Baker clamp to the rescue

Could the one-shot drive the schottky diode directly, and eliminate the transistor? Maybe with some gate/buffers if a lot of current is needed.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Yeah, I think so. Using some sort of digital drive would be nice. (pulses are born with sharp edges.) but the 'real' spad get's ~10-20V of over voltage.

I should be able to capacitively couple a quench pulse in too, and then piglet sent me some that use a pulsed transformer. (transformers are somewhat magical to me... I'm not sure how to pick one.)

George H.

How big does the quench pulse have to be? Your sketch implies a volt or two.

How fast do you want the quench to be?

Actual CMOS one-shots are pretty slow; you probably need to make your own. I had to do that recently.

Mosfet gate drivers are interesting if you want to swing tens of volts pretty fast.

Any edge-driven loop can hang up.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Right, it depends on the over voltage.. in theory the quench has to be a little bigger than the over-voltage.

10-20 ns for the whole thing would be wonderful I think I'd settle for 80-100 ns.

Yeah I made my own from an 74hc14 ~10-100 ns pulse width.

OK thanks. Hey I've got the thing working with the real spad... but it's a bit weird and delayed. (At times like this a fast four chan. 'scope would be nice!) Here,

The single orange trace is the compartor out, Brown is the base drive (width ~50 ns) And green is the anode of the spad...via a buffer opamp (I understand nothing of that trace.)

I'll be careful.

Here's an unrelated statistics thing. I was checking the dead time by setting the 'scope to 4 sec. persistence. I get pulses right after the first so dead time looks small. But I hardly get any pulses at all before the one that triggers the 'scope. That strikes me as weird... I naively expect random pulses to be random in time. I think I can hand wave my way out but...

Pic worth 1k words

This is all to slow, and I can't get above 5V... but at least it works :^)

George H.

Hey, Spad fun again!

Isn't that just Baker clamp 101? - Without the baker clamp the discharge is deeper because the transistor saturates and collector goes

Good suggestion. I seem to recall goal is 20-30V swing as fast as possible, current is minimal. I suggested a 5V logic line driver stepped up but your gate driver even better as needs less or no step up. Heck, even if step up needed a charge doubler might do.

piglet

The saturation voltage of the BJT is lower than V_BE - V_f of the diode.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

OK as soon as we order 10 'real' spads I'm sending you one.

Re saturation: Most likely, I guess I don't understand it very well. It went below 'my' expected c-e voltage... but there's ~ -100V on the other side... I'm pulling both sides down.(?) I tried first with a fet, and that also went deeper than expected.

I look at my 'scope trace and I figure, there must be stuff happening faster than I can see.

I should think of diodes as low C switches, when biased properly.

George H.