I'm planning a design to measure small DC currents (~nA) across a shunt, apply some gain with a series of zero-drift amplifiers, and read the voltage output with a DMM. The goal is to reduce the burden voltage introduced in a typical DMM when measuring current. The LTC2063 / 2064 look to be well-suited to the job thanks to low input bias current and low offset voltage. However, when running a DC sweep in LTspice I get some very significant error introduced by the bias current of the amplifier, which looks to be way above the specified maximum at ~120pA. Even when driving the input from an ideal voltage source, the bias current doesn't change. The input voltage is centered at midrail, right in the middle of the common mode input range. Am I missing something here, or is this SPICE model inaccurate?
[edited by: EHaag at 2:36 PM (GMT -5) on 22 Dec 2020]