spectreRF measurement question

[cupric]

Hi!

I was simulating an LNA using spectreRF. In the spectreRF reference
manual, there are two ways suggested to measure the voltage gain of
the circuit. One through the PSS analysis (and plotting voltage gain
at the input frequency and its harmonics) and the other by performing
an s-parameter analysis and plotting s21. Both of these measurements
give different results for the same circuit. Any suggestions on why
the difference in results.

Also, I get different results when I try to measure the noise figure
of the same circuit by (1) using s-parameter simulations and (2) using
Pnoise analysis.

thanks.

 

[S.P.]

i) SpectreRF in PSP analysis has many different gains, make sure that
you compare the same that you have in PSS.
ii) PSP/PAC/PXF will give the gain under small signal approximations
while PSS will have a no-assumptions, regular gain.

S.P.

cupric wrote:

Hi!

I was simulating an LNA using spectreRF. In the spectreRF reference
manual, there are two ways suggested to measure the voltage gain of
the circuit. One through the PSS analysis (and plotting voltage gain
at the input frequency and its harmonics) and the other by performing
an s-parameter analysis and plotting s21. Both of these measurements
give different results for the same circuit. Any suggestions on why
the difference in results.

Also, I get different results when I try to measure the noise figure
of the same circuit by (1) using s-parameter simulations and (2) using
Pnoise analysis.

thanks.

 

[Andrew Beckett]

s21 is essentially a power gain, from the port 1 to port 2. Voltage Gain
in PSS is that, a voltage gain, from a node to a node. Consequently you'll
get different answers if the inputs/outputs aren't matched.

If you're doing straight s-parameter simulation (i.e. not psp or qpsp) and
compare that with the noise from pnoise, then it won't be including any
noise folding, so I wouldn't expect the answer to be the same necessarily.

However, without knowing precisely what you've done, it's hard to
give an exact answer - there may be other explanations. I've seen many
cases where the results are in agreement (but it has to be an appropriate
circuit).

Regards,

Andrew.

On 7 Feb 2004 19:48:37 -0800, cupricwhistle@yahoo.com (cupric) wrote:

Hi!

I was simulating an LNA using spectreRF. In the spectreRF reference
manual, there are two ways suggested to measure the voltage gain of
the circuit. One through the PSS analysis (and plotting voltage gain
at the input frequency and its harmonics) and the other by performing
an s-parameter analysis and plotting s21. Both of these measurements
give different results for the same circuit. Any suggestions on why
the difference in results.

Also, I get different results when I try to measure the noise figure
of the same circuit by (1) using s-parameter simulations and (2) using
Pnoise analysis.

thanks.

--
Andrew Beckett
Senior Technical Leader
Custom IC Solutions
Cadence Design Systems Ltd