Calculating Connector Impedance & Irregular Wire Char. Impedance

[roTTer]

Hi,

I am modeling the Unshielded Twisted Pair Wire and I have hit a roadblock there. I have studied the mathematical equations for the twisted wire pair. And they give results with an accuracy of 10%.

This is in part due to the fact, that almost all equations assume that both the wires are tightly wounded and there is no space between the two twisted pairs. But in reality, if the twist/rate is low, the wires won't be wounded in a tight formation, and that results in incorrect results from the calculations.

I have char. impedance and delay(ns/m) values measured using TDR of 4 sample wires.

I understand that having a field solver to do this would simplify the calculations very easily. But is there any article/book that explains how to do it. If there is an evaluation version of a solver that would allow me to do it, that would be swell.

If you have any idea on how to go about doing it manually/excel/code, please do explain or link me to it. Google search didn't give any results.

---------------------------

I am working on calculating Connector impedance of cables. They will be used for freq. upto 100Mhz.

The connectors are not for high-speed PCBs.

I want to model something that looks like...in the attachment.

I do not have any Field Solvers to aid me in this. Is there any material/textbook or site that could help me out with this? There are too many unknown parameters for the design of this. Though I would just like to try.

Are there any software's that are good for these type of calculations. Something that offers an evaluation version would be really good! Or a way to calculate it without them would be good too!

------------

 

[the_emi_guy]

Is this part of a cable or a single pair?

If it is a single pair with no outer sleeve, like you would pull off of a wire spool, and the wires to wire spacing is not uniform, then the impedance will not be uniform.

 

[Baluncore]

The problem that you face is that your twisted pair will be made by twisting two insulated wires together. Each wire has concentric insulation. The outer insulation boundary with the surrounding air does not conform to the circular, (but eccentric), electric field about the wires. So there can be no simple analytic solution.

The electric field of a twisted pair is usually modeled as a single wire reflected in a plane of symmetry. The eccentric equipotential field of a single wire above a ground plane demonstrates the concentric insulation problem.

If you have a TDR then you could experiment with different wire diameters, against different insulation thickness combinations, to find a solution with optimum impedance.

 

[Babadag]

If you refer to twisted pair cable characteristics [impedance, capacitance, attenuation and so on] this link may be useful [?].
http://homepage.ntlworld.com/m.al-asai/papers/2000-1.pdf