Choosing comparator for a circuit which senses current from diode and quenches down bias voltage

Thread Starter

Priya_ece

Joined Jul 23, 2019
11
I am designing a quenching operation for controlling bias voltage across my photodiode. The idea is as photons hits it, the diode produces a corresponding current pulse which I want to fed as an input of a comparator which senses the current and outputs a logic which finally triggers the biasing source i.e. when the diode is biased above breakdown, Vr while it detects the photon and gives current then immediately the bias voltage go below the breakdown, Vc (<Vr) then again switches back to above Vr for another detection. And the process repeats at a very fast rate.
 

danadak

Joined Mar 10, 2018
3,682
What is the speed that this has to operate at ?

Have you already done the signal path from the detector to the
comparator input ? If not then you need to post URL to detector
datasheet.


Regards, Dana.
 

Thread Starter

Priya_ece

Joined Jul 23, 2019
11
Hello Dana,
Since the detector's output would depend upon the incident light which I am assuming to be weak. I might use an transimpedence amplifier so I could get a corresponding amplified voltage output for the current. But since I am a bit unsure about the noise level now which would be amplified too so my current focus is on controlling the voltage. The breakdown voltage, VBr is 130V. I wanna operate higher than VBr, say Va I wanna understand here that if my current pulse per incident photon is known (like in Mili or microamp range) how can I choose a switch which probably a comparator here to do the job. The job of taking the voltage below VBr as soon as it senses the photocurrent then again set the detector back to voltage above VBr i.e. Va.
 

Thread Starter

Priya_ece

Joined Jul 23, 2019
11
Also, about the speed, as it depends upon the photocurrent and our light source will kind of make an impulse signal with a gap of few orders of nanoseconds. So the hold on and reset time must be in order of nanoseconds in order to match that and not miss any signal.
 

Analog Ground

Joined Apr 24, 2019
216
Are you detecting single photons or a larger light pulse? Hard to tell from your comments. Quenching is often associated with a photon counting, avalanche photodiode (APD). If photon counting, then the light level is only relevant to how long it takes to quench and get ready for the next photon. The signal pulse size will be fixed by the APD gain after the photon is absorbed and starts the avalanche.
 

Thread Starter

Priya_ece

Joined Jul 23, 2019
11
Are you detecting single photons or a larger light pulse? Hard to tell from your comments. Quenching is often associated with a photon counting, avalanche photodiode (APD). If photon counting, then the light level is only relevant to how long it takes to quench and get ready for the next photon. The signal pulse size will be fixed by the APD gain after the photon is absorbed and starts the avalanche.

Precisely yes, I wanna count the photons. I'd be operating my photodetector on Geiger mode. I've gone through the quenching circuits and an active quench would be better but I am a bit lack of experience when it comes to buying electronics. As the idea would be, the photocurrent resulting from the avalanche just trigger the comparator which outputs a logic pulse and can trigger the bias supply with high impedance state from a low impedance state so that the bias voltage can go below the breakdown until another photon strikes.

As I can see here, if the triggering is slow or also in case of false switching in absense of new photons, this gonna affect the system performance and counting. I'd appreciate if anyone here can list out exact points I should be careful of and bit on effective design parameters.
 

Analog Ground

Joined Apr 24, 2019
216
Precisely yes, I wanna count the photons. I'd be operating my photodetector on Geiger mode. I've gone through the quenching circuits and an active quench would be better but I am a bit lack of experience when it comes to buying electronics. As the idea would be, the photocurrent resulting from the avalanche just trigger the comparator which outputs a logic pulse and can trigger the bias supply with high impedance state from a low impedance state so that the bias voltage can go below the breakdown until another photon strikes.

As I can see here, if the triggering is slow or also in case of false switching in absense of new photons, this gonna affect the system performance and counting. I'd appreciate if anyone here can list out exact points I should be careful of and bit on effective design parameters.
What you are trying to do is quite difficult and uses very specialized circuitry. Your original question was about a comparator. The goal of the circuitry is to minimize "dead time" which is the time after a photon arrives until another one can be detected. The commercial photon counting module I have used has a dead time of 50 nanoseconds and an output pulse of about 30 nanoseconds So, the comparator needs to have a "propagation delay" of only a few nanoseconds to give the quench circuitry as much time as possible. Then, you have to decide the type of logic signal at the output of the comparator (TTL, PECL, LVDS, etc.). Sounds like you have a quench circuit in mind. What does it need for a trigger input? I suggest getting started by going to Analog Devices (analog.com) and look at some high speed comparators. Also check out their application notes on these devices.

Most photon counting APDs are cooled to lower the "dark count" which is extra output pulses which occur without an input photon. You may have to cool your detector with a Peltier type cooler. Again, you are taking on a complicated and difficult project. The best type!
 
Last edited:

Thread Starter

Priya_ece

Joined Jul 23, 2019
11
Here is an example of a photon counting, avalanche photo-diode module. I think the TS is trying to make something like one of these.

https://www.excelitas.com/product-category/single-photon-counting-modules
Hello,

Thanks for replying. Yes, I realize there are a lot of factors to consider here. I'd need to temperature compensation circuit too as the detector's temperature would rise in case of long time operation.
But I want to approach everything step by step. Once I can gain control over the quench, I'd further try to reduce the false quenches (i.e. photocurrent due to noise and false avalanche).
It seems like you've worked with detector's before. Will it be possible to discuss it by any direct means?
 

Thread Starter

Priya_ece

Joined Jul 23, 2019
11
I started a "Conversation" about this application. Here is the ultimate photon counter following the APD. The superconducting nanowire.

https://marketing.idquantique.com/acton/attachment/11868/f-023b/1/-/-/-/-/ID281_Brochure.pdf
That's okay. I could buy these modules but I am planning to design myself here which may not be efficient and noise-free as these ones but atleast I could learn how the detector really works and how to operate it and take output.

Here again my concern is, my detector's output gonna be current, so if I'd want a transimpedence amplifier which could give me a corresponding amplified voltage output which can be then fed into the comparator which can compare this voltage with a reference voltage (I am thinking to decide the reference voltage as approximate voltage which suggest an avalanche current. That's again should be differentiated with the noise). If this is decided right, then the comparator will get triggered by that avalanche voltage only and can do the quenching. But it again would make a delay as it would also include the transimpedence amplification delay, right?

Also let's assume that this is fixed, and now I need to design what would happen after the comparator signals "high" state. I would use that to drive the bias voltage of the detector to get into a high impedance state, maybe some potential drop so that the supply to detector goes below the breakdown voltage but then automatically get reset until next avalanche happens.
How to achieve this?
 
Top