I am checking some Spice models of transmission lines.

I have a Belden 1583A Ethernet Cable, with the following parameters \$R=93.8 \ \Omega\text{/km}\$ and \$Z_0=100 \Omega\$, hence \$C=47.652 \ \text{pF/km}\$ and \$L=0.47652 \ \text{uH/km}\$, for \$D=1 \ \text{km}\$ (check).

I took an approximation with \$n=8\$ discrete modules of RLC circuits each with one series R + one series L + one parallel C with the above given parameters, divided by \$n\$, as depicted.

Lumped Parameters

I considered a sinusoidal source with \$v=1\ \text{Vp}\$, (peak to zero amplitude) \$f=100\ \text{MHz}\$.

I took an Inverse Hybrid G Parameters 2-port Matrix, with the following code in Matlab.

%% Belden 1583A Cat5e
% V: 1V,100MHz
clear all; clc;

VF=0.7; Z0=100;
C=1/c/VF/Z0; L=Z0/c/VF; R=93.8;

% Distributed Impedances
v1=1; f=1e8; k=8;    
ZC=1/(j*2*pi*f*C/k); ZL=j*2*pi*f*L/k; ZR=R/k;
% B: ABCD' Form / Two Port Network Matrix
% V1I1 -> V2I2
B=([1 -ZL-ZR;0 1]*[1 0;-1/ZC 1])^k;
% G=inv(H): Inverse Hybrid Form
% G: V1I2 -> I1V2
G=1/B(2,2)*[-B(2,1) -1; det(B) -B(1,2)];
% Open Loop
i1=abs(x2(1)); v2=abs(x2(2))

This brings an output of \$v_2=0.924\ \text{Vp}\$, phasor magnitude.

However the same circuit simulated in two different programs, with zero initial conditions, and for any convergence settings, brings \$v_2=0.873\ \text{Vp}\$, sinusoidal steady state. Too much difference for an accuracy setting.

Why could I be having this difference?

  • 1
    \$\begingroup\$ Are you modeling just a meter of the cable? or an entire kilometer? \$\endgroup\$ – analogsystemsrf Feb 9 at 8:47
  • \$\begingroup\$ The purpose is to have a handy model for the kilometer range, now for digital, later for analog. I am comparing some built in models and I wished to refresh an old 2-port parametrization, but I was unable to equal them... \$\endgroup\$ – Brethlosze Feb 9 at 8:49
  • 1
    \$\begingroup\$ Is there a reason you are modeling driving this with an unmatched (0-impedance) source, and with an open-circuit termination? What other simulators did you try, and did you use the same terminations for those simulations? \$\endgroup\$ – The Photon Feb 9 at 15:56
  • 1
    \$\begingroup\$ Also, is there a reason for making a unbalanced model of a balanced transmission line? The Belden spec of 94 ohms/km appears to be per conductor, but in a differential configuration you need to account for both signal and return currents passing through this resistance. \$\endgroup\$ – The Photon Feb 9 at 16:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.