T
The Captain
Guest
Is there any way to use a look up table, for example one derived from
a spreadsheet, to provide a model of a twisted pair in
Scematics/PSpice?
I've tried using least squares regression to provide a formula which
describes the behaviour of the components as their responses change
with frequency, due to skin effect and dielectric variance with
frequency. Unfortunately, the derived formulas provide a straight
line or a simple curve, neither of which actually conform to measured
line behaviour as the frequency changes.
I have a good set of measured parameters available, which cover
frequencies from zero to 2 MHz, in 10KHz steps. What I need to do is
to use these measured parameters in a twisted pair line model, with
the results at each simulated frequency being derived from the
parameters provided by the look up table.
There doesnt seem to be any simple way to do this. Using the usual
RLGC lumped components, without any frequency dependent change in
response, doesn't provide a model with anything like the required
accuracy.
If anyone has any ideas, I'd be grateful.
John
a spreadsheet, to provide a model of a twisted pair in
Scematics/PSpice?
I've tried using least squares regression to provide a formula which
describes the behaviour of the components as their responses change
with frequency, due to skin effect and dielectric variance with
frequency. Unfortunately, the derived formulas provide a straight
line or a simple curve, neither of which actually conform to measured
line behaviour as the frequency changes.
I have a good set of measured parameters available, which cover
frequencies from zero to 2 MHz, in 10KHz steps. What I need to do is
to use these measured parameters in a twisted pair line model, with
the results at each simulated frequency being derived from the
parameters provided by the look up table.
There doesnt seem to be any simple way to do this. Using the usual
RLGC lumped components, without any frequency dependent change in
response, doesn't provide a model with anything like the required
accuracy.
If anyone has any ideas, I'd be grateful.
John