Open source feeders

[cf]

I found this kickstarter project for open source feeders,

http://www.kickstarter.com/projects/1973759599/open-source-smd-parts-feeders

There also was a project page

http://www.richardspelling.com/?p=683

But I haven't been able to find the actual design unfortunately.

[Karl Lew]

I see 3d printed parts. 8)

[Jason von Nieda]

Richard posted some of the designs to http://www.thingiverse.com/thing:25995

I don't know if this is complete or not. He does mention the Kickstarter on there. 

On Wed, Jan 30, 2013 at 10:52 PM, Karl Lew <ka...@firepick.org> wrote:

I see 3d printed parts. 8)

--
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/msg/openpnp/-/-Fppe-NFHKUJ.

For more options, visit https://groups.google.com/groups/opt_out.

[varun s]

guys found this  http://www.briandorey.com/post/Panasonic-Pick-and-Place-Feeder-Hack.aspx

[Karl Lew]

Varun that is very interesting on a couple counts:
1) I cant find that part on ebay today "10488BD059" but another ratchet feeder was available for $699, so sourcing is a concern
2) they were going to make their own from laser cut but it was cheaper to hack

Personally, I would pay more for a repairable feeder with easily sourced or manufactured components. In addition, an OSHW design can be improved by the community to suit individual needs with public collaboration and review.

[Karl Lew]

Ah, thanks Jason, I missed the Thingiverse stuff.

Here is what I saw:
1. tiny motor with worm gear turns sprocket wheel
2. sprocket wheel advances tape
3. unclear how tape cover is removed.

[cf]

It seems like the back side of the thing has a pickup spool for the cover tape, dont know how it works

http://thingiverse-production.s3.amazonaws.com/renders/98/e4/be/81/0b/main_-_right_preview_featured.jpg 

 

If you look cloesly you can see the cover tape is twisted and it goes back inside the feeder

http://thingiverse-production.s3.amazonaws.com/renders/6d/70/a6/de/86/IMAG0287_preview_featured.jpg

[Karl Lew]

cf, that makes more sense now.

Here's my current hypothesis: 

1. tiny motor with worm gear turns sprocket wheel

2. sprocket wheel pulls back cover tape
3. cover tape pullback advances component tape

That makes more sense because the thingiverse doohickey has a thin slot that probably takes the cover tape (I originally was confused about how the component tape went through that slot and why it needed to bend around the sprocket) 

[Jason von Nieda]

Richard wrote quite a bit about his feeders and even posted some videos describing them and how they worked. I don't have the exact threads handy, but you can find them by his name.

https://groups.google.com/forum/#!searchin/openpnp/richard$20spelling

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/msg/openpnp/-/TJtIBPywuHgJ.

[Ami]

Hi Karl, hi everybody,

Richard was here before you came with Sylvester,
He explained how it works before, check the older postings.
I remember he explained there's a dc motor with gearhead and wormdrive and sprocket, etc etc.

I think he should read our messages from time to time, if you cry loud enough he should hear you.

[Karl Lew]

Ami/Jason, thanks for the deja vu.

I was about to say he was using a dc motor, etc. with some sort of feedback, but there you go.

What stunned me is Richard's claim of $30 per feeder, because his feeders seem quite capable and I would have no idea how to get that price. The Shapeways cost of his feeder would easily be $10-$20. The Ponoko cost would be even higher. And gear motors on Sparkfun are about $15. And none of that includes the PCB, the worm gear or the drive sprocket. I'd guess that any attempt to source this stuff entirely vanilla-style without eBay scrounging would be about $60-$70 for just the head. That would not include reel holders or any other stuff.

[Karl Lew]

I added Zippy to the OpenPnP Hall of Machines since we keep rediscovering Richard's innovations.

Guys, don't be shy. Please add links to your own machines. It really helps us browse all the wild creativity out there.

https://github.com/openpnp/openpnp/wiki/Machines

(cough) you, too, Ami. I would love to see more about your machine in one place. 8)

[Ami]

;-) Hi Karl,

Man, I don't like drawing, and I hate show-and-tell. Kindergarden was a nightmare I want to forget ;-)
I'll see if Sylvestre is not shy.

