On 12/27/2015 02:35 PM, EBo wrote: > On Dec 27 2015 10:47 AM, Jon Elson wrote: > Hmm, that sounds really GOOD! I think my repeatability is > probably closer to .005" or so. Just BARELY good enough for > anything finer than SOIC lead pitch. > I think you can probably get much, much better than that. Well, this is a 15 year old machine with some significant production time on it. It is servo-controlled, but has high-lead ballscrews, so the resolution of the encoder counts isn't all that high. It is a 1700 Lb machine with NSK double-row linear ball glides for the X and Y axes.
<openPnP> I am curious about what they got right and wrong in your opinions. I have heard of the project, but never had a need... Well, I don't think they have much error recovery in it. That seems to be a REAL key feature. Detect (either by vision of vacuum sensors) that you don't have a good pickup of the part, dump it and try again. After a couple mis-picks, call for help. >>> The entire concept of using G-code for a PnP machine is not a good >>> idea on >>> any level. >> Right. It is what is there, but not a proper fit to the task. > So what do they use? The only system I know well is the Philips CSM (and Yamaha CM) machines. (Yamaha made the early Philips machines, the hardware is very close, the software is similar.) They have a board file, that lists the feeder #, X and Y and rotation coords to place the part, and which head to use. (My machine has 3 heads which are mounted on the same X-Y positioner. They have belts that make them all rotate together, too. They have independent Z motion by air cylinder. Then, there is a component (or feeder) file, that has a lot of data on each feeder position. It tells the tape width, by which the machine can compute the centroid of the part in a stock feeder, the proper head rotation to pick up the part with (so the long side of the component will be aligned with the long alignment jaws of the head) how many times to pump the feeder advance lever, how much vacuum indicates a good pick-up of the part, whther mechanical alignment is needed on this part and other stuff. if the part is in a tray, it needs to know the size of the tray in number of pockets for X and Y, the pitch of the X and Y pocket spacing, the current pick-up pocket in the tray, and the coordinate of the first pocket in the tray. I teach that with the teach camera. Since the machine can pick up 3 parts at a time, it sometimes does that to optimize head motion. Picks up 2 or 3 parts when over the feeders, then deposits the parts when over the board. > > Do you have one you could test on? Or something you can set up as a > test bench? I would be interested in following your progress, but I am > unlikely to be free to contribute. I do know of someone that > wants/needs a custom parts machining center. This type of logic process > would likely be more amenable. > > Well, I really do NOT want to tear apart this perfectly running machine to try some experiments. Without somebody to make a big head start, it seems like it would take a LONG time to try to implement something like this. Jon ------------------------------------------------------------------------------ _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers