On Fri, 24 Nov 2006 15:49:21 +0000 (GMT), Jim Lesurf
wrote:

In article , Serge Auckland
wrote:


Impedances as above, but as to what frequency they apply above, it
depends on the capacitance and inductance characteristics of the cable.
The formula is Fc=2/2Pi x root(LC) according to my reference, where L is
the inductance/unit length and C is the capacitance /unit length.
Characteristic impedance is root (L/C) above the cut-off frequency Fc.

I don't recall the formula, so the above may be correct. However its form
surprises me as I'd expect the value to depend upon the resistance (R')
and/or shunt conductance (G') per unit length as well as the capacitance
(C') and inductance (L') values...

My recollection is that, if we neglect shunt conductance and dielectric
losses, then the characteristic impedance turn over would be at the
frequency where jwL' is comparable with R'.

Slainte,

Jim

The formula is sqrt((R + jwL)/(G + jwC)), where G is the conductivity
of the dielectric, and R is the resistance of the wire - all for the
standard length that yields the L and C values.

I've worked and example with some typical values, and you can see that
the effect only comes into play at lowish frequencies. By the time you
get up towards the top of the band, where it might matter, the cable
has returned to its true impedance, and the lumped model is no longer
needed to describe it - the transmission line is just fine.

http://81.174.169.10/odds/cable.html

d

--
Pearce Consulting
http://www.pearce.uk.com