Michael Post subject: Has anyone been able to simulate a Twisted
Pair Cable? Posted: Tue Sep 20, 2005 3:27 pm Good day. Has anyone
been able to simulate a twisted pair cable with ADS, Microwave Office,
Eagleware, etc.? I am trying to examine the effect of different
twists and how they would affect coupling between the two wires.
Thank you for your time and help. Michael Top
guest Post subject: simulation of twisted pairPosted: Wed Sep
28, 2005 5:56 pm Because you're dealing with the physical configuration,
you can't use a simulator which isn't physics-based for your task.
Sounds like you need an 3-D electromagnetics simulator, if you want
to investigate this in detail. But - the largest effect will
be the increase in wire length to cover a given distance. You should
be able to calculate this easily. The second-order effects will
be the increase in capacitance and inductance. It may be that
you don't need anything other than the largest effect. At any
rate, Good Luck! Top dstraight Post subject:
Re: Has anyone been able to simulate a Twisted Pair Cable?Posted: Sun
Oct 02, 2005 6:48 pm Captain Joined: Sun May 09, 2004
1:23 am Posts: 10 Location: Morgan Hill, CA (Silicon Valley, Bay
Area) Michael wrote: Good day. Has anyone been able to simulate
a twisted pair cable with ADS, Microwave Office, Eagleware, etc.?
I am trying to examine the effect of different twists and how
they would affect coupling between the two wires. Thank you for
your time and help. Michael What exactly are you looking
for when comparing different twists in a twisted pair cable? A 3D simulator
might help, but I would think that you'd just get different impedances
for different twists and that would be about it. Will the twisted pair
be kept far away from metal objects? That would make a difference too
since a twisted pair isn't shielded like coax is. David
Top Graham Post subject: Posted: Thu Oct 13, 2005
7:12 pm Colonel Joined: Fri Sep 02, 2005 7:25 pm
Posts: 34 Location: Hampshire UK Twisted pair like telephone
wire pairs is around 110 ohms. The impedance of shielded twisted
pair is not much different to the shielded variety. Cat5 network cable
amounts to a relatively high loss, but useful cable for short range
applications, is of this type. You can get individually shielded
pairs cable. In all these, the presence of the shield is to protect
the cable from external interference fields and to limit the amount
of conducted and radiated EMI from the cable. The shield does not primarily
define the impedance, which is a function of the conductor diameter,
the spacing, the dielectric constant of the plastic separating the conductors.
It would affect it a little. The "shield" around coax should
not be thought of as a shield. It is one half of an unbalanced transmission
line. It is what happens if you think of one of the conductors of a
balanced pair getting bigger, and morphing its shape (in cross section),
to surround the other, to eventually assume the shape of coax. The lines
of electric field then become radial between the centre conductor and
what now surrounds it. This coax form of outer very definitely has a
profound effect on the impedance, by its dimension and the dielectric
within. Grounding the outside when there is a standing wave on
a coax cable is a major sorce of EMI trouble. Triax, which is a coax
with a second outer braid, allows the outermost braid to be grounded
(usually one end only), and is an (expensive!) shield - the coax equivalent
of the shielded twisted pair. In theory, if one conductor of
a twisted pair gets up against a metal object more than its partner,
you get some unbalanced current. In practice, with a pair having a high
rate of twist, it is remarkably immune to coupling. I once tried out
some HF cable distribution stuff that had six pairs around a mandrel
plastic, in a jacket that was carbon loaded to make it "pre lossy" so
that it did not change its characteristic much as it aged in use. Provided
all the unused pairs were terminated, the balance between pairs was
better than -60dB. We could run a small electric drill on one pair while
sending a vestigial sideband TV signal up the cable the other direction.
Definitely a "don't try this at home" stunt! Then again, it only
takes ONE leg of ONE pair in a fat cable to get disconnected to have
all the rest get mangled with crosstalk! Ask any telephone engineer.
Posted 11/12/2012
|