I am trying simple integrator circuit with cadence using op-amp from its generic analog library (ahdl lib). This simple design works fine with sine input that I get nice cosine output.
But feeding it with a pulse source doesn't result in an expected triangular wave.
Does anyone see anything wrong. The circuit and the wave scan is appended.
Later I will implement SC integrator.
The circuit implements a low-pass rather than an integrator, it won't generate a triangular wave anyway.
But you should show the input waveform for clarity. Also it's not clear, if the OP is suited for single supply operation.
I am trying simple integrator circuit with cadence using op-amp from its generic analog library (ahdl lib). This simple design works fine with sine input that I get nice cosine output.
But feeding it with a pulse source doesn't result in an expected triangular wave.
Does anyone see anything wrong. The circuit and the wave scan is appended.
Later I will implement SC integrator.
You cannot implement an analog "ideal" integrator. This would imply an amplifier with really infinite gain. Thus, you only can choose an active lowpass with a corner frequency (pole) as low as possible. However, this approach is limited by offset properties of the opamp.
Compromize: Lowpass with a pole frequency at least 50...100 times lower than the wanted integration time constant (which is identical to the invers of the angular frequency which leads to a gain of unity.)
Added after 18 minutes:
But be careful: The dc gain of the opamp (with feedback) shouldn't be above app. 1E3 (because of offset).
Hint 1: Usage of an OTA wouldn't help too much, since it is not an ideal current source (this is a common misunderstanding).
Hint 2: In case you are going to use the integrator within an overall dc stabilizing feedback loop, no resistor across the integration cap is necessary!.