even thought about the FR4 based designs, nothing worked as good or as cheap as the Yamaha feeders, these things are mechanical art work. 

[Robert Waters]

Are those feeders pneumatic? What is the specific part numbers for the $40 feeders? I see many of them in the >$100 range.

[Jason von Nieda]

This idea that the Yamaha feeders are $40 is a bit misleading. Yes, you can get them for $40 but you also need a fairly expensive block to mount them in, solenoid valves and control circuitry for each one, an air source, air filters and dryers, etc. It adds up to a lot more than $40 each. That's why a self contained feeder is so compelling.

Jason

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, 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/bb053b2a-4370-4855-a76f-69de5432fe75%40googlegroups.com.

[Peter Betz]

Great discussion!

This is very much on my mind. I have a sprocket design that I have 3D printed and seems to work just fine. I am just trying to nail down the motion of the feed. My feeders have IR light gates for the teeth but they do not block the light very well as they are 1mm thick. I have mounted encoders on the motors and I am just trying to play with them to get them their inertia to stop overrunning the encoder positions. I hope to play with this more tomorrow. I just had some sprockets laser cut out of stainless as well but they are a bit loose on the shaft.

I like the idea of PCB material, very clever, but I think the limitation will be the small features on the gear, and whether or not the fab house actually will be able, or want to, make them.

Peter.

[Bernd Walter]

On Tuesday, December 5, 2017 at 9:37:39 AM UTC+1, mikeselectricstuff wrote:

Problem with.lasercutting from final thickness sheet is you probably don't want a square tooth profile. A laminated 3d profile made from multiple sheets of stencil material might be worth looking at.

My siplace feeders are using a 2D cut sprocket wheel plus a sandwiched drum.

[Bernd Walter]

On Tuesday, December 5, 2017 at 5:25:29 PM UTC+1, Jason von Nieda wrote:

This idea that the Yamaha feeders are $40 is a bit misleading. Yes, you can get them for $40 but you also need a fairly expensive block to mount them in, solenoid valves and control circuitry for each one, an air source, air filters and dryers, etc. It adds up to a lot more than $40 each. That's why a self contained feeder is so compelling.

And the compressor can be too loud for some people.
I had to buy a silent running compressor for that reason, which came with an extra cost of course.

[Trampas Stern]

https://www.ebay.com/itm/Philips-Assembleon-FV-12-LOT-OF-4-PA2903-81-12mm-SMT-Feeders-Pneumatic/201964078531?hash=item2f05ff45c3:g:s-gAAOSwi8VZTr1y

You have to watch ebay but they do come up regularly 

[Trampas Stern]

Yes the block from robodig for 30 feeders is $128.  The tubing and relays add to this about $10-$20. per feeder.  So the base is around $25 per feeder.  

https://www.robotdigg.com/product/1190/Pick-and-place-machine-Feeder-mounting-block

However often people will have more feeders than slots on machine such that they keep parts on feeders full time and just swap them in as needed.   

If however you want the maximum number of feeders on the machine it is hard to beat the cost of a pin drag feeder, it is simple and it works. 

Personally I have found that I want automatic feeders for high run parts, 0.1uF, 10k resistors, etc.  Then for the parts that are used once per board, like a special RF choke, I would prefer to have it on a drag feeder plate.  Note I have started backing away from strip feeders as that I often have to swap out strips, and if you pull cover tape off the top, they parts go flying when you move them.   Hence for small run parts I want "on demand" that is the cover tape stays covering part until needed.  

What I would like is a pin drag feeder design where I could 3D print base easy to add more parts to machine.  For cover tape I am fine with weights method for the moment but would like a nicer solution.  Again this is not for high run parts, for them use the Yamaha feeders. 

[Bernd Walter]

I've just checked the Yamaha.
There are some slight differences.
The wheel is just about 30mm diameter.
The pick position is in front of the highest wheel part.
The tape goes straight to the wheel, but bents over it after the pick position.
On the picture I've opened the shutter on one of the feeders.