I've been sharing with you guys detail by detail on different aspects of my experience building it,
I'd like to think that it's useful to all of you guys, it's just that I haven't had the time to put it together in a page, writing it down, etc.
Perhaps one day, OK, no promise, and anyway it's not quite done, I have other things in mind to improve it.

For now I'm going fishing mackarel with Sylvestre.
Ciao!

[cf]

 

I found another design made out of lego and lift card spools :)

 

http://www.youtube.com/watch?v=GO7BruVYRxA&feature=youtu.be

 

I found it another forum that is discussing SMT tech a lot and DIY feeders

 

http://forums.adafruit.com/viewforum.php?f=42&start=0 

[Karl Lew]

Yay! Lego!

The FirePick mascot is the Lego Crash Dummy minifig. 8)

Note the use of shape memory to activate the feeder. I have long wanted to use nitinol, but it's never made itself into any of my projects. It's slow and requires long spans to achieve useful contraction. See the long lengths of wire behind each feeder and the time (seconds) it takes for the wire to recover for the next feed. Great video, but I'm sticking to servos and steppers!

Thanks for the link, cf. Looks like Lady Ada has her own PnP village. 8)

[Ami]

How does it work?
Is it a motor puller that advances the tape?

[Karl Lew]

The wire contracts when heated by the current passing through it. The contraction advances the ratchet and tensions the pullback spring. When the current is turned off, the spring pulls the tape forward slowly as the nitinol relaxes in about 1-2 seconds. This mechanism is very precise since nitinol contracts about 3-5% when heated. That is why you see about 8-10cm of wire behind the feeder. The current required to heat the wire varies with wire diameter, but the pull is quite strong, so the controller is simple and just sends 100-300mA to heat the wire. Elegant.

[Karl Lew]

No. I am wrong. The black cord pulls back the cover tape and is part of a security badge pullback thingy. You might use fishing weights. 8)

The nitinol is probably in a linear actuator like this
http://www.imagesco.com/nitinol/linear_muscle_wire_actuator.html

[Karl Lew]

note that the motor stroke length is, haha, 4mm. Exactly what you need for 8mm component tape advance.

[Ami]

You're right, it has nothing todo with lego.
Marketing gimmick!

But the "muscle" motor is interesting. How many times can it do like that?
Will it get uricemia after long run?

[Ami]

You know what?
I think the badge yoyo is the most interesting part of the video.
With drag feeder and cut tape, it's the best solution for low volume feeder.

why not?

[Michael Grant]

Wow, I just watched the video again and realized the gantry is nearly identical to my design...

--

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/msg/openpnp/-/3iRfDi59wHkJ.

[Michael Grant]

Here are my collected thoughts on feeder via drawing.

The gears, worm and motor are off the shelf eBay specials:
http://www.ebay.com/itm/34-1-PRECISION-MOULDED-PUSH-FIT-WORM-GEAR-SET-10-pieces-/160948863213?pt=LH_DefaultDomain_0&hash=item25794cc0ed
http://www.ebay.com/itm/COMPACT-3-6-VOLT-DC-HIGH-TORQUE-MOTOR-10-pieces-/160917013856?pt=LH_DefaultDomain_0&hash=item257766c560

The small motors should have plenty torque with a 34:1 reduction.
As the motors are 24mm diameter,  staggering the pullers would give a tighter fit per tape...

Michael

[Michael Grant]

And here it is with the motors staggered allowing for tight spacing between feeders.
Feeder 1, 3 & 5 are driven off the bottom, 2, 4 & 6 from the top...

The gears are a bit narrow but still wider than the tape cover.
A guide would keep the tape cover aligned with gears.

Michael

[Michael Grant]

Hmm, a test with plastic gears on the cover tape reveals it requires a good bit of pressure to bite.
Not an insurmountable amount but that cover tape is really slippery...
A metal gear may have more bite, one metal will probably suffice.

At least with the slot sensor it won't matter if it slips.

Michael

[Michael Grant]

Easier than metal, the plastic gears are nylon 66.
A bit of heat (~268 C) and a sprinkle of fine sand might turn the gear surface into a sandpaper traction wheel...

