Schottkys in series

Sure. Leakage goes up softly as voltage increases, so schottkies self-balance.

Hmm it didn't work... well all the voltage drop was over the first one. I put three 1n5711 in series and then 1 meg to ground.. and hung that on the high side of my spad. I was expecting it to look like 1 pf (or less) of loading... it was a lot worse (maybe 10 pF?) hang on... recharge time of 8 us. ~20pF!

George H.

I wonder how some vanilla 1N4148 would fare?

piglet

Oh snap.. my f-up I forgot the 2 pF coupling cap between 16 pF 'scope probe and the HV node.

It looks fine now. (sorry)

George

Makes no sense.

Sure. The leakage increases smoothly with bias, and the capacitance is low, so they equalize themselves.

Cheers

Phil Hobbs

That's a different question. If you want the _voltages_ to equalize and not just the leakage currents, you have to do it manually.

Cheers

Phil Hobbs

If they share back-voltage equally, it's because of matched leakage (not tightly specified, but you could parallel 'em with resistors) and/or matched capacitance (not tightly specified, but you could parallel 'em with capacitors).

Best practice is to match those characteristics and thermally-couple all the pellets (make a stick rectifier of two or three in series). Second-best is to add in those parasitic components, R and C.

Diodes don't suddenly blow up at their rated voltage--it's just an arbitrary limit set on the continuous I(V) curve that will prevent damage. Putting them in series forces the leakage to be equal, and if it's at a safe level for one, it's safe for all.

Cheers

Phil Hobbs

Not sure what you are trying to say.

I have not tried Schottkys in series, but goto 12. Do you rely on some sort of zenering in reverse?

Who are you really?

Cheers

Phil Hobbs

Not zenering, but schotttky leakage (unlike PN) increases about exponentially with reverse voltage, so schottkies in series equalize and are happy. The usual damage mechanism to over-voltaged schottkies is thermal, and in a series string all diodes see the reverse leakage current of the lowest-leakage part, so it's OK.

People also sell high-voltage PN diode stacks that don't have distinct equalizing parts. In ancient times, diodes might have had some negative-resistance avalanche mode that made series strings unreliable without added RCs. Or maybe the hazard was/is an urban legend.

PN diodes can sometimes act like step-recovery diodes and make nasty reverse-recovery spikes, even at 60 Hz. So paralleled capacitors are sometimes seen to supress that EMI.

Yes here we call those 'rattle capacitors' used to eliminate the ratlling noise. in radios.

OK, now it is clear to me, nice if it works, there must be some special cases where 2 schotttkys in series are useful, bit less voltage drop than 1 HV si diode perhaps.

Something to add to the tricks book, thank you.

You call me really?

A nobody. You can safely ignore it...

Tim

circuit,

Reverse biased diodes fail by avalanche breakdown, just like zener diodes rated above 10V and higher. The difference between regular diodes and avala nche rated rectifier diodes and zeners is that the avalanche break-down - w hich starts off very localised - doesn't stay localised for long enough to generate sufficiently intense local heating - a hot spot - to produce perm anent damage.

The voltage required to sustain avalanche is a little lower than that which intiates it, so there is a small negative resistance effect, but not enoug h to do anything interesting unless you design your diode very carefully

Ummm- doesn't "soft" characteristic increase chances of avalanche for series connection? Think so.

No, schottky leakage current is close to exponential on voltage, and the failure mechanism is more thermal than avalanche. Series strings self-equalize. Even individual diode self-heating, which increases leakage, is a stabilizing effect.

The reverse characteristic is completely different from the forward characteristic. Look at figure 4.