Am Mi, 14.03.2007, 08:34, schrieb walter steffè:
> Hello Stefan
Hi Walter,
>> Have you already looked at the QucsTranscalc tool? There is synthesis
>> and analysis for rectangular waveguide available. It also tells if
>> there
>> occur beside the TEM mode any TE or TM modes.
>
> Not yet. I will look at it
Ok, good luck.
>> When creating a waveguide component we can probably easily make a good
>> model for the TEM mode. When higher order TE or TM modes occur it gets
>> perhaps difficult because each mode has different Zl and gamma.
>
>> > The only problem I see is that the S parameters I have spoken about
>> are
>> > those
>> > that are normalized versus the characteristic impedance. This can also
>> be
>> > written as: Zc_TE=beta/beta_wg, Zc_TM=beta_wg/beta.
>> > The rules for the computation of the S matrix normalized versus Zc are
>> > similar
>> > to those explained in your paper but you must take care that Zc is a
>> > function
>> > of frequency and that it can be different on the different ports.
>>
>> I don't yet understand it fully... :(
>>
>> But the Sparameter simulation results are normalized to the impedance
>> given in the Sparameter port (e.g. 50 Ohm). When Zl is different you
>> get appropriate reflections, etc., also depending on the frequency.
>
> To better clarify what I am asking for look on the annexed schematic.
> It is a very simple two port circuit.
> First thing I am not able to specify the impedance (Zp=50 ohm or anything
> else) associated with the ports. I think there is the need of an
> additional
> type of port (waveguide port) which gives this possibility.
The Sparameter port is different from what you used (you used a succircuit
port). You need the "power source" from components/sources. In the
properties there is a Z to define.
When you press F2 (simulate) then you will get the Sparameters of
the circuit normalized to the ones which you specified.
BTW: When you press in your schematic F2, then the simulator complains
about a missing "Pac" which is the power source I mentioned.
> When this is
> done
> a way to proceed is the following:
>
> 1) Compute the admittance Y (which is a 2x2 matrix) of the 2 port circuit.
When you have the simulated Sparameters, then you can add an equation
on the schematic:
Y=stoy(S)
resimulate and then you have the admittance parameters.
> 2) Renormalize Y with respect to the ports characteristic impedance Zp
>
> y = sqrt(Zp) Y sqrt(Zp)
>
> here Zp and sqrt(Zp) are diagonal matrices
> the diag entries of sqrt(Zp) are the square root of Zp entries.
> NB the y matrix is dimensionless
You don't need to renormalize, since Sparameter simulation can do
it one its own. If, whyever you just use e.g. the default 50 Ohm S
parameters, you can renormalize then using an equation:
S_new=stos(S,100,50)
which renormalizes from 50 to 100 Ohms.
> 3) Compute the scattering matrix S as
> S = (Iy)/(I+y) I is the identical 2x2 matrix
This won't be necessary...
Hope this helps, Stefan.
