I've got a prototype board of a light tracking device based on a 2 dimensional PSD, but it suffers from important bias currents on one of the electrodes which could possibly clip the input to the supply rail at any moment. I would like to compensate for that offset in order to be sure the signal is never altered, but is it going to have any effect on the PSD's behaviour? How does that work out really?
What I would like to do in particular, is replace the direct ground connection on the non inverting input of the current-voltage converter by an adjustable voltage divider between 5 and -5V on the following schematic:
(also in dropbox) - the gain of the tail amplifier is set to 100.
I have tried injecting currents with a 10Mohm resistor to -5V on the electrode but even if the offset is no more, it seems it tempered with the detector somehow because I picked up immediately an important (several Vpp output!) 50Hz noise when it wasn't there before. That noise decreases when the device is obscured, so it really sounds like some kind of imbalance.
In a nutshell, if I offset the potential on one electrode (well, the opamp will do it for me), is it going to change anything else?