something I tossed together
Richard Spelling
running gcode extrapolated from PCB software location data and robot
layout files with program I wrote. looking forward to openpnp
conversion when it gets to that point.
currently blind and does bump alignment.
seem to work pretty good, still need a bit of tweaking. waiting on the
60 8oz lead sinkers I bought so I can have more than three reels
"live". Planning on building actual boards within a week.
Richard Spelling
Bryan
I presume that's some kind of flux or pre-wash in the trough to the upper right?
Bryan.
some people can't open the video file, here:
https://www.youtube.com/watch?v=wW-v0k08m-g
...
Richard Spelling
Still tweaking, but it's coming along.
the "trough" is the bump alignment hole.
On Mar 19, 10:16 pm, Bryan <gru...@gmail.com> wrote:
> Nice :-D A machine that works at all is better than no machine at all (as
> is still the case here.) And this one seems to do a pretty fine job
> already. :-D
>
> I presume that's some kind of flux or pre-wash in the trough to the upper
> right?
>
> Bryan.
>
Bryan
Running blind is fine by me, if it has stuff like this to take up the slack. I guess a simple photo-diode could be used to detect if a part was picked up at all or not, too. So many possibilities for the man on a budget. ;-)
Of course, if the clever camera optics is available when I finally get around to building a machine, I'd obviously use it. But I find the eloquence of these 'simple' mechanical solutions fascinating.
Bryan.
Richard Spelling
Actually, you could easily use a vacuum sensor, as the vacuum in the
line goes up when it has a part.
Still, the whole point of the bump aliment, you have enough vacuum to
grab the part even if you only have a small section of it and you
needle is hanging off the side. You bump it to fix that. Also, helps if
your needle isn't perfectly aligned with the axis.
It does slow it down by about half. When I get it tweaked to pick up the
parts (at least the small ones) pretty close to the center of the part,
I will disable the bump alignment for those particular ones.
re:optical feedback
Optical is nice, and may go there eventually. But, millions of CNC
routers can't be wrong. As long as you stay in the envelope, you can do
precise work repetitively.
Plus the reflow process is pretty forgiving. We do them by hand now, and
you only have to get them so the leads are somewhere on the pads. You
can actually have quite a bit of mis-alignment without it ever showing
in the final product.
On 03/20/2012 02:23 PM, Bryan wrote:
> Ah! That's clever. So it moves the part to each of the four walls to get
> it centred on the pick-up tool then? A simple, physical approach to what
> is otherwise done with complex optics. Nice. I learned something new.
> (again :-)
>
> Running blind is fine by me, if it has stuff like this to take up the
> slack. I guess a simple photo-diode could be used to detect if a part
> was picked up at all or not, too. So many possibilities for the man on a
> budget. ;-)
>
> Of course, if the clever camera optics is available when I finally get
> around to building a machine, I'd obviously use it. But I find the
> eloquence of these 'simple' mechanical solutions fascinating.
>
> Bryan.
>
> On 20 March 2012 16:42, Richard Spelling <rls...@gmail.com
> <mailto:rls...@gmail.com>> wrote:
>
> Thanks.
>
> Still tweaking, but it's coming along.
>
> the "trough" is the bump alignment hole.
>
> --
> You received this message because you are subscribed to the Google
> Groups "OpenPnP" group.
> To post to this group, send email to ope...@googlegroups.com.
> To unsubscribe from this group, send email to
> openpnp+u...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/openpnp?hl=en.
--
Visit my online store for solar electronics: http://www.richardsfoundry.com
------------------------------------------------------------------------------
NOT sent from a Blackberry. Sent from a free hand-me-down PC running
free open source Ubuntu Linux... :-P
Bryan
Without having actually tried it personally, I do like the idea of having a camera head, if only to make alignment and human monitor easier. The OpenPnP software is able to handle multiple boards of "random" alignment, as I understand it. So having cross-hairs from a USB microscope or whatever to wiz about and mark off reference points would seem to be a good thing, regardless.
I have also theorized that one of the greater values of optical assistance in high volume production scenarios would be to allow logging of events that "need a human to check", rather than leaving it all to human inspectors. My guess is that such would save a lot of time in the first stage QC areas. (I'm just a self taught hobbyist you see -- hence the theorizing. :P)
I've also witnessed a surprising number of things that can go wrong at the component pick-up point -- especially as the speed is wound up. I guess this is the sort of thing that 'staying in the envelope' is about, as you put it.
Finally, ironically, since I started in SMD about a year ago, my hand placement and manual re-flow (oven) skills have increased to the point where I think I'm doing most one-off boards faster than I could set up and run a PnP machine. I've gone from, "Sheesh! This is far too tedious without a machine!" to "Toss me your boards for populating. No problem!" :-P (Even the basic skills of knowing how to keep ones hand dead steady for the task is a huge leap forward for me!) But I still definitely want to go the automated direction for potential larger build runs -- and just for the plain fun of it, of course.
Bryan.
Thanks. I don't think I invented the idea. :-)
Actually, you could easily use a vacuum sensor, as the vacuum in the
line goes up when it has a part.
...
re:optical feedback
Optical is nice, and may go there eventually. But, millions of CNC
routers can't be wrong.
...
Richard Spelling
problems with the cheap chinese bore scope camera, so I tossed it and
just do alignment by "eye" for now.
> <mailto:r...@richardspelling.com>> wrote:
>
> Thanks. I don't think I invented the idea. :-)
>
> Actually, you could easily use a vacuum sensor, as the vacuum in the
> line goes up when it has a part.
> ...
>
> re:optical feedback
>
> Optical is nice, and may go there eventually. But, millions of CNC
> routers can't be wrong.
> ...
>