PIN Photodiode: needs biasing?

PIN

Is

that

Robert,

What I have noticed with unbiased silicon photodiodes in general is that they are pretty linear up to about 0.2 or 0.25 V output, no matter what current + load resistor combination is used. By the time the output gets up to 0.3 or 0.4 V, the response (output vs. incident photo power) is pretty well flattening out. You don't mention what output current or voltage you are getting, but hopefully this info helps you determine if you should need biasing for linearity.

That being said, it has been 12+ years since I worked with a PIN photodiode, and I'm not sure if they behave vastly different, in terms of linearity, vs. a regular PN type. If speed is an issue, then as you noticed biasing will help you with that.

Regards,

Mark

PIN

Is

that

Provided you built a good TIA where the voltage at the photodiode node does not move the reverse bias is primarily intended to reduce the diode capacitance and thus improve speed. It also allows you to reduce the compensation cap Cfb for the TIA, making it a bit more hot. But that really only matters in the 100MHz+ speed ranges, not so much for an LF series amp.

I finished a photodiode design about a month ago and this one went to

100MHz. I moved the bias around a bit but even there it didn't matter. I just kept it in case the client wants to go higher one day.

--
Regards, Joerg

http://www.analogconsultants.com

PIN

Is

that

The datasheet I found doesn't start at zero. Do you have a link?

PIN

Is

that

The datasheet I found doesn't start at zero. Do you have a link?

It is a crummy optically-scanned image document. Do you have a better one, perhaps one that is not just an optical scan?

Robert Scott Ypsilanti, Michigan

That is the same one I found. Since the graph doesn't go to zero, how did you conclude the shift for 2v back bias?

Hello miso,

They obviously measured with an ampere-meter with 100kOhm input resistance. This is nonsense of course in a photo-diode application when you measure over a wide range of light(photo current). As soon as there is a voltage drop of a very few 100mV on the load resistor, the measured current isn't correct because the photo diode gets internally forward biased by this value. Either a negative bias voltage will help to prevent that or a short circuit current measurement (transimpedance amplifier).

Best regards, Helmut

Robert is talking about the photocurrent (bottom of p. 3 in the .pdf file), whereas you seem to be talking about the graph of leak current (top of p. 3).

Mark

Does this mean the efficiency of the detector increases with bias? Or are they lumping leakage in with photocurrent?

one,

If you mean the ~1 uA increase between about 20V and 30V reverse bias, it is definately not leakage (dark) current, which is only several nA.

The increase is probably due to some avalanching (electrons multiplying), rather than a higher quantum efficiency of the photodiode. Note how close the bias is getting to the max rating of

32V.

Mark

one,

OK, that makes sense.