doubts in simulators


Joined Sep 8, 2017
Just a beginner myself, but so far LTSpice seems pretty good. It only runs natively on Windows and MacOS but if you're using a Linux box you can always install the Wine package, which is what I did and it works just fine.

For quick-and-dirty simulations I recommend the online Falstad simulator. It does have its limitations however, so more complex circuits may require breadboard verification.


Joined Mar 14, 2008
Spice base simulators such as LTspice (free) or PSpice (fee) are generally the most accurate for analog circuit simulation.
But no simulator will give results equal to actual hardware since there are parameters, such as power dissipation, and stray parasitics that simulators don't account for (although you can add those into the circuit for a more accurate simulation).
Also part models don't generally include all the various parameters of a real device, such a breakdown voltage.
But use of such a simulator will generally give results close to the operation of the real circuit and can be quite helpful in discovering circuit errors before you build the circuit.

The main caveat in using a simulator is that you have to have a general idea of how the circuit is supposed to operate and thus recognize if the simulator is giving erroneous results that don't correspond to something a real circuit would do.
For example, if it's telling you that the circuit current is a million amps, that's likely in error (unless the real circuit is designed to carry a million amps, of course ;)).
You can't just blindly accept the simulator results.


Joined Aug 21, 2017
About Linux and LT-Spice. At least to my Ubuntu the Wine is not strong enough to dull the processor. For the sure Spice is incapable to work there on Ubuntu. Therefore I use the VirtulaBox with the corresponding real Windows for Spice.


Joined Aug 21, 2017
RE:""The main caveat in using a simulator is that you.....""
No, how far I see, the main problem is very rarely the `mad` results if any You had failed with something serious. But about always we have a problem of correct component models. See, the P-Spice use the model set what contains only electric data like I max, C(gs) etc, and most of producers readily shown `em. But there is large difference in LT-Spice what demands a lot of physical, chemical and geometrical data of components, like charge carrier concentration, barrier layer sizes etc, what most of producers count the sacred secret. That is, by one hand, the answer why LT-Spice gives far more trustable simulation results, but in another hand it means that for many components there is need to apply a P-Spice models into LT-Spice (it allows such manipulation) but then the accuracy is degraded down to the ordinary P-Spice level.
However, to say the truth, by my own revolutionary consciousness, I personally vote for any free of charge software, and that is ultimate reason why "P-Spice must die".

RE: Bordodynov
Hi mate! You live seems only 800 km to south-east from me!