Romano Giannetti

Wow, great! How did you compiled it on 20.04? Did you add the Qt4 source? (I am interested for other apps too...)

Maybe you can have a look at my take at an AppImage: https://rmano.github.io/qucsAppImagesBuild/ The authors are working quite hard lately toward the QT5 migration here: https://github.com/Qucs/qucs , but for the moment you can use the appimage as a stopgap.

New dcraw sources

This is expected; it is a classical visual aliasing problem. To correctly sampling a 100Hz signal you need at least 200 sample per second, that is, one point every 5ms... But better more. You are plotting 100 samples over 1000ms, including extremes, that is, one sample every a bit less than 10ms... One point per period of your signal. If you use 101 points you will have even more surprising output (the graph should be a constant zero value, but you'll see math rounding errors). Use 1000 points and...

This is expected; it is a classical visual aliasing problem. To correctly sampling a 100Hz signal you need at least 200 sample per second, that is, one point every 5ms... But better more. You are plotting 100 samples over 1000ms, including extremes, that is, one sample every a bit less than 10ms... One point per period of your signal. If you use 101 points you will have even more surprising output (the graph should be a constant zero value, but you'll see math rounding errors). Use 1000 points and...

This is expected; it is a classical visual aliasing problem. To correctly sampling a 100Hz signal you need at least 200 sample per second, that is, one point every 5ms... But better more. You are plotting 100 samples over 1000ms, including extremes, that is, one sample every a bit less than 10ms... One point per period of your signal. If you use 101 points you will have even more surprising output (the graph should be a constant zero value, but you'll see math rounding errors). Use 1000 points and...

Yep. Look what happens if you use 101 points:

Here is the sim with the ideal op-amp