I'm doing this for the learning as well. I think that this tool will do a reasonable job of discrete designs.
Some of these designs, like the Fuzz Face, or SBD Fuzz, which run the input signal through a cap directly to the base of a transistor whose emitter is grounded, are going to be VERY dependent on the characteristics of that transistor, as they are working without any feedback to even things out (such as a resistor from emitter to ground). These types of circuits are probably going to sound different depending on which actual transistor is used, even if it has the exact same part number as another one.
In any case, I find it useful to say "what is the effect of changing the emitter resistor"? etc. etc. even if the simulation is not 100% accurate.
For op-amps, if you are running them without slamming the outputs into the rails, I would expect reasonable results from even a generic model. Once you go into extreme overload (which seems like it happens ALL THE TIME around here) then the actual behavior is going to be more complex than a simple model.
I have an old Radio Shack tube amp that has a tube complement something like a Tweed Deluxe. But I don't just want to rewire it to be a Tweed Deluxe, so I'm going to try to do something a little unique or off the wall, (or just try different tone stacks prior to committing it to solder) and then see how close the simulation came to the real result.
|