circuit simulators with electrolytic capacitors?

There is such a thing as trying to overdo the simulation. The simulator is to help understand the behaviour of a real circuit, certainly, but at some level of attempting to achieve reality it is not worth the trouble. This requires judgement on the part of the user.

If you want to use electrolytic capacitors in a 555 timer circuit (for example) then they are always bound to be a disappointment, because of temperature effects, dielectric absorption, leakage current and grossly inaccurate capacitance. Electrolytic capacitors are not intended for that purpose (timing), but as reservoirs and bypass capacitors. A designer needs to understand this and find another way. So use a standard capacitor in the simulation model and that will be fine, so long as you understand it will have practical limitations.

In the case of LTspice (and probably others), power sources and standard components including capacitors can have parasitic values set as properties of the component, such as series and shunt R of a capacitor. There are even included models for some brand name aluminium electrolytics that set the values automatically and change the symbol to polarised. I don't think you can expect more than this.

Here are some points:

Simulators don't usually show a special alarm or warning for cases of the ratings exceeded, such as when a device is turned to smoke or a million volts is present or a million amps is drawn. A certain level of user understanding is expected. The original purpose of spice simulators was for very experienced designers to develop integrated circuits more realistically. You should check for these "ratings" situations yourself. It requires judgement, something not normally included in software. The same idea applies to an electrolytic capacitor. It is up to you to make sure reverse bias conditions do not occur. You cannot expect this to be modelled, given there is no standard or simply modelled behaviour even. Why doesn't someone add "ratings" to their simulator? It would slow the simulation down a lot I expect. But more importantly, how do you determine a transistor is getting destructive voltages or power, when this depends on airflow, heat sink, etc? I think best to see it is beyond the normal scope of a spice simulator. I do think the type of user is changing, with beginners using simulation for learning. Consider simulation is a tool, very useful, but there is no substitute for actually building the circuit. This will include all the aspects of stray coupling, power supply deficiencies, interference etc.

The equivalent series resistance (ESR) is easily modelled even if these are not included in the capacitor model, by adding that externally. Dielectric absorption is added by a parallel network of R and C in series. Leakage by a parallel resistor, and so on. Yes you have to guess or measure these.

If you do model ESR (which can be important) you really need to model the supply (source impedance), the wires and other stray or parasitic properties. For this reason it is my advice to only simulate specific points of interest, rather than the whole detailed circuit.

http://www.electro-tech-online.com/general-electro...

http://www.newark.com/pdfs/techarticles/cornell/im...