Hi All, > Al Davis wrote: > It works for me. Even that LM358 model you supply works for > me, after a simple change. That model looks strange to me. > The real netlist would be simpler. It is amazing how much work > some people will do to hide the real circuit.
A schematic would be even better. I've toyed with the idea of reverse engineering it but at this stage have more pressing tasks requiring my attention. If anyone else wanted to do it that would be great. > In the old days, all Spice files reserved the first line for a > comment, so the first line is ignored. Gnucap honors this, and > drops the first line. For your LM358 model, I had to add a > blank first line. If this isn't the way it is supposed to be, > let me know. It is easy to change. All I need to do is take > out that line of code that skips the first line. The simulation files generated by GSpiceUI always begin with a banner so it's OK if GNU-Cap ignores the first line. > You may have forgot the "op" command. You must do an "op" > analysis before doing an "ac" analysis, unless you want to > simulate it with the power off. Actually, you can do anything > first and Gnucap will use whatever it was at last as the > operating point. This behavior is different from Spice, but is > of great value in simulating nonlinear circuits in AC. You're correct, I did forget the "op" command in the AC analysis. I modified the code last night and it has made a dramatic difference. I'm doing some testing and adding opamp models to the library at the moment. The mod. will be in the next release. If anyone want's it now let me know and I'll send it to them. Thank's for your comments Al. Mike W.
