# photodiode shunt resistance equation

Discussion in 'Electronic Design' started by Wanderer, Feb 23, 2008.

1. ### WandererGuest

What is the equation for the photodiode shunt resistance relation with
temperature? I know it's logarithmic and got kT/q in it but I can't
seem to find it.

Thanks

2. ### MooseFETGuest

Look for the (non-photo)diode equation.

3. ### WandererGuest

Thanks for the reply. It's probably that simple and staring me in the
face. I know the diode equation I don't see how to relate it to photo
diode shunt resistance so I can get the shunt resistance noise. If I
use ohms law and the bias voltage is zero I'm back to square one.

At one point Aldert Van der Ziel in his book converts the dark current
noise into a conductance to simply his model.

Idark * q/2kT

Is that 1/Rshunt? Is the shunt resistance just a different way to
write the dark current noise?

Thanks

4. ### John PopelishGuest

Wanderer wrote:
(snip)
I don't think so. The incremental shunt resistance is the
slope of current as voltage changes. This is in addition to
the approximately constant (but noisy) current that is
almost independent of voltage but very dependent on temperature.

5. ### Phil HobbsGuest

Dark current noise in photodiodes is almost never the limiting noise
source in a measurement. What are you actually trying to measure?

Cheers,

Phil Hobbs

6. ### WandererGuest

I'm building an incoming inspection tool for one of these

http://jp.hamamatsu.com/products/division/ssd/pd041/pd064/pd065/S9295-01/index_en.html

I'm computing the noise through a power spectral density matlab
program. I see the dark current drop like it is suppose to, but I only
get 6% change in noise from 25C to 5C. I'm told I should see a much
greater change because of cooling the shunt resistance. I know it's
just some bugs in my code, but I'd like to do some calculations to get
an idea of expected results.

Thanks

7. ### WandererGuest

I don't know. Karl Spangenberg has an Appendix on shot noise in
Fundementals of Electron Devices 1957(I have a bad habit of collecting
old electronic books, though sometimes their better than the
internet). He takes the derivative of the diode equation and gets a
conductance qI/kT and calculates the rms noise to get 2qIB. Though he

"The above demonstration is not really a derivation but rather a
rationalization. It assumes that the phenomenon giving rise to
resistance noise is the same as that giving rise to emission noise. It
will at least make the formula seem reasonable and dimensionally
correct."

8. ### Helmut SennewaldGuest

Hello,

How do you calculate the noise?

Is it the integrated noise voltage over a certain bandwidth?

What's the lower and upper frequency limit of your
measurement and noise calculation?

I remember that Rshunt will fall by a factor of two
for every temperature rise of 6 to 10 degree.

Best regards,
Helmut

9. ### Phil HobbsGuest

Judging from the very high feedback resistor and the rapidly rising
noise curve on Page 2 of the datasheet, that gizmo is dominated by op
amp noise above about 10 Hz.

The photodiode capacitance and the feedback resistor form an RC
differentiator that multiplies the voltage noise of the op amp, and
(with this device) once you get as far as the audio region, the PD noise
hardly enters the picture.

Cheers,

Phil Hobbs

10. ### WandererGuest

I take the FFT over 100hz bandwidth convert it to a Power Spectral
Density by taking the absolute value, throwing away half , multiplying
by 2, etc. I then have a function which breaks up the data into
different bandwidth regions, sums the points in the different regions,
divides by the number of points in the region, by the bandwidth of the
region,divides by 2 and takes the square root. I've been experimenting
with different size regions.

11. ### WandererGuest

It's a low frequency dc application. I'll try focusing my analysis
below ten Hz. I've been focusing around 50Hz because I don't trust the
edges of the window.

Thanks