[Michael Grant]

I looked at the last couple of PCB designs I did and counted about 25 to 35 individual smd place-able parts.
With a bit of effort these designs could be trimmed to around 20 to 30 parts maybe.
About or over 90% of these parts are on 8mm tape as I tend to gravitate towards discreet parts for cost.
The other ~10% are MCUs and ICs, very few large size caps as ceramics have improved so much in capacity and price.

My machine width (minus head offsets) gives me enough space on the front to run about 48 feeders (8mm tape + ~5mm between).
After that I'd have to fit stuff in the back and/or use trays.

I had a quick peek at the options for driving 48 motors and sensing 48 slot interrupter inputs.
The MCP23017 ($1.30) is a 16 channel i/o expander for i2c, it would work for sensing as well as switching the motors via L293DD.
Four motors can be driven by a single L293DD ($3.50) or 3 with the weaker T6817 ($1.95) which happens to include the serial expander and can be daisy chained.

The half bridge is ideal (as opposed to low side fet) for switching the motors as when off it would short the motor for a more controlled stop.
At some point I'll make another isolated USB Arduino layout with i/o for 48 feeders and 5v smps.

Michael

[Paul Kelly]

I think the badge yoyo is the most interesting part of the video.

 

Seriously guys. If you think that a badge yoyo with 300mm of travel is acceptable for your application then:

A: You would be better off with  a static/strip feeder

B: You don’t need a pick and place machine, you need an assisted manual placer.

 

We regularly end up with tape and cover tape beards that reach the length of the workshop after ony an hour or two of operation. Having to re attach a clip on every feeder, every 300mm of tape would be a misery....

 

Paul Kelly | Design Engineer
CASWA Pty Ltd
9/4 Roper Street, O'Connor  WA  6163
M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com

 

                                        www.caswa.com

P Be earthsmart.  Please consider the environment before printing this e-mail

--

[Paul Kelly]

Our “gears” are more like square toothed timing pulleys.  On larger feeders (ie >= 24mm) the cover tape gets folded in half and runs through soft rubber pinch rollers...

PK

 

Paul Kelly | Design Engineer
CASWA Pty Ltd
9/4 Roper Street, O'Connor  WA  6163
M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com

 

                                        www.caswa.com

P Be earthsmart.  Please consider the environment before printing this e-mail

 

--

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/msg/openpnp/-/lLNSC4ui4DUJ.

[Ami]

:-D I love it when PK rises up like that to keep us motivated ;-)

[Paul Kelly]

J Just trying to help keep it real!

PK

To view this discussion on the web visit https://groups.google.com/d/msg/openpnp/-/wmir_RWuAzYJ.

[Miroslav Pejic]

Hi Paul,
How about a picture of this "badge yoyo"? Didn' find this.
Thanks,
Miroslav

[Ami]

Hi Miroslav,

It's just a badge holder with retractable cord that is used to pull the smt-cover tape. I call it badge-yoyo.
You can see it in the video link sent by cf. It's the black round thing that pulls the smt cover tape.


Basically OpenPnP can do the tape advancing using drag-feeder. (the head lower a pin to drag the tape)
The remaining thing to do is to pull out the cover tape.
Normally this is done using fishing weight (around 50 gr) to pull the cover tape until the weight reach the floor.
The retractable badge holder can replace this fishing weight, which is alright if we don't have a lot of boards to run, or for some of the less used parts (1 item per board).
This thingy can pull like 60cm of tapes. For 8mm tape that gives 150 parts.
The only drawback is that your machine will look like chinese harp,

But Papa PK is right, the badge yoyo is not improving a lot from our current situation: ie. after a while we need to re-attach the clip like we raise the weight.
The challenge for us pnp-maker is to make a device to pull out the cover tape automatically, using either motorized system or mechanical system so we don't have to tend for the machine once it's launched.
We shall not give up from this task at hand! So forget the yoyo!
(Just between us, if you want to place order, it cost around 30 cents per 100 pcs, please don't tell PK ;-)

[Karl Lew]

Michael, the more I look at your design the more I like it. The alternating up/down worm gear drives are ingenious!
And I really like that you have put yourself into the business of making us driver boards. I want one!

Why are the gears so large? They seem to be about 35mm OD. Would 18 mm work just as well?

And do you need any 2d or 3d parts that I could design for your cover tape cookie monster?

