A few Short Questions on HFSS

[enyxep]

deembed hfss

Hi guys, i've been using hfss for almost 2 week now but i still haven't figure out some stuff . so i'm hoping some of u could help me see through the problems :wink:
so here it is:

in defining wave port:
- What does it mean by deembed port ?

in setting up analysis setup:
-what is adaptive pass? is it the number of test iteration?

-what does delta S represents? what would normally the max value?

-what does lambda refinement do? is it the same lambda as in wavelength?

If solution doesn't converged. What does it actually mean?

Last one: How to plot 1-dimension Electric field against freq? e.g. Ex v.s freq in the result report.

that's it.. many thanks guys..

regards.
enyxep[/b]

 

[madengr]

hfss delta s

- What does it mean by deembed port ?
The reference planes are moved to the end of a de-embedding line which you define. You measurements are now relative to this point instead of the ports. You should not de-embed past any discontinuities.

-what is adaptive pass? is it the number of test iteration?
The mesh density is increased (or adapted) then the error in sparameters (Delta-S) between the present pass and the previous pass is computed. The premise is that increasing the mesh density should provide a more accurate answer. If there is not mush change in Delta-S then the solution must me accurate.

-what does delta S represents? what would normally the max value?
Total S-parameter error. Don't know if it's RMS or not but the default value is 0.02 which I think is 2%. Beware that the wave port solution must also be increased if you change Delta-S max or else your re-normalized (to 50 ohm) results will be more error prone.

-If solution doesn't converged. What does it actually mean?
It means that increasing the mesh density did not result in a solution (after N passes) with less than maximum Delta-S error. You need at a minimum two passes in a row with Delta-S less than the maximum error.

Note sure about the rest.

 

[enyxep]

hfss adaptive passes did not converge

thx again madengr,
you made it clear

But anyone can enlight me with the other questions?
cheers..

 

[antennas321]

adaptive pass in hfss

I would recommend getting a copy of some HFSS help material (an Ansoft rep should be able to help with that). That way you would avoid some of the pitfalls of using a complex tool such as HFSS. Setting up the excitations and boundary conditions are very important in insuring that you are accurately modeling a real-world situation.

For example, if you are modeling an antenna, then you may need to seed the mesh on the PMLs in order to obtain accurate far-field parameters. HFSS converges on S-parameters only.

Good luck.

 

[chwf]

Re: hfss delta s

this answer is Very very helpful for me! I can't repay you! But Bundle of THANKS for the answer!

- What does it mean by deembed port ?
The reference planes are moved to the end of a de-embedding line which you define. You measurements are now relative to this point instead of the ports. You should not de-embed past any discontinuities.

-what is adaptive pass? is it the number of test iteration?
The mesh density is increased (or adapted) then the error in sparameters (Delta-S) between the present pass and the previous pass is computed. The premise is that increasing the mesh density should provide a more accurate answer. If there is not mush change in Delta-S then the solution must me accurate.

-what does delta S represents? what would normally the max value?
Total S-parameter error. Don't know if it's RMS or not but the default value is 0.02 which I think is 2%. Beware that the wave port solution must also be increased if you change Delta-S max or else your re-normalized (to 50 ohm) results will be more error prone.

-If solution doesn't converged. What does it actually mean?
It means that increasing the mesh density did not result in a solution (after N passes) with less than maximum Delta-S error. You need at a minimum two passes in a row with Delta-S less than the maximum error.

Note sure about the rest.