Post subject: Cascade S parameters Posted: Thu Jul 17, 2008
Joined: Tue Sep 04, 2007 7:54 am
I am trying to analyze the casecade
S parameters of a passive filter cascaded with a LNA. The passive filter
is followed by the LNA. S11 of the passive filter is 22dB and S22 is
around 21dB. LNA S11 is around 18dB. Wht will be the S11 of filter LNA
cascade?. Cascade S11 will vary depending on the phase of filter S22
and LNA S11. If the phase varies from 0 to 360 what will be the maximum
and minimum casecade S11?
Post subject: Posted: Fri Jul 18, 2008 10:46 am
Joined: Tue Feb 27, 2007 6:16 am
the cascaded S parameters of the resultant two port network, it is easiest
to first convert from S parameters to the appropriate Z, Y or T parameters,
work out the resulting matrix and then convert back to S parameters.
For parallel connection of two port networks, convert to Y parameters
and sum the resultant. For series connection of two port networks, convert
to Z parameters and then sum the resultant. For a cascaded connection,
convert to T parameters then multiply the last network by the one previous
and so forth (order is important). Refer to pg 169 Intro to Radio Frequency
Design by Wes Hayward (beware typo pg 172, this one held me up for many
Most books on RF design will include some section on
how to convert between parameters. If you use Matlab I can post some
code I was using to convert between parameters.
Regards, Darcy Randall, Perth, Western Australia
Post subject: Posted: Fri Jul
18, 2008 5:09 pm
Joined: Tue Jul 08, 2008 7:15
Location: Boulder Colorado
If you put a passive
filter in front of an LNA, the S11 of combined network will be dependent
on the match between filter output and the LNA input. I don't think
it is possible to predict composite S11 without full set of S parameters
for each network.
I am not sure what objective is, but if filter
were perfect (S11=S22=0, S21=S12=1) in passband, then S11 of network
would be S11 of LNA, in other words, the S11 of the amp can't be improved
by the filter unless an attempt is made to match the filter to the amp
for optimum return loss in passband.
If you are inclined to
do some math, HP app note #95-1 gives a very good tutorial on S-Parameter
If you are like most of us, and would prefer for computer
to do math, there are a number of free s-parameter design tools available.
I have not used it, but I understand that Ansoft has a "student" version
available. There used to be a program called ARRL designer that was
quite capable of doing simple s-parameter calculations. I would bet
that if you check this (RF Cafe) site you will probably find a program
that is capable of doing this analysis.