[Michael Grant]

Regarding tape beards, I calculated about 20 meters for a 5k roll of parts at 4mm spacing.

Using a spool instead of pinch, if the cover (0.08mm thick) were to spool up on a receiver with a core diameter of 1cm, it would finish under 5cm diameter.
But besides the effective gearing change as it rolls larger, the take up spools would require emptying...

I should test the sand coated gear today, as well as gear to rubber roller.
Plastic on plastic requires a bit too much pinch force to gain the friction for my liking

Karl, the gear size was just eBay stock, smaller should work as well, but larger will have more area to increase traction..
Don't purchase the gears from the links I posted earlier, the worm is not m1 compatible as I've drawn it, the seller said he will be putting some m1 worms up today.

I will design the frame for cnc milling around the 9mm acyrlic I have in stock, but it should be equally translatable to 3d printing.
I think the design can have two windows for motors, where only one is populated to stagger.

For the controller, the T6817 is the cheaper part but is only stocked by digikey.
I think I will use the l293dd and the mcp23017, it will be a little more expensive but the design and programming will be simpler.

[Michael Grant]

I suppose the controller should also have a push button input for each feeder.
The button would be on the feeder and would advance the tape manually for loading..

Perhaps this button can be multiplexed with the slot interrupt sensor, maybe the same.
If the button and the sensor were wired parallel, when pressed or dark sensed, the motor rolls until a low to high transition occurs.

Opto sensing is easy but it just occurred to me that capacitive sensing could also work.
And would be cheaper when using a chip like the AT42QT10xx.

[Jason von Nieda]

Michael,

You are using a slot sensor to detect the holes in the tape, right? Keep in mind that some tapes are transparent. Richard Spelling ran into this with his feeders.

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/msg/openpnp/-/WFV8pH3JIpwJ.

[Paul Kelly]

Actually, I just remembered another metric we’ve  calculated. It’s very relevant to the openPnP space (for some of you).

The stuff we make has between 30-150 components per board, with most board having <<100 components.

It’s all pretty normal gear: a micro, a power supply or two, a couple of special IC’s like RTC’s or amplifiers. Nothing exotic.

 

You’ve seen the video, our old girl runs at about 1000CPH.

 

We reckon it costs between $1000 and $2000 per hour in parts to run the machine. Ie that’s the dollar value of parts it places on a continuous basis. Now your boards may be different, but I’d bet they would still be in the high hundred$ per hour and remember, we have wholesale accounts with the big suppliers, so we get better pricing... Now, the machine produces between $5000 and $15000 of boards per hour, so I’m not (or at least my accountant isn’t) complaining..

 

In the context of that, a $2000-$3000 pick and place machine seems unnecessarily thrifty to me.

Not trying to demotivate, just a: explain where I’m coming from a little better, and b: point out that if an OpenPnP kit cost $5000 or $10000 (and came with feeders) then it would still be a bargain...

 

PK

 

Paul Kelly | Design Engineer
CASWA Pty Ltd
9/4 Roper Street, O'Connor  WA  6163
M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com

 

                                        www.caswa.com

P Be earthsmart.  Please consider the environment before printing this e-mail

 

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

Sent: Wednesday, 13 February 2013 5:10 PM
To: ope...@googlegroups.com

To view this discussion on the web visit https://groups.google.com/d/msg/openpnp/-/wmir_RWuAzYJ.

[Paul Kelly]

Feeders:

Plastic base, tape slides in channel.

Rubber pinch rollers. The lower one presses down on tape+cover strip.

Cover strip wraps around lower roller and passes between the lower and upper roller.

 

Drive is by a thin gear on the outside edge of the feeder, driven by a stepper underneath the tape.

Positioning is by optical sensing of the holes.

 

Having the pinch rollers directly drive the tape as well as pull the cover off helps deliver more force to the tape advance.

Stepper driver (cheap) does primary positioning.

Optical hole sensor does secondary positioning.

Vision system picks up the slack.

???

