by nyazzip » Sun Apr 13, 2008 2:32 am
i've never fully understood the "point to point is best" argument. i understand that original PCBs can break down over the course 4 or 5 decades, and can be harder to repair/modify, but lets say i only intend to keep my PCB amp for a mere 30 years, and have no intention of "modding" it.... if PCBs were so horrible, i doubt they would be used with such frequency, in military and ET applications. to me, "point to point" equals many more inches/feet of wire the signal has to squirm around in, and many more square cm of unshielded/poorly shielded solder area exposure, before it finally reaches the output tranny. am i wrong to look at it this way?
i worked for motorola and for a brief while it was my job to operate a machine that applied a solder paste to a blank CB(no "P" yet) via robot controlled stencil and squeegee...after this artwork was done, the components were robotically placed (a few manually)down the conveyor line, atop the "wet" paste, then the whole shebang got baked in an IR conveyor @300C or so, melting the metal particles suspended in the solder paste, and presumably "cooking off" the rest of the solder vehicle.
the boards seemed pretty durable and the yield was high; by far the biggest error was with the robot placing components innaccurately....granted, many chips were much smaller than a pinhead, and placed at a rate of many cycles per second, with an accuracy of 0.2mm or better...the component-placing bots sounded and rather looked like machine guns during a production run; i think they were for the most part pneumatically powered
anyway, i came away from that gig with a respect for the PCB. cheap? yes. inferior to 1930s style "point to point"? not sure why that would be.