[Jason von Nieda]

Here's what I found:

Yamaha CL8 feeder: I can't get this apart. I have no idea how to do it without destroying it. So, I measured the sprocket as best as I could:

Tooth tip to tooth tip: 33.3mm

Tooth count: 24

Tooth length: ~1.4mm

Sprocket diameter, not counting teeth: 30.2mm

Neoden 4 Feeder: See attached image. Also see my previous teardown at https://imgur.com/a/osEGC

Hope this helps!

Jason

On Mon, Dec 4, 2017 at 9:31 PM Jason von Nieda <ja...@vonnieda.org> wrote:

I have some Yamaha CL feeders and an Neoden 4 feeder that I'll be happy to measure. It doesn't look like the Yamaha's sprocket comes out easily, but I should be able to measure it's diameter at least. The Neoden 4 one comes apart easily, so I'll get good dimensions on that one. Can't do it tonight, but will keep it on this list for this weekend.

Jason

On Mon, Dec 4, 2017 at 9:27 PM Bradley Parcels <bpar...@viopix.com> wrote:

Hey Bernd,

   Thanks for the input... Good call on the aluminum core - roughly doubles the price, but in this case, that's still under a buck a piece delivered... I guess it will all boil down to how much they care about the complex milling (of either material).

   I had seen somewhere that the minimum bend radius for most tape was in the ballpark of 25mm (thus 50+mm diameter) - but what I had in mind wouldn't actually bend the tape around the sprocket - just engage tangentially to the circumference... Anyone know any reason such a setup would still require the larger diameter? I realize that it would require a different tooth profile than the larger, but save that, I couldn't really think why it would matter? I'm guessing that on the siplace feeders, the tape wraps that sprocket to some extent?

    In any case - if anybody happens to have easy access to a proper sprocket (i.e. have a feeder disassembled already - please don't sweat it if you'd have to tear things apart!), and could be convinced to throw it on a flatbed scanner or something - I'd likely go with whatever diameter we can get a "known good" profile from... I was thinking of the smaller diameter purely for torque and cost reasons... but I'd almost certainly go with a known good profile (at whatever diameter) over having to potentially iterate a smaller custom design!

Cheers,

Brad

   

   

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, 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/1adc496f-f3f4-4c7f-855c-d0567f605de0%40googlegroups.com.

[Brad Parcels]

Hi all,

  Very sorry for the delayed response (“real” work – bah!)… But thanks so much to you all for all of the input!!

 

Jason – that’s perfect… thanks!

 

Bernd – thanks again – those pictures are also very helpful!

 

Trampas – fair enough… You’re probably right (about not being able to do it cheaper)… but I always enjoy a challenge 😊! Also, as Bernd brought up - noise is very much an issue in my case… I’m usually only playing with this stuff at crazy hours (i.e. 3AM now) and my setup is in our basement (which has regrettable sound dampening qualities). Despite having also sprung for a “silent” compressor (while significantly quieter than most - silent is quite a stretch) – I’m fairly sure that one of my favorite body parts would be in serious danger (from the lady of the house) if were to try running it at this time of night!

 

Your other comment about drag feeders does bring up somewhat of a tangent that I also meant to touch on… I’ve gone back and forth between the “fully compartmentalized” style of feeder, and simply adding another (quite wide, but short – maybe 50mm travel) axis, that has a drag pin actuator (i.e. solenoid) for each strip… Does the same job as the drag feeder, but can operate independently while the head is placing… Seems like it would probably be very (if not the most) economical – but I don’t love it, and I’m not sure why… I’m sure the idea must have been raised before – anybody have any thoughts as to why/why not go that way?

 

SMDude – edge plating may certainly help – good call

 

Mikeselectricstuff – may be right about the fibers… but edge plating (as mentioned above) may help there… never know till we try I guess!