PK

 

Paul Kelly | Design Engineer
CASWA Pty Ltd
9/4 Roper Street, O'Connor  WA  6163
M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com

 

                                        www.caswa.com

P Be earthsmart.  Please consider the environment before printing this e-mail

 

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

Sent: Wednesday, 13 February 2013 5:10 PM
To: ope...@googlegroups.com

To view this discussion on the web visit https://groups.google.com/d/msg/openpnp/-/wmir_RWuAzYJ.

[Michael Grant]

Jason, how did I overlook that?
I have some holding tantalums directly to my right...

So capacitive might be the best solution for these.
Either that or a slot sensor that detects the part directly.

Michael

[Paul Kelly]

Someone asked for this pic..

 

Paul Kelly | Design Engineer
CASWA Pty Ltd
9/4 Roper Street, O'Connor  WA  6163
M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com

 

                                        www.caswa.com

P Be earthsmart.  Please consider the environment before printing this e-mail

 

From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Michael Grant

Sent: Thursday, 14 February 2013 2:44 AM
To: ope...@googlegroups.com

--
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/msg/openpnp/-/xTFF0g1F8MQJ.

[Karl Lew]

Paul, thanks for the 1000 words. Your gear is clear.

[Jason von Nieda]

Hi Paul,

In that context, I agree. But that's not the context I'm working in. It's comparing apples to oranges. If you are making hundreds or thousands of boards a week you need professional gear that never goes down. Time down is money lost.

What I am targeting is the person who wants to make a handful of boards a week. Or maybe even less. Maybe they, like me, just don't like being hunched over a table with magnifiers and a pair of tweezers to assemble prototypes.

I don't want a pick and place to make money with. I want one to make my life easier. I have the need to assemble boards now and then, but it's not my bread and butter. This gives me a way to do that without completely breaking the bank.

When you frame the question as "What does a hobbyist need?" I think the answer is very different from "How do I run my business with this?"

Jason

[Paul Kelly]

What I am targeting is the person who wants to make a handful of boards a week. Or maybe even less. Maybe they, like me, just don't like being hunched over a table with magnifiers and a pair of tweezers to assemble prototypes.

Ahh, now this is where I think the misunderstanding may be. I’m not talking about magnifying glasses and tweezers, I’m talking about an assisted manual pick and place system.

When we started out, I was lucky to be able to assemble 10 units a day with tweezers and a magnifier.  When I bought/modified my assisted system I could do 10 boards in  an hour.

 

It’s hard to get your head around the difference that having a vacuum pencil that only moves in Z, A, X, and Y makes.  A moving wrist support and simple manual feeders finish it off and make it a vastly better proposition... For bonus points, add a video camera with a display.  I really believe that this is the appropriate technology for the user group you describe.

                                        www.caswa.com

[Karl Lew]

http://www.manncorpdirect.com/product-info.php?manual-pick-and-place-pid220.html

Above is such a manual SMT machine.

Although effective, manual SMT lacks a certain fun factor. I think many of us are entranced by the idea of entering an Eagle file, a BOM and pushing a GO button. Voila!  New prototype after plopping it on the electric skillet! 

[Jason von Nieda]

That could well be. I've never used one, but I just watched a few videos and it looks like a very helpful machine. That said, I'm still not interested in hunching over a desk and doing the work by hand.

I guess what it comes down to is that this something I want to do. Like many open source projects - this is an itch that I feel like devoting my time to scratch. In the end, it doesn't have to be useful to anyone but me.

Jason

[Paul Kelly]

 

http://www.manncorpdirect.com/product-info.php?manual-pick-and-place-pid220.html

 

Above is such a manual SMT machine.

And outrageously over priced it is too.

 

Although effective, manual SMT lacks a certain fun factor.

And I do acknowledge that.. I just think its important that folk understand that you don’t actually need a PnP robot to make small qtys of boards (ie <100) boards..

 

I think many of us are entranced by the idea of entering an Eagle file, a BOM and pushing a GO button. Voila!  New prototype after plopping it on the electric skillet! 

J You wait ‘till you do your first setup. We’re still tweaking our board files a year later...

 

PK

--
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/msg/openpnp/-/Ol_pr3mB8vYJ.

[Paul Kelly]

That said, I'm still not interested in hunching over a desk and doing the work by hand.

 

Yeah it really is quite hard work. But if we have to make just 10 boards, we use that set up... The carousel component tray is a really good way to handle loose components..

