cst discrete waveguide port

Hello, I am trying to do TDR on a Twisted Shielded Pair(TSP). I had some decent results when I used arbitrary wire dimensions (that is, they matched experiment qualitatively). Now that I have entered the actual wire dimensions, the results are very different (Conductor Radius = 0.25 --> 0.6mm, Dielectric Radius = 1 --> 0.85mm). My question concerns the choice of excitation port. I have been using a discrete port, which is essentially a point source and propagates many modes simultaneously. In the experiment, the TSP is coupled to the TDR using a coax connector which propagates TEM only (or mostly). Does it not make more sense to use a waveguide port? If so, how could I implement this in CST? I need one wire to be excited with a ramp/step voltage and the shield to act as ground. Any thoughts on how I might do this with a waveguide port?

Thanks,
Nate

Look into using a Multi-Pin Port. That will give you the ability to define unique signals in a plane consisting of several (or many) conductors. That way, you can apply any signal you want to a particular propagating mode in your combination of conductors. This actually runs inside of a field port.

Good luck,

--Max

Thanks Max. Yeah, we ended up using a multipin port single-endedly to excite only one pin while leaving the other open.

Nate

For cables simulations I would recommend to use MTL-based methods, not FDTD.

Could you please provide problem specifiation detailed - cable topology, length, phisical parameters? Or attach your project?

Hi e_m_c. Basically, we're using MWS CST 2006B to simulate a twisted pair wire typical of those used in aircraft wiring. The dimensions are as follows: 0.25mm shield thickness, 0.6mm inner conductor radius, 0.25 dielectric thickness of permittivity 2.7.

Since we are interested in finding shielding faults using TDR, FDTD seems to be the best option at this point. I am trying to correlate the simulations with the lab data, but this is turning out to be a difficult task. I probably shouldn't post the project since I don't have permission from the project lead, but I can answer any questions.

BTW, anyone have any advice on why I keep getting the error "Port 1 is inhomogeneously filled within the first 3 cells"? I have simply cut a crossection of the wire (through shield, dielectric, and inner conductors) to make a planar surface for a multipin Waveguide port.

Thanks,
Nate

e_m_c: Do you mean "TLM" instead of MTL? Yes, TLM techniques can model cable harnesses pretty well. I think the Micro-Stripes software (a TLM approach) had a model for multi-conductor wiring harnesses. I don't know how much control you have over wire placement within the bundle, though.

I think Micro-Stripes was recently acquired by CST from Flomerics. So the Micro-Stripes TLM solver is still available, but you have to get it from CST.

--Max

I?m new user of CST MWS, and I?m interest to simulate the response of a TWP,
Is it possible someone can show me how I can create the twisted geometry?
Thanks in advance.
Smk