Diode recovery pulse generator

Ok, so if I understand the procedure properly, you are to charge a diode
with some forward current, then reverse the terminals, putting it in reverse
recovery, and when you have gobs of current flowing, and the charge carriers
suddenly run out, and it says oh *** and makes a huge dI/dt and switches
off. Or something like that.

Well here's my circuit to test it.
http://webpages.charter.net/dawill/Images/Diode%20Recovery%20Pulse%20Generator.gif
Note the circuit is optimized for turn-on only (the 2N4401 output, without
FET load, has a risetime of about 8ns, comparable to my signal generator,
coincidentially), so repeat rate is pretty crappy (~200kHz).

With FET, Tr is about 50ns. Er.. I forget if that's before or after Rg.
Drain risetime is pretty spanking, of course.

So, when the FET slams on, current in the two turns of hookup wire quickly
rises, and the diode goes reverse... after some time, it plinks and the
inductor discharges as a negative-going flyback pulse, after which the
voltage falls further as the MOSFET saturates, then turns off and everything
relaxes until the next bit of excitement.

But the thing is, I went through pretty much all my diodes and the best I've
seen is a pulse around 40ns across at the base (about 20V tall with supply
as shown). I've got the best results from high speed damper diodes (1.5kV,
<200ns trr, etc.), and the worst from power rectifiers (1.5kV, trr ~1us).
Schottkies of course just ring (gimme a break, it's lashed up on a
protoboard), with no RR to speak up the FET just slams on.

Where's the 1ns *** everyone else seems to be getting? Step recovery
diodes?

Tim

--
Deep Fryer: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms


.

Quantcast