EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: AndersJ on April 23, 2021, 10:58:40 am
-
I have a integrator as shown in the picture below.
Red curve: 5V digital input.
Green curve: Integrator output.
There is a delay before the integration starts.
The delay is longer for the larger RC.
What causes the delay?
(https://www.eevblog.com/forum/projects/op-amp-integrator-advice-please/?action=dlattach;attach=1214012;image)
-
Under normal conditions delay is caused by the limited bandwidth of the operational amplifier, which is especially problematic in an integrator because the operational amplifier must be compensated for unity gain which limits the bandwidth.
In your example however, there is also delay caused by recovery from saturation.
-
Oops, bad schematic.
The diode is NOT shorted.
Assuming my issue is saturation,
would a 4,5V diode clamp across C reduce/eliminate saturation?
-
Clamping accross the cap could suppress the saturation. However with 5 V supply and the virtual ground at 2.5 V it would need more like a 2.4 V diode or similar.
-
Yes,
virtual ground, of course.
I was thinking GND, when I suggested 4,5V.
I just tried this, with 2 x (yellow LED + 1N4148).
Almost full swing, and no delay.
Will the saturation harm the op-amp,
shortening its life,
or have other other negative effects?
-
Will the saturation harm the op-amp,
shortening its life,
or have other other negative effects?
There is no risk of damage. Saturation creates recovery time and may cause increased supply current which may or may not be acceptable.
-
When is an integrator not an integrator?...
Saturation of the output, means the op-amp can no longer keep up with the input, and the linearity assumption used in the integrator derivation is broken. Input voltages are no longer equal, and that is the hint as to where the extra time is going.
To minimize windup, clamp the inputs together with a pair of diodes. To eliminate it entirely, use a transconductance amp (with sharp, well behaved saturation!) into a capacitor, then a buffer.
Tim