Shopping for low C PFET

On Wednesday, August 22, 2018 at 5:32:14 PM UTC-4, piglet wrote:
> On 21/08/2018 20:16, George Herold wrote:
> > On Tuesday, August 21, 2018 at 6:05:06 AM UTC-4, piglet wrote:
big snip
> > George H.
> > ps
> > I've been f-ing around today trying to go from measuring 'dc' photo current
> > (in my SPAD as only APD) to photon counting.  But it's a bit of a hard hill
> > to get over.  If I have a maximum count rate of 1e6 that's ~0.1pA..
> > (e~ 1e-19) and APD gain of 100 gives me 10 pA.  Not very much, and I'm
> > grubbing around in dark currents/ offset voltages and bias currents.
> > 
> > GH
> > 
> 
> Since all you want is to replensish a little bit of lost charge I wonder 
> if it can be done from the low side with a small capacitor?
Yeah I think that's Mario's (see paper's upstream) first low 
side quench and reset circuit.  He pulses in some charge 
through a cap.  It's a diode and he takes advantage of that.
(sometimes my model of a diode is a switch, and that stops 
me from thinking in reverse. :^) 
Then my power supply has to swallow the reverse over-voltage?       
> 
> John Larkin's resistor pullup to EHV clamped to the working voltage is 
> gloriously simple.
Umm, I don't know/ recall it.  

> 
> piglet

I need (maybe) a delayed reset and blanking pulse.  
I've got 74hc14s, diodes, R's and C's,

I was going to ask for links with circuits,
but it's simple enough, (I think) I can use the RC 
delay on one side and the diode on the other.
That still leaves a lot of permutations.  

George H. 


  

On Thursday, August 23, 2018 at 9:34:18 AM UTC+10, George Herold wrote:
> On Wednesday, August 22, 2018 at 5:32:14 PM UTC-4, piglet wrote:
> > On 21/08/2018 20:16, George Herold wrote:
> > > On Tuesday, August 21, 2018 at 6:05:06 AM UTC-4, piglet wrote:

<snip>

> I need (maybe) a delayed reset and blanking pulse.  
> I've got 74hc14s, diodes, R's and C's,

A 74HC14 isn't all that quick. There are faster flavours of CMOS around and ECLinPS is even quicker.

The faster stuff doesn't cope with large voltage swings, so you have to cascode the output into a quick transistor with a decent collector-to-emitter breakdown voltage. I dug up something suitable in a earlier thread on this subject.

> I was going to ask for links with circuits,
> but it's simple enough, (I think) I can use the RC 
> delay on one side and the diode on the other.
> That still leaves a lot of permutations.  

The literature was full of them.

Keeping the actual avalanche period as short as possible seems to be a good idea.

The charge carrier generation that creates and sustains the avalanche seems to dump energy into the channel, which shows up as after-pulsing, and the less of that, the better.

-- 
Bill Sloman, Sydney