600 ohm Twisted Pair Telephone Line

[Mike Monett]

To All,

(Sorry - I posted this to ABSE by mistake. It should have come here)

A while back we were discussing the 600 ohm impedance of telephone
lines, and John L. mentioned the resistance of 24 ga wire had a
significant effect on the impedance. Turns out he was right. The
resistance dominates at low frequency for long lines.

Howard Johnson discusses this in "High Speed Signal Propagation:
Advanced Black Magic". One chapter is "Performance Regions" which
describes the critical regions in transmission lines, how they
affect transmission loss, and how to overcome some of the loss
problems.

At audio frequencies, long lengths of twisted pair act like an rc
circuit, and the resistance of 24 ga works out to about 640 ohms.

This is discussed in section 3.5, "RC Region", at

http://www.informit.com/articles/article.asp?p=101149&seqNum=5&rl=1

The entire chapter is provided online courtesy of Prentice Hall. It
is available at

http://www.informit.com/articles/article.asp?p=101149

and also

http://www.phptr.com/articles/article.asp?p=101149

Very interesting reading.

Mike Monett

 

[Tony Williams]

A while back we were discussing the 600 ohm impedance of
telephone lines, and John L. mentioned the resistance of 24
ga wire had a significant effect on the impedance. Turns
out he was right. The resistance dominates at low frequency
for long lines. [snip]
At audio frequencies, long lengths of twisted pair act like
an rc circuit, and the resistance of 24 ga works out to about
640 ohms.

Hello Mike. I haven't read the articles you mentioned, but
similar assumptions were standard in old telecomms textbooks,
(well, the same in the two I have anyway).

For a telephone cable the conductors were small diameter and close
together, so R>>wL and wC>>G. For such lines, Zo can be shown to
approximate to.... Zo = sq-root[R/wC] at an angle of -45.