Hi all,
I am using the circuit physics in comsol 5.1 as part of a complicated model, but to double check I understand what the circuit physics is doing, I decided to simulate a very simple RLC circuit, just to see that comsol gives the values I'd expect.
I followed a RLC problems class I found online, which gave a L, C, and voltage max + voltage source frequency and then compared my graph to the graph they plotted. Please see attached graphs (Ignore the key on my graph, only the first three terms were plotted).
Is there any obvious reason for the differences? The peaks are about the same height for each respective resistance but the peak widths are different compared to the problems classes. I have checked other examples online and they all show that the smaller peaks do not cross over the larger peaks.
My set up is very simple, I have the Ground node, capacitor, inductor, resistor, voltage source in series and I have run a parametric sweep over the 3 values of r (5, 10, 20 ohms), and my frequency domain study goes from (0,1,3000Hz) for each value of r.
Thanks in advance.
Hello Frazer,
You are right, the three curves should not cross one another. The only way I can see this happening is if the curves are displaced vertically, i.e., are you sure that the currents, the y-axis quantities, are not offset from one another? Perhaps the software is outputting the relative currents, i.e., relative to the maximum current for each resistor value, instead of the absolute values of the currents. I know nothing about Comsol; I am just taking an uneducated guess from the symptoms.
Regards,
Tom Cuff
I ran your circuit through SPICE, and got the traces shown below.
The curves don't intersect, because the resistor reduces the 'Q' of the circuit, which reduces the overall amplitude of the current at resonance. I assume that the thing that plots your curves arbitrarily set the x/y scale factors.
Hello!
The reason is probably becouse both graphs are ploted in fact for different quantities: right one is absolute current vs frequency (ie. RMS value - phase is missed). Left one is a real part of the current (current in phase to the voltage only). See atached graphs: top is a RMS current - curves do not crosses each other, below for the same circuit current curves are crossing each other.
Best regards!
Hello Frazer,
You now have three possible explanations for your observation about the curves crossing. Have you been able to determine which, if any, of these possible explanations is correct? I think you owe us a reply.
Regards,
Tom Cuff
Hi, sorry for the late reply, was away on annual leave for two weeks. thank you all for your replies, I don't usually get much of a response off of researchgate!
Thomas - I don't think COMSOL has moved the axes arbitrarily as I am superimposing the curves for each circuit onto the same axes and each curve is plotted with the same expression over the frequency range (with a different resistance value).
Mark - I agree the graphs should look like that, I ran it through LTSpice and got a similar graph (see attached).
Grzergorz - That actually could be it... COMSOL likes to hide whether or not it's dealing with purely real or complex numbers. I think it might have plotted just the real part of the complex current.
In summary, I'm going to use LTSpice as an intermediate step to deal with the circuit based physics.
Thank you all.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs