On Fri, Mar 31, 2017 at 11:32:08AM +0000, David Griffith via cctalk wrote: >One of the biggest challenges for the last >run was getting the QFP-packaged 100-pin chips[1] in a state such that the >pick-and-place robot wouldn't throw a fit about slight differences in lead >position. The stuffing house insisted that I send them new chips.
What an odd problem! Well I have no idea what it feels like to be a CV system so I don't know how to make one happy, but if it were my design I'd lean toward (1) re-design using still-available chips (and hoard the hell out of them this time), and/or (2) hand-soldering. Everyone's jumped in with lots of QFP soldering thoughts. Have some more: - www.oshstencils.com is awesome. - RoHS solder is *fine*. Maybe it wasn't in its infancy but it is now. I don't understand all the hate. I've been using it for almost everything for ~10 years now. Yes I have to use higher temps but that's the iron's problem (PTA8 tip instead of PTA7 on my WTCPT, and twiddle the knobs on the other thingies). - I do QFPs (0.5mm and 0.4mm) using a cheap eBay binocular microscope, Ameritronics solder paste, and a toaster oven controlled by an Arduino with a homemade shield on it (with a MAX6675 for a K-type Sparkfun thermocouple which I position touching the PCB somewhere, START and RESET buttons, a header for LCD status display, and terminals for an SSR which turns the toaster oven on/off). It's *fantastic*. Push the START button and everything's done when you come back (it follows the usual temp profile automatically). I often have to do some soder-wicking afterward (maybe I should be using thinner stencils, and I always overdo it when I use a syringe directly), but the chip stays aligned as the paste melts and the success rate has been excellent. And solder bridges are self-fixing. John Wilson D Bit