 

Also with regards to the jumping components – not sure if anybody has done anything more extensive on it – but I was quite surprised to learn that many (most?) of the really tiny stuff (i.e. 0603/0402 caps/res/LEDs) are actually attracted to magnets (so ferrous?)… In some of my earlier prototyping, I threw a magnet under the tape between the peel/pick locations, and it seemed to keep the components in place pretty well… I only tested with a few different types of components – so I’d be curious to know if anybody has come across any that don’t work? (Note – none of those that I tested are strongly magnetic – but at least with an N52 neodymium under the tape, they wouldn’t fall out - even when I shook them like a madman).

 

Michael Anton – wow – that’s hugely cheaper than any laser cutting I’ve come across… that link could be handy regardless – thanks!

 

In any case, I’ll keep you posted on any progress on my end!

 

Cheers,

Brad

 

 

 

 

 

 

 

From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Jason von Nieda
Sent: December 9, 2017 3:58 PM
To: ope...@googlegroups.com
Subject: Re: [OpenPnP] Re: Feeder sprockets

 

Here's what I found:

--
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/B7-6KBnYbVI/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/CA%2BQw0jw7kzDTnZgkQLCR7vURg_OhEZUv0DCnSSQLBcwGZCR1jQ%40mail.gmail.com.

[Michael G.]

Hi,

 

Also with regards to the jumping components – not sure if anybody has done anything more extensive on it – but I was quite surprised to learn that many (most?) of the really tiny stuff (i.e. 0603/0402 caps/res/LEDs) are actually attracted to magnets (so ferrous?)… In some of my earlier prototyping, I threw a magnet under the tape between the peel/pick locations, and it seemed to keep the components in place pretty well… I only tested with a few different types of components – so I’d be curious to know if anybody has come across any that don’t work? (Note – none of those that I tested are strongly magnetic – but at least with an N52 neodymium under the tape, they wouldn’t fall out - even when I shook them like a madman).

is it common to put a magnet below the tapes in commercial feeder? if so, I would like to add a feature to insert a small magnet into my feeder to prevent parts from jumping (especially a problem with plastic tapes).

[Brad Parcels]

Hi Michael,

   Not to my knowledge (but to be clear – that’s very limited… I’ve never seen a commercial feeder in person). I just happened to run across an article on Hackaday where they were actually picking and placing using an electromagnet instead of a vacuum… At first, I didn’t really believe many parts would actually be that ferrous… But after testing a few parts myself, I thought that it may just work as a nice alternative to a cover-plate that has to shuttle back and forth. Again – haven’t tested extensively – but would love to hear if it helps!

 

Cheers,

Brad

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/12284bdd-4733-4871-8c4b-dcd2d4d30df8%40googlegroups.com.

[Mike Harrison]

On Wed, 13 Dec 2017 05:05:27 -0800 (PST), you wrote:

>Hi,
>
>
>> Also with regards to the jumping components – not sure if anybody has done
>> anything more extensive on it – but I was quite surprised to learn that
>> many (most?) of the really tiny stuff (i.e. 0603/0402 caps/res/LEDs) are
>> actually attracted to magnets (so ferrous?)… In some of my earlier
>> prototyping, I threw a magnet under the tape between the peel/pick
>> locations, and it seemed to keep the components in place pretty well… I
>> only tested with a few different types of components – so I’d be curious to
>> know if anybody has come across any that don’t work? (Note – none of those
>> that I tested are strongly magnetic – but at least with an N52 neodymium
>> under the tape, they wouldn’t fall out - even when I shook them like a
>> madman).
>
>
>is it common to put a magnet below the tapes in commercial feeder? if so, I
>would like to add a feature to insert a small magnet into my feeder to
>prevent parts from jumping (especially a problem with plastic tapes).
>

I think it's due to nickel in the plating.
I just dipped a magnet in my big box of mis-picks and pretty much everything was attracted to some
extent.
I don't think it's a good idea as some parts are a lot more magnetic than others , so a magnet that
has sufficent pull to hold a less magnetic part will hold a more magnetic one so hard it won't be
pickable. Obviously inductors woudl be a non-starter ( and also risks magnetising the core, which
would affect saturation current0

Might be a handy hack for the odd "difficult" part but I'd be highly surprised if any commercial
feeders do it.

