Photodiode Amplifier

The critical point is to mimimise the stray capacitance between the virtual ground summing point and real ground.

Minimising this capacitance minimises the broad-band noise on the ouput of the transimpedance amplifier, and maximises the bandwidth you can get. The Burr-Brown (now Texas Instruments) data sheet for their OPA655 FET-input op amp

goes through the analysis (which starts on page 8 of the data sheet and fills most of sheet 9).

In general, you should aim to tie the case of your photo-diode to real ground. This is usually tied to the anode of your photodiode.

What would you gain by connecting it the other way ?

Rene

Some photodiodes have metal shield cases, and sometimes they have only two terminals, with one tied to the shield. In that case you'd want to be selective about which one you ground.

Hi Rene,

The application is a monolithic one. I have another diode whose anode is fixed connected to p-substrate (which is the chip ground point). The current flowing in the two diodes need to be subtracted. Hence this diode has to have its cathode connected to psubstrate (im just going to use two antiparallel diodes to do the subtraction..).

Thanks qq

Rene Tschaggelar wrote:

In article , wrote: [....]

No, it often isn't. One surface of the photo diode is facing towards the light source through a clear (nonconductive) window. The other surface is facing towards the mounting hardware. If there is much electrostatic noise in the environment, you are better off to ground the face of the photodiode and take the extra capacitance. This assumes the choice wasn't made for you by the mechanical design.

You can get away without using ground by going differential. See my schematic in ABS.

Regards,

Boris Mohar

Got Knock? - see: Viatrack Printed Circuit Designs (among other things)

void _-void-_ in the obvious place

Your output voltage will want to range from 0 to negative voltages, if I have my anodes and cathodes straight.

Yes, it seems your output polarity will be reversed. Can your opamp swing below ground? The other consideration, in either zero-bias configuration, is the fact that you're stuck with the zero-bias capacitance of the diode, which is typically much higher than if you reverse bias the photodiode. This will severely limit your BW and will require more compensation of the feedback resistor to prevent instability.