# Transmission Line Pi-Model in Spice: Urgent!!!

Discussion in 'General Electronics Chat' started by Management, May 5, 2009.

1. ### Management Thread Starter Active Member

Sep 18, 2007
306
0
Hi All,

What is a good way to look at the performance of a transmission line in Spice?

I have a model exported from Sonnet to PSpice. I would like to see the "real" loss associated with that transmission line pi-model.

For an example I would like to use some random FPGA I/O line. If the Voltage is X then what is Y out.

Should I place a step voltage in and see what the voltage out would be? If it is a, or suppose to be a 50 Ohm Line, should I terminate it with 50 Ohm and see what that output is with that step input? This would also give me the rise time associated with the line as well correct?

Thank you.

2. ### KL7AJ AAC Fanatic!

Nov 4, 2008
2,047
295

I'm not sure how good Spice is at modeling lossy transmission lines, but there are plenty of excellent tools out there for doing so. Are you just trying to calculate loss, or pulse distortion?

An ideal, perfectly terminated transmission line should just look like a resistor with a certain amount of propagation time (which in your application is probably negligible)

Eric

3. ### Management Thread Starter Active Member

Sep 18, 2007
306
0
Ok let me take a step back.

I sometimes get confused between the characteristic impedance of a line (matching right?) and the dc resistive impedance of a line. Any material describing the difference? Can someone do a quick explanation?

I have a 2 single-end striplines (plural) that is 2 inches long. The line thickness on line A is 15 um and the line thickness on line B is 5 um.

Does a thinner strip increase the dc resistance of a line? Why?
What I'm looking at is the GND planes and the thickness of the stripline conductor. Both being equal each other depending on which line (A or B) your looking at. Over the same length, should line B experience great resistive losses?

Also, from my understanding if I use a PEC and define the losses of the dielectrics then, looking at S-parameter S21, any losses I see would be due to the dielectric. Am I correct is this assumption?

And if I use air (dielectric constant of 1) and define the metals conductivity, then any losses I see in S21 should be due to WHAT exactly? How can i find the resistive losses of the line?

4. ### Management Thread Starter Active Member

Sep 18, 2007
306
0
I know this is a lot of questions but I keep confusing myself.

What I want to get is the amount of loss on a transmission line in dB per m or inch (whatever).

Thanks again.

5. ### KL7AJ AAC Fanatic!

Nov 4, 2008
2,047
295
The DC resistance of the stripline should be UNMEASUREABLE. The resistance to ground should be, for all practical purposes INFINITE. The characteristic impedance should be ENTIRELY defined by the transmission line formula, unless you're running SUPER high R.F. frequencies.

Eric