shop talk from my electrical engineering days:
I think my favorite power-related mishap involved the board layout engineer using the wrong DC barrel plug in the layout (center-negative instead of center-positive)
none of us caught it during layout review because we were focusing on the complicated things like the FPGA and the main component
the board is normally powered by USB so everything was testing fine on the first prototypes
until like Friday afternoon when we were like "okay let's make sure the DC power failover works" and we go to plug it in and the fuckin power-source-selection IC pops and leaves a crater
that was fun
My second internship we had a power supply known as the Trace Breaker. It was a 15V, 50A linear supply that weighed a fucking ton. You'd use it on boards with a short somewhere and ramp the voltage slowly while using a thermal camera. I actually had a board I designed fail and we couldn't break a trace even at max amps, because the PCB shop failed to etch a corner of the board under some connectors! It was a huge power/ground short that was like 20% of a layer. XD
@zetasyanthis @rey for stuff like that, i think an H-field probe is a lot easier to use
if a lot less impressive
remind us to tell you about liquid crystal hotspotting sometime
-F
@zetasyanthis @rey We once put a piece of #30 gauge wire wrap wire across the terminals of an "ultra high current" (capable of up to 1500 A) SMU and programmed it to just do pulses of current increasing until it breaks
it made it to several amps before it just melted
-F
@zetasyanthis @rey this actually had a purpose, believe it or not: we wanted to determine a rough measurement of the fusing current of a similar type of wire
figured this was a good enough approximation so put it in there and let it do its thing
it actually fried the socket we poked it into in the process, which indicates there was a lot of contact resistance
-F
@zetasyanthis @rey you should hear the clunking of the contactor when we turn it on
it can only do very short pulses of 1500 A, of course
It came with banana plug cables with a little mechanism inside the plugs that makes them expand when you twist a collet, to increase contact pressure. they're also on some #6 AWG welding cable instead of the more conventional for banana plugs #14 AWG
-F
@zetasyanthis @rey it's an Agilent N1265A ultra high current unit by the way, hooked up to a B1505A semiconductor parameter analyzer
-F
@zetasyanthis @rey i wish you could get stuff like this without being a big company
but this thing was bought used, at like a 50% discount, and it still cost more than a house
-F
Oh I bet it did. There was a point where a previous employer of mine had about a million bucks of test equipment that I had full run of. It was a blast. FSW26s spectrum analyzers from Rhode and Schwarz, Agilent signal gens, Tek and Agilent oscilloscopes, not to mention multiple Cesium references and cleanup oscillators...
@zetasyanthis @rey we work for a startup so we don't have *that* much fancy equipment, the 1505 is about the fanciest thing we have
the scope we use is a lecroy ws4104hd, a 1 ghz thing from lecroy that cost $10k because we got it on a deep discount
other stuff in the lab is more pedestrian, there's an hp high-voltage power supply from the 1960s (yes, actually the 1960s, it still works perfectly), a handful of Instek power supplies, a Siglent function generator
-F
@zetasyanthis @rey still doesn't work completely right but it does the job. it's older than we are, so i'm just glad the refrigeration system is still working--i'm pretty sure it's insulated with asbestos so we don't really want to open it up for repairs
-F
@zetasyanthis @rey this wasn't a thermal chamber, it's a more specialty piece of eqiupment. it's got a tiny chamber (maybe 1 L in size) on the end of an articulated arm, that can be lowered over a portion of a circuit board and control the temperature of *just* the DUT by blowing hot or cold air on it
it's a fascinating piece of equipment, and it has quite an astonishinig temperature range--from -80 to +225 °C, nominally. ours doesn't quite reach the low end of that
-F
@zetasyanthis @rey it does take time to get to the low end of the range. when going below ambient, it even has a compressor startup sequence that takes 30 minutes, and then it takes about another 30 minutes to get down to -70ish -F
@zetasyanthis @rey oh we also have a temperature forcing system that we got for $1k used, plus $1k shipping and two dozen hours or so of repair work and learning how to do plumbing to fix it
-F