Hello,
I'm designing a CPWG in the 30 GHz range in HFSS. I noticed that in my design the characteristic impedance is variable across my simulation range from 15-50 GHz. I think this is to be expected but more importantly the impedance of the line seems to be quite sensitive to the port definition. If I use the suggested wave port width the port overlaps with the coplanar grounds and the impedance drops from 52 ohm to 14 ohm. Any suggestions or tips on how to accurately do this and ensure no artifacts from the wave-port definition?
The wave port should overlap with the coplanar grounds so I would trust the simulation results from that. As for the width it should be so it spans a bit on either side of the grounds (I am not sure what suggested width you're using). You can trust the simulation results if increasing the wave port width does not affect the results by much. You can put even more trust in the simulations if you use a model of the connector with the CPWG.
In CST it is possible to look at the port mode fields that have been calculated. It probably is in other codes. Keep making the port bigger until the fields stop spreading wider when you make the port bigger. Also keep making it bigger until the port impedance stops changing.
Mohamed K. Emara Malcolm White Thank you for the response! I think the port is behaving as expected now! But I did notice the s11 plot looks almost sawtooth like. The RL is below -20 dB so i'm not sure if that's an issue but I was expecting a smooth curve with resonance points.
It might be resonance between the two ends of the line. Is the sawtooth regular and is the period (frequency difference between peaks) the same as the frequency that has half a wavelength between the ports?
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators