samcheetah said:
ive only used Multisim but its really wayyyyy too expensive. i just use the limited demo version :lol: for most of the cases it serves my needs. but i really do get frustrated by the inability to save my schematics. i have heard about UltiBoard and UltiRoute which work with Multisim but i havent used them.
actually Optikon i cant suggest u any simulator because i dont have the experience with simulators that you have. i want to ask u something. in about every text book there are spice models of circuits written in plain text. ive read a couple of articles about spice and in most of them ive read the spice is better than the breed of WYSIWYG simulation software because of (i) low consumption of resources and (ii) greater options for analysis. but these days the processing power and the memory is not much of an issue. and simulators like multism provide a multitude of analysis tools. so is there any need of knowing spice?????? why build up opamp models when they are already built for you.
i think working with multisim is more "natural" in the sense that you work in multisim like you are working on a breadboard with real components. i think that speeds up the learning process and u dont need to learn spice. what do u guys have to say about this.
Samcheetah,
For most general purpose work with simulators, there is not much need for actually knowing spice these days. I believe back when it was new, software did not exist that would interface nice fancy GUI's and make life easy for the user. So the old-timers had to know spice format & model writing. There are several reasons why one might be interested in building an op amp model.
1) The opamp you want to simulate doesnt exist in model form and the vendor you buy it from cannot/will not create a spice model.
2) The opamp you want to simulate does exist but the characteristics you are interested in model are done poorly or not at all. For example, multi-pole rolloffs. Sometimes (albeit rare) the single pole model doesnt apply or can fool you.
Generally, I will try and use a model that exists or one that is close enough to give me an answer that will resemble my circuit (might not even be the same part number). If I can convince my self that the modelling is good, I will go for it, if I have doubts I will go deeper into the model to find out (hopefully!)
So in summary, you don't need to learn spice if the software guys did an excellent job. With that said, the more you work with electronics and modelling for simulation, the more you will discover its limitations and occasionally have to go into the spice / model aspect of it. Not to mention, having a good working knowledge of how spice works is valuable for those (seemingly more frequent) times when spice won't converge on a solution!