0V a
ing
Earlier this year I revisted zener diode noise, (zeners in the avalanche range.. above 8 Volts.) The noise was very part dependent. With maybe a factor of ten spread for one zener voltage. (My model for this is that the breakdwon is happening in one little section of the zener. There is some amount of charge 'stored' in that region, and the pulse size is just how much charged is stored there, so different devices have a different regions.. different pulse sizes.) But with a 20V zener (1N5250) the pulse sizes (and noise) are pretty much all the same. (less than 10% difference.) My modle for this is that the whole device breaks down when it zeners, so I get a pulse that is the capactiance of the device, which is pretty much the same for all of them.
I'm guessing the pulse from the reversed biased LED are similar... the whole LED is breaking down.
There's a nice article by K.G. McKay (Phys Rev. (94) 887 May 15, 1954) about avalanche break down in Si.
ces
axy
y,
eas
ial
at
east
or
fed
Oh the pic's I posted were w/o any amp. (I needed an amp to drive the
50 ohm input of the counter.)The amplitude grows as the bias voltage gets closer to the avalanche knee. So sometimes even a volt or more.
wast
ed
kink
s and
Grin, the fall time is even worse. All done with a 40MHz 'scope, though that doesn't appear to account for slow rise* time. I needed the 100k ohm bias resistor, with a 10k ohm the pulses were sharper, but of constant amplitude and sometimes longer pulse width. (I don't have any idea why it is like that.)
George H.
(*Well my 'scope shots are inverted so by rise time I assume we are both talking about the turn-on time.)