[Brad Parcels]

Hi Mike,
Good call - the nickel plating makes sense - fairly ubiquitous... I was also concerned that it would hold "too-hard" - but at least for those that I've tested, it wasn't a problem (it is held-off by a certain distance, not directly against the magnet)... I'd also agree that it probably isn't wise to use a magnet for all parts... But for the really small res/caps/LEDs, it seems like it might be a plausible solution for "jumpier" feeders... Again - I'd be curious if anybody can find any that don't work (If it's only a solution for certain size parts, it better at least be generalizable to all of those size)!

Cheers,
Brad

-----Original Message-----
From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Mike Harrison
Sent: December 13, 2017 6:14 AM
To: ope...@googlegroups.com
Subject: Re: [OpenPnP] Re: Feeder sprockets

--
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/B7-6KBnYbVI/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/al923dtj12trba5fk1sf6vq7vgd1ip2brb%404ax.com.

[Anthony Webb]

I would not do it, we had magnets holding down our strip feeder holder, and it was causing parts to actually flip over in the tape! 

Sent from my iPhone

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/12284bdd-4733-4871-8c4b-dcd2d4d30df8%40googlegroups.com.

[Michael G.]

okay, one more question on this: what are the shutters all about? there are shutterless variants of commercial feeders and some seem to be coming with shutter. the shutter is to prevent parts from popping out, right? in which designs shutter are used and in which not? are there any criteria? or is a shutter useful in general for every design? I have never seen a commercial feeder unfortunately.

[Brad Parcels]

Hi Anthony,

    I noticed the same – but only if the part crosses a magnetic boundary (i.e. if you tried to use multiple magnets in a line, instead of a single larger magnet)… But it seems to be okay if the part remains in a single magnetic domain between peeling and pickup.

 

    Just to be clear – I’m not entirely advocating for a magnetic solution here either… Just trying to pass on some of the things that I’ve stumbled across!

 

Cheers,

Brad

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/35BB8A45-105F-40F6-AB97-75FACD37E98F%40gmail.com.

[Bernd Walter]

On Wednesday, December 13, 2017 at 11:07:36 AM UTC+1, Bradley Parcels wrote:

Hi all,

  Very sorry for the delayed response (“real” work – bah!)… But thanks so much to you all for all of the input!!

 

Jason – that’s perfect… thanks!

 

Bernd – thanks again – those pictures are also very helpful!

 

Trampas – fair enough… You’re probably right (about not being able to do it cheaper)… but I always enjoy a challenge 😊! Also, as Bernd brought up - noise is very much an issue in my case… I’m usually only playing with this stuff at crazy hours (i.e. 3AM now) and my setup is in our basement (which has regrettable sound dampening qualities). Despite having also sprung for a “silent” compressor (while significantly quieter than most - silent is quite a stretch) – I’m fairly sure that one of my favorite body parts would be in serious danger (from the lady of the house) if were to try running it at this time of night!

Those silent compressors all seem to be from the same Chinese company, sold under different labels.
When I'd bought mine I've searched a bit and quality as well as noise differs a lot between brands.
My impression was that the air intake filter is a big part of the noise difference and maybe you can tune that on your.
I think on my 18l one I can place about 600 parts until it restarts.

 

Your other comment about drag feeders does bring up somewhat of a tangent that I also meant to touch on… I’ve gone back and forth between the “fully compartmentalized” style of feeder, and simply adding another (quite wide, but short – maybe 50mm travel) axis, that has a drag pin actuator (i.e. solenoid) for each strip… Does the same job as the drag feeder, but can operate independently while the head is placing… Seems like it would probably be very (if not the most) economical – but I don’t love it, and I’m not sure why… I’m sure the idea must have been raised before – anybody have any thoughts as to why/why not go that way?

 

SMDude – edge plating may certainly help – good call

 

Mikeselectricstuff – may be right about the fibers… but edge plating (as mentioned above) may help there… never know till we try I guess!

Don't think edge plating will work, because you simply can't plate the whole outline.
The plating has to be done with the PCB in the panel.

