Is the quantum efficiency of a photodiode dependent upon its temperature?

[Topher925]

Like the title states, is the voltage produced by a photodiode, or any PV device for that matter, dependent upon that devices temperature? To rephrase, if you have a fixed amount of photons coming in contact with the business end of a photodiode, is the voltage or current generated by the photodiode dependent upon the diode's temperature?

 

[dlgoff]

I would think so since the ideal diode equation is temperature dependent.
http://en.wikipedia.org/wiki/Diode#Shockley_diode_equation"
Wouldn't you expect better efficiency at low temperatures?

 

[berkeman]

At the very least, the reverse leakage current will increase with temperature, so even if the photons are generating electron current, some of that current gets stolen back as reverse leakage before doing anything useful.

Do you have a specific application in mind, Topher?

 

[Topher925]

Do you have a specific application in mind, Topher?

Yes, I'm trying to use a photodiode for a sensor but the thing seems to be drifting all over the place. Its QE seems to change quite a bit over a period of hours and I can't figure out why. The data sheet says the QE should only change about 0.05%/C at the wavelength I am using but its more like 50-75%. The lab temperature I'm using it in only varies by 2-3C over a period of a day so there shouldn't be that large of a drift. I'm hoping I just have something hooked up wrong.

 

[berkeman]

Hmm, that is weird. How much reverse bias are you using in your current-to-voltage converter amp for the photodiode? Are there other sources of noise, either RF or light, that could be causing the drift?

 

[Topher925]

Hmm, that is weird. How much reverse bias are you using in your current-to-voltage converter amp for the photodiode? Are there other sources of noise, either RF or light, that could be causing the drift?

At the moment I'm just using an instrument amplifier (voltage sensing) because it has fantastic noise rejection but I'm going to try using one with current sensing. My original set up used a shunted resistor with the photodiode in reverse bias but I still think I got a lot of drift from that set up too.

The sensor including the photodiode is optically sealed and there isn't any significant noise. I tested the entire circuit and its rock solid, as in no drift for a period of weeks with drastic changes in temperature. Hopefully when I get some current sense amplifiers the drift will go away.