Ansoft Designer, Microwave Office, or ALPAC - RF Cafe Forums
Post subject: Ansoft Designer, Microwave office, or ALPAC
Posted: Wed Oct 19, 2005 4:24 pm
I want to get the s parameter
of the transistor.
If I have the spice model of the transistor,
am I able to get the S parameter of the transistor using any of
S-parameters from SPICE modelPosted: Wed Oct 19, 2005 10:35 pm
Actually, all you need is SPICE. Set up the appropriate measurement
circuit, simulate, and you've got it.
Post subject: Posted: Thu Oct 20, 2005
Joined: Thu Oct 06, 2005 6:34
A very good example available in the installation
folder of microwave Office....
Features\NEW--Models--NEW\New Nonlinear Models.emp
Post subject: Re: S-parameters from SPICE modelPosted:
Fri Oct 21, 2005 3:06 pm
Actually, all you need
is SPICE. Set up the appropriate measurement circuit, simulate,
and you've got it.
But I won't get
s-parameters out of pspice,
subject: SPICEPosted: Fri Oct 21, 2005 8:27 pm
are essentially nothing but reflection coefficients: ratios of voltages
and currents. You can get these out of SPICE, you do have to do
some work to set up the necessary measurement scaffolding.
Hmm - this may be an area worth writing an article about...
Post subject: Posted:
Sat Oct 22, 2005 12:43 am
Oct 06, 2005 6:34 am
Try this in Microwave Office,
1. Build the schematic using spice model &
setup input & output biasing options...
2. Simulate the
3. export the s-parameters from output files section...
If you have problems send me the spice model with biasing values..I
will send you s-parameters..
Post subject: Posted: Sun Oct 23, 2005 6:48 am
Joined: Fri Sep 02, 2005 7:25 pm
Location: Hampshire UK
The guest is right in that in principle
once you have a SPICE model, you can contrive the s-parameters,
provided the SPICE model is good enough. That is, you end up with
something that agrees with measured s-parameters. Working from s-parameter
end has huge value in designing matching arrangements for best noise
or IIP3, at the same time keeping an eye on stability. Then you
can get back to the non-linear simulations / bias / and layout effects.
This is OK, but having a "one-click" approach to "export" implies
having a tool like Microwave Office. This cannot can be taken for
granted. Its a bit of a journey to set up a series of simulated
s-parameter measurements using contrived SPICE circuits, and then
take the leap of faith that they agree with a measured device.
Do they really? So what of the data .s2p files we get? Are they
churned out by a manufacturer who had a SPICE model, and then made
a simulator do "export"? I think they measure s-parameters first!
For passive networks it might be OK, but I would be cautious of
s-parameters derived in this way for any semiconductor device, coupler,
magnetic core device, etc.