I simulate peak detector circuit form well known Horowitz & Hill book. I use LTSpice for simulation. I use AD8045 op amp. According to data sheet AD8045 has input bias current of 6.3uA max, but in the simulation results I observe current of order of 1mA flowing into the op amp's inverting input. What is the reason for that? Incorrect model or I misunderstand the circuit operation?
Here is the schematic:
V3 gives the input pulse (1.3V amplitude, 1us long), V1 and V2 give bipolar supply ±3.3V to the op amp. R3 and R4 are for current measurement only.
I use AD8045 SPICE model from AD site. Simulation results:
Green - AD8045 input signal.
Blue - AD8045 output.
Red - C1 voltage
Cyan - R3 current.
I assume that R3 current is equal to op amps noninverting input current.
When input pulse ends (starting at 1us), R3 current (same as AD8045 input current) of about 1mA shows up and discharges C1 rather quickly.
Why does this happen if AD8045 has input bias current of 6.3uA according to data sheet?