hfss cpw size with very thick substrate

[oby]

cpw field distribution hfss

Hi, I am not actually new to HFSS, but I am struggling with a problem.

By designing a cpw transmission line I have problem with the cpw size.

g-s-g= 50 um
h(substrate)= 400 um

following all the tips in the various tutorial the width of my waveport shopuld be at least 200 um...which is fine.....for the height of the port I would have 4*h=1600 um).
The ratio of the two length gives me a very tall port with wrong results in the port mode and impedance.

I don't want to use an internal pumped port because I am investigating the different port impedances I could get varying the ports parameters.

Any help...or any experience regarding to this???

Thanx guys

 

[wlcsp]

Hi, I am not actually new to HFSS, but I am struggling with a problem.

By designing a cpw transmission line I have problem with the cpw size.

g-s-g= 50 um
h(substrate)= 400 um

following all the tips in the various tutorial the width of my waveport shopuld be at least 200 um...which is fine.....for the height of the port I would have 4*h=1600 um).
The ratio of the two length gives me a very tall port with wrong results in the port mode and impedance.

I don't want to use an internal pumped port because I am investigating the different port impedances I could get varying the ports parameters.

Any help...or any experience regarding to this???

Thanx guys

It is only rule of thumb. Try to solve the ports only. Plot the magnitude of the E-field, preferably in log scale. The strongest filed would be around the metal. Then I would re-size my port accordingly to capture enough fringing field until 3 orders of magnitude smaller. Next, I would increase the mesh on the port to get more accurate result in the end.
So, the port does not necessarily have to start from the bottom of the substrate since the field may not even reach the bottom, unless the substrate is grounded.

wlcsp

 

[oby]

Hi wlcsp,
thanx for your advices....

I was sure there was a logic way to set up the port for cpw (and microstrip).

Your advice seems a sensible approach.
Just last question....you suggest to include 3 orders of magnitude for the E field in the port area, by plotting the field.
Is this just a visive approach, or there is a a way to realize we are including the E field til the 3 orders of magnitude??

thanx

 

[nhomuathu_hanoi]

I agree with Wlcsp.
I guess that you used high electric permitivity substrate and the proposed operating frequency is also high.
With a high electric permitivity, most of the CPW field (E & H) will be concentrate around G-S-G surface. In this case, I think you don't need to extend your wave port size like the common case. Try to reduce the port's height until you have the desired port characteristic impedance and a appropriate field distribution
regards!