I need to integrate 1 pA currents from a TE-cooled photodiode for 100ms and the IVC102 seems ideal for my needs. Reasonable drifts are not a problem since I will be calculating a differential signal from the last 100ms measurement. I tried modeling the setup with the IVC102 Spice model in Tina 8 Classic but I get a 600 mV drift in that 100ms even with the switched input open. That is a lot more than I expected based on 1 pA input bias currents. The spec sheet mentions using the IVC102 for 100ms measurements so I expect that it will work when I try it but a model would be useful. Is the model wrong... have I misunderstood the specs... or am I missing something in the proper setup of the TI website Spice circuit in Tina 8?