
1. Re: Normalization/Reference Impedance
cristian.filip Jan 8, 2014 7:44 AM (in response to potluri.ravikumar)Hi Ravi,
Either one will work fine as the simulator will know from the Touchstone file what was the value of the reference impedance. Accordingly to the Touchstone spec 11, pages 4 and 5, the data file must contain an option line formatted as follows:
# <frequency unit> <parameter> <format> R <n>
R n specifies the reference resistance in ohms, where n is a positive number of ohms (the real impedance to which the parameters are normalized). The default reference resistance is 50 ohms.
Here is an example from a PDN Zparameter extracted using HyperLynx:

! HyperLynx V8.164 SISPICE V2010.06.16
# HZ Z RI R 1.000000e001

Cheers,
Cristian

2. Re: Normalization/Reference Impedance
potluri.ravikumar Jan 8, 2014 9:09 AM (in response to cristian.filip)Hello Cristian,
What I am trying to do here is to cross check the DC drop from SParameter simulation data and find out reference impedance. I did DC drop simulation between VRM and Load and simulation suggests DC resistance of the path to be around 10mOhms. Please find the file attached. The S21 magnitude from a series R resistance with Zo reference impedance is 1/(1+R/(2*Zo)). Assuming that the actual correct DC resistance value is 10mOhms, the S21 magnitude is around 0.5dB and with Zo = 0.1Ohm and 3dB with Zo=0.01Ohm. So, is this the way to select reference impedence of a SParameter model of a power plane if we want to do transient analysis to find out power supply noise?
Thanks and Regards,
Ravi

Report_DC_Drop.txt.zip 963 bytes
