I'm designing a Potentiostat and one of it's key parts is a current-to-voltage conversion stage. In this stage I need to detect currents ranging from hundreds of picoamperes to some miliamperes. The frequency range should be of 1Khz or less, so not much of a problem here.
My first option is using a transimpedance amplifier with the AD8641 and different sets or resistors for the feedback loop. I was thinking of inserting an analog switch such as the ADG782 into the feedback loop and use it to select different resistors for each current range (100M, 10M, 1M,..). But the problem is that this analog switches have leakage current of tens of picoamperes and present a capacitance between terminals than can distort the measurement. I've never used analog switches for this kind of work, so is it possible? It may seem to be inadvisable to introduce these kind of components into the feedback loop. Or at these low frequencies it doesn't really matter?
If this can't be done this way, my other option is using different transimpedance amplifiers each one with a different resistor according to the desired current range and use the analog switches before and after them to choose which one to use. For the lowest current range, the same problem with the leakage current arises: tens of picoamperes leaking through each unused input may represent a high percentage of the measuring current. Am I correct in this?