Is that reverse blocking up to 5V?
After 5V, the LED breaks down and conducts. Like a zener I suppose..
But wouldn't a series resistor limit reverse breakdown mode energy at
the LED? Or is there more going on like a solid state chemical
reaction that damages the LED (not from heat)?
D from BC
A reverse junction is very easy to damage. I doubt the series resistor
would save the LED from damage.
I spent quite a bit of bench (aka prober) time investigating metal
fuses on chips. [Trust me, this will become relevant.] A particular
product had odd behavior after trim. It turns out the inductive kick
from the wire leading from the popping circuit was damaging parasitic
diode junctions on the chip. [They would be leaky, which caused
problems at hot test.] The solution was to change the polarity of the
zapper, so that once the fuse was popped, it would forward bias the on-
chip diode, absorbing the remaining energy from the zapping capacitor.
That is, the energy not used to zap the fuse. Diodes are reasonably
rugged under forward bias condition, but reverse bias is another
story. [Incidentally, the fuse popped in under a uS. I don't recall
the rise time.]
Take a BJT and break down the EB junction on your curve tracer. If the
current is kept low, say under 100uA, there isn't much damage. If you
whack the junction, the transistor will be permanently altered in
normal operation, should it still be lucky to work.
When you have a large reverse bias, there is a tendency for the weak
link to pop. That is, the junction in question is not a point source
but distributed. You get a high field, and some piece of the junction
pops. Maybe a crystal defect was present where it zapped.