[Bernd Walter]

On Wednesday, December 13, 2017 at 4:26:20 PM UTC+1, Michael G. wrote:

okay, one more question on this: what are the shutters all about? there are shutterless variants of commercial feeders and some seem to be coming with shutter. the shutter is to prevent parts from popping out, right? in which designs shutter are used and in which not? are there any criteria? or is a shutter useful in general for every design? I have never seen a commercial feeder unfortunately.

There is a very short answer: if the movement uses fast acceleration, you need a shutter.
The Yamaha CL series uses shutters, because the pneumatic actor bumps against movement limits.
The more modern Siplace S feeders use shutters, because they bump the transport motor against a hard limit.
The older Siplace feeders are shutterless, because the use a smooth drag mechanism and are rather slow.
Some Siplace replacement feeders are shutterless, because they use steppers with smooth acceleration/deceleration.
And those feeders are super heavy, so they don't shake by machine movements.

There are not many videos of PnP machines showing the feeders and even less with shutterless feeders.
This is one:
https://youtu.be/8geFoSQASiA
The feeders are electric, but in this video actuated from the head, which is because it can also actuate much older non electric feeders.
At the end you will see that they didn't work very well for the LED tape.
I'm partly running the same feeders and was surprised to see that a part is always ready to pick and the tape is advanced after picking.

Those are such feeders running on my OpenPnP machine:
https://youtu.be/kiv9UX46JWY

Some details on how they work.
In some I had been using a reduced voltage to get a better view on the motion.
Unfortunately they don't like to be run with the side plate removed, so I had to press on it to keep it moving.
https://youtu.be/BKNlFyXO1fY
https://youtu.be/eUQJxQEHgqo
https://youtu.be/Y1WWvaRjqRM

[Michael G.]

went over to a new topic not to clutter this thread any more... https://groups.google.com/forum/#!topic/openpnp/NxWOBHsgeyg

[Michael G.]

hi

just stumbled over this on-topic site: http://tim.cexx.org/?p=798
not checked in detail, but maybe it's valuable for sprocket design...

[Joel Mckay]

Over at HaD someone requested our clubs version of the spoke-wheel sprocket from an abandoned project.
Thus, I pushed what I had left over, and some tips here:
https://github.com/Joel-Mckay/Pick-And-Place_OSH_feeder

The minimum tooth count we used ensures there will always be tooth engagement, and thus does not easily jam on thicker tape (see 0802 in photo).
Trust that getting it any smaller is probably going to be rather difficult, and there will be some tear-out unless supported on both sides.

My only build complaint was the severe undercutting from the etching process knocks down the tooth profile a bit.
I really wanted to try the toner transfer process for hobby accessibility, but the project was already over budget.

Best of luck, =)
~J~


On Monday, December 4, 2017 at 5:09:15 PM UTC-8, Bradley Parcels wrote:

Howdy guys and gals,

  Correct me if I'm wrong, but by way of a quick search or two, it looks as though the feeder sprocket "problem" is still sort of open?

 

  I was just wondering if anybody has tried using our old friend FR4 as the sprocket material? I was thinking that if you round out the base of the teeth so that it can use standard routing (i.e. at a cheap PCB Fab), you could just slap that between two discs that provide the base (either printed, or FR4 as well), and screw/glue it all together:

 

 

 

For a 30mm diameter x 1mm, it would only be about $50 for 100(!) of them delivered from PCBWay – and half of that is the DHL charge. (Not to suggest that’s the place to use – I just picked a fab at random)…  This is assuming they don’t surcharge for the routing, but I have seen at least some intricate routing in their sample pictures, so I’m optimistic.

 

Given that it’s a PCB, you could also get a bunch of customization and complexity for free – i.e. a keyway, holes to screw together, holes for an opto-sensor, tracks for some kind of drag contact or capacitive sensing, routing a side disc as a 2GT belt pulley, etc. etc.

