Feeding A Square Wave Into An Integrator

Papabravo

Joined Feb 24, 2006
21,225
How about using a voltage source with the PULSE generation parameters.
The source you are using lacks several of the parameters like rise time, fall time, on time, and period. Like this:

1697758258084.png

It should be noted that a 741 does not behave well when operated with a single power supply.
 

Attachments

Last edited:

crutschow

Joined Mar 14, 2008
34,452
Is this triangle wave normal at the peaks?
Ideally no, but for the frequency response of the LM741 it is normal.
Its low frequency response allows feedthrough of the fast rise input change at the square-wave edges directly through the feedback capacitor, which then appear at the output.
The 10ns rise/fall times in you simulation has a very high frequency Fourier component of about 35MHz.
If you look at the spike at the (-) op amp input and compare that with the output spike, you will see they are essentially identical (second attachment).

If you reduce the square-wave rise and fall times to 10µs, which gives the op amp time to respond to the change, my simulation shows no spikes at the triangle peaks (first attachment).
 

Attachments

Last edited:

Thread Starter

elec_newbie

Joined Oct 19, 2023
21
Ideally no, but for the frequency response of the LM741 it is normal.
Its low frequency response allows feedthrough of the fast rise input change at the square-wave edges directly through the feedback capacitor, which then appear at the output.

The 10ns rise/fall times in you simulation have the Fourier component of a very high frequency.
If you look at the spike at the (-) op amp input and compare that with the output spike, you will see they are essentially identical (second attachment).

If you reduce the square-wave rise and fall times to 10µs, which gives the op amp time to respond to the change, my simulation shows no spikes at the triangle peaks (first attachment).
Excellent. TY for your help folks!
 
Top