# "Eigen Q" in HFSS

Status
Not open for further replies.

#### zhul3

##### Member level 5
hfss how obtain q resonator

Hi guys,

I am using HFSS to design a 4-pole coaxial cavity filter.

Now I am in the last step designing the feedline of the filter. (please see the pic.) I set a PML boundary on the coaxial line and sweep its "height" to achieve a desirable Loaded Q.

Equations

QL=sqrt(im_f^2+re_f^2)/(2*im_f),
im_f=(im(Mode(1))+im(Mode(2)))/2
re_f=(re(Mode(1))+re(Mode(2)))/2

are used in the solver for calculating loaded Q. (The parametric sweeping (10 passes) takes super long time with the PML boundary.....)

Surprisingly, I noticed that after setting a PML boundary, a "eigen Q" option is appearing in the solver setup window and gave Q values of each mode after simulation.

My question is, what do these "Eigen Qs" mean? Unloaded Q or Loaded Q?

Can I just set one mode and simulate its Eigen Q, then consider such Q as loaded Q instead of calculating it with the above equations?

Thanks very much.

Thanks

free space impedance 377 ohms

zhul3 said:
I set a PML boundary on the coaxial line and sweep its "height" to achieve a desirable Loaded Q.
You don't need to set a PML boundary for coaxial. The easy way is to define
Impedanace Boundary (Resistance: 377 Ohm, Reactance : 0)

My question is, what do these "Eigen Qs" mean? Unloaded Q or Loaded Q?
It depends what are you calculating. If you set up both material conductivity and
coaxial load then HFSS will give you Q total (1/Q_internal + 1/Q_external = 1/Q_loaded)
If you define coaxial load only (and PEC material) then you will have pure Q_external.
And if you set up short (PEC) to your coaxial output and real material for a cavity then you will get pure Q_internal (aka Q_unloaded).

377ohm impedance

[/quote]
You don't need to set a PML boundary for coaxial. The easy way is to define
Impedanace Boundary (Resistance: 377 Ohm, Reactance : 0)[/quote]

I am calculating the loaded Q with a matching load, that is why I am using PML (perfect matching load). How can you be certain 377ohm (free space impedance) is matched perfectly with the resonator? otherwise the Q calculated is not the desired one.

[/quote]
It depends what are you calculating. If you set up both material conductivity and
coaxial load then HFSS will give you Q total (1/Q_internal + 1/Q_external = 1/Q_loaded)
If you define coaxial load only (and PEC material) then you will have pure Q_external.
And if you set up short (PEC) to your coaxial output and real material for a cavity then you will get pure Q_internal (aka Q_unloaded).[/quote]

I don't really understand how to "set up" such issues like conductivity, load, short to your output.... Would you please clarify them with more details.

Thanks very much. Really wants to be clear of these questions.

zhul3 said:
How can you be certain 377ohm (free space impedance) is matched perfectly with the resonator?
Because main mode in coaxial is TEM-mode exactly the same as in a free space.
So, TEM-wave has 377 Ohm impeadance and if you set up it as a boundary in coaxial, it would be perfectly matched until there are no High Order Modes exited
(just make you coaxial long enough).

I don't really understand how to "set up" such issues like conductivity, load, short to your output....
1. Conductivity :
Select surface/Boundaries/Assign/Finite Conductivity
3. Short (set HFSS by default)
Select surface/Boundaries/Assign/Perfect E

### zhul3

Points: 2
eigen q in hfss

navuho said:
zhul3 said:
How can you be certain 377ohm (free space impedance) is matched perfectly with the resonator?
Because main mode in coaxial is TEM-mode exactly the same as in a free space.
So, TEM-wave has 377 Ohm impeadance and if you set up it as a boundary in coaxial, it would be perfectly matched until there are no High Order Modes exited
(just make you coaxial long enough).

I don't really understand how to "set up" such issues like conductivity, load, short to your output....
1. Conductivity :
Select surface/Boundaries/Assign/Finite Conductivity
3. Short (set HFSS by default)
Select surface/Boundaries/Assign/Perfect E

Thank you very much, navuho.

Status
Not open for further replies.