 

At that price, I’m tempted to give it a shot the next time I put a PCB order in – unless somebody has already gone down this path and can advise against it? Or has a good/cheap/quick source been found for the real deal that makes this all for naught? Barring that - with all the expertise here, is there any suggestions on the profile, diameter, width, and the like? Has there been any sort of standardization amongst what people have been using?

 

Cheers,

Brad

[Thomas Langås]

We (I'm working on the project that was featured on hackaday now) are using a laser cut sprocket that is 1mm thick, and it's cut out of stainless steel.  It fits very nice on the various tapes we tested.

We have ordered two different versions to see if we get better accuracy. The one in the article has 1mm holes, the two other versions we have ordered will have 0.5mm and 0.2mm respectively.

Also, we are looking at 0201, but currently testing with 0402.

-Thomas

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@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/076442ff-fb75-4506-a884-25404806a4c6%40googlegroups.com.

[Tracker J]

Hi Thomas,

Very intersting approach, somehow similar with my owm experiments.
I am using some 3D printed sprokets so it's quite hard to play with under 0805.
From where did you ordered the sprokets? How much? If you have some spare ones I will be more than happy to buy few.

Cheers,
TJ.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

[Thomas Langås]

We are only ordering a few at the time now, during prototyping. We got a price of about 1.5usd pr piece in small volume (200).  I plan on setting up a pre-order and announce when we do the first small volume order, to allow others sourcing the sprocket. We are looking into adding pre-order for a kit version of the full feedet as well. It seems like there's a few people interested.

-Thomas

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@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/d280a949-c693-4197-b43e-d05f3f735cdb%40googlegroups.com.

[Tracker J]

Sounds great, looking forward to give it a try :)

Do you have any price estimation for such a kit?

Also, from where will be done the shipping? EU? USA?

This is how is looking my 3D printed sproket after passing thru it 5 full reels (15000 pcs of different types of 0805 caps and resistors):




​

All the best,

TJ.

--
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/B7-6KBnYbVI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+unsubscribe@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/CANN7TQnbvVXooSs-mDmWZQLZScuB8aTp2U20PMrxHCUQueFUDQ%40mail.gmail.com.

[eSlavko]

TJ do you have photo of 3D printed sprocket as new, and worn one to see difference? and what material is used?

[Thomas Langås]

On Feb 11, 2018 13:49, "Tracker J" <track...@gmail.com> wrote:

Sounds great, looking forward to give it a try :)

Do you have any price estimation for such a kit?

Our intention is to sell the first ones at production cost, not to earn money.  So 20-30 usd, maybe.  Our goal is sub 20 usd pr feeder.

Also, from where will be done the shipping? EU? USA?

Norway, Europe. Unless we figure out something else.

This is how is looking my 3D printed sproket after passing thru it 5 full reels (15000 pcs of different types of 0805 caps and resistors):

Yeah, our sprocket will be more durable for sure.  :-)

[Tracker J]

I don't have around any before and after pictures, but the above ones are after 15000pcs usage and I can say that the difference is zero to none.

I don't know exactly the 3D printing parameters and resin type, it was not done by me, I will ask and will come back later with details.

[Tracker J]

On Sunday, February 11, 2018 at 4:16:18 PM UTC+2, Thomas Langås wrote:

On Feb 11, 2018 13:49, "Tracker J" <track...@gmail.com> wrote:

Sounds great, looking forward to give it a try :)

Do you have any price estimation for such a kit?

Our intention is to sell the first ones at production cost, not to earn money.  So 20-30 usd, maybe.  Our goal is sub 20 usd pr feeder.

Sounds great. I would definitely give it a try when available

Also, from where will be done the shipping? EU? USA?

Norway, Europe. Unless we figure out something else.

 

Sounds encouraging for me in terms of shipping costs . Originated from a  EU Member state to avoid the EU Customs clearance hassle would be even better :)

This is how is looking my 3D printed sproket after passing thru it 5 full reels (15000 pcs of different types of 0805 caps and resistors):

Yeah, our sprocket will be more durable for sure.  :-)

As been made of steel, probably yes :).  Meantime I must say that I'm also impressed by the durability of these resin 3D Printed ones.

TJ.