 

You can buy creams for your itches.... J

PK

 

 

                                        www.caswa.com

P Be earthsmart.  Please consider the environment before printing this e-mail

 

From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Jason von Nieda
Sent: Thursday, 14 February 2013 11:43 AM
To: ope...@googlegroups.com
Subject: Re: [OpenPnP] Re: Open source feeders

 

That could well be. I've never used one, but I just watched a few videos and it looks like a very helpful machine.

 

I guess what it comes down to is that this something I want to do. Like many open source projects - this is an itch that I feel like devoting my time to scratch. In the end, it doesn't have to be useful to anyone but me.

 

Jason

 

[Jason von Nieda]

I think many of us are entranced by the idea of entering an Eagle file, a BOM and pushing a GO button. Voila!  New prototype after plopping it on the electric skillet! 

J You wait ‘till you do your first setup. We’re still tweaking our board files a year later...

Paul,

This is something I hear over and over from pick and place operators but I have never really heard why. Can you expand on some of the challenges a bit? I'd love some details on what makes this a hard problem.

Thanks,

Jason

[Paul Kelly]

I think many of us are entranced by the idea of entering an Eagle file, a BOM and pushing a GO button. Voila!  New prototype after plopping it on the electric skillet! 

J You wait ‘till you do your first setup. We’re still tweaking our board files a year later...

 

Paul,

 

This is something I hear over and over from pick and place operators but I have never really heard why. Can you expand on some of the challenges a bit? I'd love some details on what makes this a hard problem.

 

 

I think it’s basically because placing a component on a board has a lot of variables. What angle you pick the component up. How much vacuum, how quick you move, how you center the part, how hard you press it down, what order you place the components around it in to minimise bounce and nozzle strike.  This is on top of all the stencil and paste variables.

 

A good system will allow you a reasonably large latitude for each of the variables, whist still outputting a board that reflows and passes test.

So, by the second or third panel, you normally have all the “I need to rotate this part 90 degrees” and, “Whoa, we need to go up 2mm here” problems sorted... 

What happens next is that you keep tweaking the file to reduce the CHANCE of a tombstone or dry joint.  As you make more and more boards, you start to see patterns in what causes them to fail test and you tweak the file to compensate. Sometimes it can take a hundred panels to get it good..... Then you get a batch of components with slightly different cases..... but that’s not what we are talking about here.

 

PK

 

Thanks,

Jason

 

 

 

 

PK

--
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/msg/openpnp/-/Ol_pr3mB8vYJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
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.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Karl Lew]

Paul, thanks for the enlightening explanation. I did not realize that component placement order mattered because of nozzle jackhammer earthquakes. Is that something that software could handle by simply placing in order of part size? I hope there is some correlation between part size and stability. Maybe smallest first?

[Jason von Nieda]

Thanks Paul, much appreciated. Gives me a lot to think about.

Jason

[Ami]

The part height also change (tantalum cap for example), because we don't necessarily use the same brand/reference number.
So you have to readjust the file practically for every run.

[Karl Lew]

Ami, then it is actually valuable to have a self-adjusting PnP head to handle varying part heights. Yay! Happiness. One less worry.

[Jason von Nieda]

Except then you have to deal with the earthquakes. So many tradeoffs.

On Thu, Feb 14, 2013 at 9:16 AM, Karl Lew <ka...@firepick.org> wrote:

Ami, then it is actually valuable to have a self-adjusting PnP head to handle varying part heights. Yay! Happiness. One less worry.

--
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/msg/openpnp/-/Z5Q0g6jUQPIJ.

[Ami]

Guys, make it work first!

Self-adjusting-head-that-doesn't-cause-earthquake we'll invent that later.

[Ami]

Ordering parts is like ordering a box of chocolate, you never know what you're gonna get.

Sometimes it's tantalum cap which gets taller, the other day was the chip inductor that gets smaller.

[Karl Lew]

Ami, I am not expecting any earthquakes because the spindle floats and is gently pushed down by a small spring. No jackhammer. Contact shock is minimized simply by going slowly near the board. The board should not be bent at all so there won't be any part release recoil earthquake.
http://www.firepick.org/design/z-axis

I am still waiting for Ponoko parts so that I can verify this.

