turns out a problem we've had at work for several days is because philips's semiconductor division (long since divested) decided that pin 1 on a TO-92 ought to be the rightmost pin
why would you *do* that
-F
every other TO-92 we've ever used has had pin 1 on the left pin, which is also the case for all TO-220, TO-247, TO-263, SOT-223, SIP, and bespoke packages with pins all in a row that we've ever used
why would you do it backwards
-F
this is not a major problem because
- jfets are super resilient and can handle that
- the jfet in question is literally the cheapest one we could find on digikey and we have like twenty more of them
- TO-92 packages are such that there's no special mounting requirements so we can just flip it around
it's just frustrating that this project has been stalled for over a week because of this (and other projects taking up time)
-F
anyway this meant that we had the gate and source terminals swapped on a jfet which of course means everything was broken because trying to apply a negative gate bias as is normal for an n-channel jfet was instead just forward-biasing the gate
-F