Insert a 10 K resistor in series with the feedback from R sense. The secret is to insert a 1 K ohm resistor in series with the mosfet gate. You need to separate the DC constant-current sink from the AC pulse generator.ġ) I have built many circuits like you have but stayed with a DC stable op-amp like a TL072. To put it simply, you cannot have your cake and eat it too with this type of circuit. R_load is not actually a resistor, it has a nonlinear I-V curve) (Example component choices: THS4541 opamp, BSS816 mosfet. Is there some way to avoid the offset issue in a circuit like this, and have a current source which is essentially exactly proportional over a very wide dynamic range? Is there some way to use a current feedback amplifier instead of voltage feedback? Span errors or deviations from linearity don't matter as much, but zero shift errors are really bad for my use case. I could also add an extra switch on the high side to get a true zero, but that doesn't solve the dynamic range problem (at best I'd have 4 decades of dynamic range, and more likely only 3) and also the saturation problem. I think drift over temperature and time would make it so this would need pretty frequent adjustments. I could try to null the opamp offset either using the null pins or slightly offsetting the input.
This results in a delayed and slower rise time at the beginning of current pulses in some of the breadboard versions of this I have built. When the input goes positive (eg at the beginning of a short current pulse) the op amp would take time to get out of saturation, and then to slew from the negative rail to a high enough positive voltage to turn the MOSFET on. If the offset happens to be positive, when the input is zero, the opamp would try to drive to a negative sense voltage and its output would hit the negative rail. This is not much of a problem when the current is 1A (and the voltage across the sense resistor is 1V), but becomes a pretty big problem when the current is 1mA and the sense voltage is 1mV. The opamps with the right speed for what I'm looking for have offsets of 100-500uV. The problem is the offset voltage of the opamp which introduces an offset to the output current. I would like the output to be adjustable over a pretty wide dynamic range, let's say 4-6 decades (amperes to microamperes), and most importantly I'd like the output to be a true zero if the input is zero. Simulate this circuit – Schematic created using CircuitLab I have a current source circuit with an op amp and pass transistor.