[Ami]

That "depends" what you call earthquake.
(Man, because of you I can't keep a straight face when saying that)

That long gears have friction don't they?


I don't have problem with smaller parts like 0603 because they stick like ants.
But electrolytic cap is the worse: their pads are too small compared to their body.
SOD80C sometimes also gives problem if the paste is not enough, it simply rolls or stands up during reflow.
One of the critical phase is applying the stencil.

[Karl Lew]

Ami, oops I missed that. You are quite right.

Perhaps I may need to...umm...grease the long gear (cough)? That may assist in a (cough) smoother delicate part placement?

[Paul Kelly]

The part height also change (tantalum cap for example), because we don't necessarily use the same brand/reference number.
So you have to readjust the file practically for every run.

Only if you don’t use pneumatics to place.

PK

To view this discussion on the web visit https://groups.google.com/d/msg/openpnp/-/ahGOAa7UwaAJ.

[Paul Kelly]

Vibration is less of a problem if the board is well supported, one of the
things we've learned to do is to populate a panel starting with all the
boards at the top and bottom, then do the ones in the middle last. Yes, we
try to put heavy things like inductors on last, but we can't always do that
if, for example, there's a tall electrolytic right next to it...

PK

--
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/msg/openpnp/-/0Cgx6kb5jsEJ.

[Michael Grant]

A strange coincidence...

I have some gears of unknown origin.
They appear to be module 1.25 but I may have deduced this wrongly.

The coincidental part is that the teeth seem to line up perfectly with the 4mm tape spacing.
As the picture illustrates, turning down the edge should leave enough tooth to engage a worm.
Quick and easy work on a small lathe...

[Karl Lew]

Paul/Michael, both of your machines seem to have no cover tape guides on the pinch gears. How does the cover tape stay aligned during a long run and not slip out of the pinched gears?

[Jose Lopez]

Hi All;

Just want to share the following link and a few pictures...that may tweak the open feeder design's direction...

http://europlacer.com/en/ii-feed.html

[Karl Lew]

HLO, thanks for the link! It is very helpful.

I like how the cover tape is redirected down to gears BELOW the feed plane. Also interesting is the slight angle of the cover tape pull. That makes sense because it provides greater shear on one side of the cover tape, which might lessen the force needed to initiate each cover tape pull. Also, I do see flanged plastic pulleys and what looks like a stepper-driven pinch gear.

[Jose Lopez]

The design is also very fast to setup...
simple feeder head, now if we apply the best things of this design and add the motor driver so we can use each feeder on its own....not in magazine...

On 15/02/2013, at 10:01 AM, Karl Lew wrote:

> HLO, thanks for the link! It is very helpful.
>
> I like how the cover tape is redirected down to gears BELOW the feed plane. Also interesting is the slight angle of the cover tape pull. That makes sense because it provides greater shear on one side of the cover tape, which might lessen the force needed to initiate each cover tape pull. Also, I do see flanged plastic pulleys and what looks like a stepper-driven pinch gear.
>

> --
> 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/msg/openpnp/-/v1M7msJfvPcJ.

[Paul Kelly]

The rollers that the cover tape travels over have flanges, the gears keep significant tension on the cover tape and this ensures that it tracks straight from the last roller through the gears.

Paul Kelly | Design Engineer
CASWA Pty Ltd
9/4 Roper Street, O'Connor WA 6163
M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com


www.caswa.com

Be earthsmart. Please consider the environment before printing this e-mail

-----Original Message-----
From: ope...@googlegroups.com [mailto:ope...@googlegroups.com] On Behalf Of Karl Lew
Sent: Friday, 15 February 2013 5:35 AM
To: ope...@googlegroups.com
Subject: Re: [OpenPnP] Re: Open source feeders

--
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/msg/openpnp/-/9qqrmXAjuwAJ.

[Jason von Nieda]

The thing I am not seeing is... what drives the tape? Surely it's not driven by the pulling of the cover film?

Jason

On Thu, Feb 14, 2013 at 3:01 PM, Karl Lew <ka...@firepick.org> wrote:

HLO, thanks for the link! It is very helpful.

