Surprising blocking capacitor behavior

How to use openEMS. Discussion on examples, tutorials etc

Moderator: thorsten

Post Reply
huszaghmatt
Posts: 24
Joined: Fri 19 Jul 2019, 00:22

Surprising blocking capacitor behavior

Post by huszaghmatt » Tue 31 Mar 2020, 02:33

I'm attempting to simulate a microstrip line with a series blocking capacitor (LumpedElement) and am getting strange results. I've attached the csx structure and efield time dump to demonstrate what's happening. In particular, after the signal excitation, the capacitor seems to act like an AC voltage source, creating an energy oscillation in the system whose average energy decreases very slowly. The simulation takes hours to get to -50dB whereas the same simulation with a series resistor takes minutes. I've only shown the first few oscillations in the video here to keep the file small. The signal excitation is from DC to 18GHz, but I've also tried with other frequency ranges and gotten the same results. I understand that the capacitor will get charged up somewhat, but the sub-GHz energy should be fairly small since it's a gaussian pulse. Additionally, when I set the minimum frequency to 4GHz, for instance, I get a similar result. Here's a plot of the return and insertion losses (I aborted the simulation after 2hrs where the energy was oscillating around -30dB). Black line is s11 and blue is s21.
scratch.png
scratch.png (16.49 KiB) Viewed 1441 times
Any thoughts on what might be going wrong here? Have I used the lumped capacitor correctly? Let me know if there's any other detail I can provide that would help.

blocking_cap.gif
blocking_cap.gif (590.27 KiB) Viewed 1441 times
Attachments
csx.zip
(3.03 KiB) Downloaded 87 times

huszaghmatt
Posts: 24
Joined: Fri 19 Jul 2019, 00:22

Re: Surprising blocking capacitor behavior

Post by huszaghmatt » Sun 05 Apr 2020, 06:53

I discovered that if I removed the PEC caps from the lumped element the behavior is as expected. I'm a bit confused, however, why the end caps do this.

Post Reply