Questionable results with simulation software

[Bourbon daddy]

I am doing an assignment for college, (though my question is more general, rather than based on questions I have been set, hence the reason this post not being in the homework forum).

It is basic DC circuit analysis using, Kirchhoff, Thevenin and Norton theorems. Nothing taxing, so I know my calculations are correct. Also, using different methods on each circuit is giving me the same results,confirming my findings.

However, to back it up, I have also created the circuits using Yenka simulation software and I find that the results differ. Not by much, generally less than 5% of the expected value, but occasionally as much as 8.5%. I am not too concerned, but does anyone know why there is a slight anomaly between the results. Am I just expecting too much from the software or is it something more than this. For instance, the calculations being more of a theoretical estimation and the simulation software being closer to what the readings would be 'in the real world'.

Thanks guys

 

[milesyoung]

... For instance, the calculations being more of a theoretical estimation and the simulation software being closer to what the readings would be 'in the real world' ...

All the software does is find a numerical solution to a problem you already have an analytical solution for. If there's that large of a difference in results, then it's either a terrible solver for your problem (which seems odd for the ideal, linear networks you're probably working on) or there's other elements to the circuit used in the simulation which you aren't including when solving it analytically.

Have you tried using LTspice or some other SPICE simulator to compare results?

 

[Bourbon daddy]

I haven't tried LT spice, I will download it and give it a go. Yenka is a very basic software, similar to crocodile technologies, that I was introduced to when designing circuits with logic gates. It is not really up to much, I usually just use it to illustrate the circuits that I am solving.

Like I said, the circuits that I am calculating are all very basic: 2 voltage sources, 3 resistors max. So i could understand why the results are so inaccurate.

Thanks for the advice though, being fairly new to this, I am always on the look out for good software I can use.

 

[milesyoung]

I'm not familiar with the software from Yenka - the default settings for its solver might be too lax (as in _a lot_), but I'm just speculating. If you have the same problem with LTspice then I'd very much like to see your analytical solution vs. the simulation results.

 

[Bourbon daddy]

of course. I have just downloaded LT spice IV. I will watch a few tutorials and get to grips with it, but it will probably be later tonight, or maybe even tomorrow that I get to complete the questions that I am currently doing.

 

[Jony130]

Show as a circuit diagram you use in Yenka. Is impossible to get wrong result in DC analysis in simulation software.

 

[Bourbon daddy]

I have just set up a parallel dc circuit consisting of two power supplies, 108V and 84V, and three resistors, 6, 99 and 30 ohms. Measuring the voltage and current of the various components using LT spice I am getting exactly the results I was expecting. I guess it is time to stop using Yenka for network analysis. Thanks for the advice Miles young.

 

[Jony130]

Well very strange, because I get exactly the same result in both programs.
Maybe you have "ON" internal resistance in Yenka.

 

[AlephZero]

Is impossible to get wrong result in DC analysis in simulation software.

Ahh ... the innocence of youth ...

But if you are getting results 5% wrong for a circuit that simple, either the software is complete junk, or (more likely) the simulation you ran was not the same as the one you wanted to run.

 

[jasonRF]

Ahh ... the innocence of youth ...

But if you are getting results 5% wrong for a circuit that simple, either the software is complete junk, or (more likely) the simulation you ran was not the same as the one you wanted to run.

Yep ... Perhaps the simulation has an internal resistance in the source, or some other aspect that it is modeling that does not match what the OP thinks they are simulating. I am not familiar with that software, but the OP should "open up" the elements and see what parameters they have ... it is likely not doing what they think it is. It is possible the software reallly is that bad - but I would be surprised.

jason