I like how the cover tape is redirected down to gears BELOW the feed plane. Also interesting is the slight angle of the cover tape pull. That makes sense because it provides greater shear on one side of the cover tape, which might lessen the force needed to initiate each cover tape pull. Also, I do see flanged plastic pulleys and what looks like a stepper-driven pinch gear.

--
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/msg/openpnp/-/v1M7msJfvPcJ.

[Michael Grant]

I was thinking the same thing...

[Paul Harrison]

I've used one and it does help - however I have another problem that's caused me to look into DIY Pnp - I have essential tremor and am rapidly approaching the time when I wont be able to place TQFP packages, even with meds. Besides, I like a challenge :P

Paul H

On Wednesday, February 13, 2013 9:43:17 PM UTC-6, Jason von Nieda wrote:
> That could well be. I've never used one, but I just watched a few videos and it looks like a very helpful machine. That said, I'm still not interested in hunching over a desk and doing the work by hand.
>
>
>
> I guess what it comes down to is that this something I want to do. Like many open source projects - this is an itch that I feel like devoting my time to scratch. In the end, it doesn't have to be useful to anyone but me.
>
>
>
> Jason
>
>
>
>
>
> On Wed, Feb 13, 2013 at 7:26 PM, Paul Kelly <te...@caswa.com> wrote:
>
>
>
>
>
>
>  
>
>
>
> What I am targeting is the person who wants to make a handful of boards a week. Or maybe even less. Maybe they, like me, just don't like being hunched over a table with magnifiers and a pair of tweezers to assemble prototypes.
>
> Ahh, now this is where I think the misunderstanding may be. I’m not talking about magnifying glasses and tweezers, I’m talking about an assisted manual pick and place system.
>
> When we started out, I was lucky to be able to assemble 10 units a day with tweezers and a magnifier.  When I bought/modified my assisted system I could do 10 boards in  an hour.
>
>  
> It’s hard to get your head around the difference that having a vacuum pencil that only moves in Z, A, X, and Y makes.  A moving wrist support and simple manual feeders finish it off and make it a vastly better proposition... For bonus points, add a video camera with a display.  I really believe that this is the appropriate technology for the user group you describe.
>
>
>
>
>  
>
> When you frame the question as "What does a hobbyist need?" I think the answer is very different from "How do I run my business with this?"
>
>
>  
>
> Jason
>
>  
>
>  
>
>
>  
>
>  
>
>
>
> On Wed, Feb 13, 2013 at 1:25 AM, Paul Kelly <te...@caswa.com> wrote:
>
>
>
>
>
> Actually, I just remembered another metric we’ve  calculated. It’s very relevant to the openPnP space (for some of you).
>
> The stuff we make has between 30-150 components per board, with most board having <<100 components.
>
> It’s all pretty normal gear: a micro, a power supply or two, a couple of special IC’s like RTC’s or amplifiers. Nothing exotic.
>
>  
> You’ve seen the video, our old girl runs at about 1000CPH.
>
>  
> We reckon it costs between $1000 and $2000 per hour in parts to run the machine. Ie that’s the dollar value of parts it places on a continuous basis. Now your boards may be different, but I’d bet they would still be in the high hundred$ per hour and remember, we have wholesale accounts with the big suppliers, so we get better pricing... Now, the machine produces between $5000 and $15000 of boards per hour, so I’m not (or at least my accountant isn’t) complaining..
>
>  
> In the context of that, a $2000-$3000 pick and place machine seems unnecessarily thrifty to me.
>
> Not trying to demotivate, just a: explain where I’m coming from a little better, and b: point out that if an OpenPnP kit cost $5000 or $10000 (and came with feeders) then it would still be a bargain...
>
>  
> PK
>
>  
>
> Paul Kelly | Design Engineer
>
> CASWA Pty Ltd
>
> 9/4 Roper Street, O'Connor  WA  6163
>
> M 0402 177280 | P +61 8 9277 0900 | F +61 8 9467 0550 | E te...@caswa.com
>
>
>
>
>  

[Jose Lopez]

The driver system is in the magazine, the "tray" or basket that holds the reels and the feeders snap into...

This design separates the feeder head and the drive system so setup time is faster..

[Karl Lew]

It is also cheaper because you can have more tape holders than feed motors