On Jun 2, 2014, at 12:28 AM, Jim Brown <k9yc@audiosystemsgroup.com> wrote:
> SO - for minimum crosstalk, a very robust shield is critically important, and
> this is just as true at HF as it is at UHF.
In theory, it is obvious that the more radiation you have from a cable, the
stronger the EMF around the cable, and the potential would appear to be there
for crosstalk from cable to cable. This subject has been studied since about
1937 in Princeton, NJ, where ATT was concerned about crosstalk in long lengths
of coax used in repeater applications at frequencies up to about 200 Khz. When
I was in university, those studies formed the sum total knowledge from research
conducted for telecommunications.
Now, as you know, there are three three elements which must be considered in
crosstalk measurements, near end crosstalk, far end crosstalk, and what you
would call transfer crosstalk. What I know is the length of the cables and
their proximity to one another, is a major factor, as is the sheath (in
telephone cable parlance) which covers them, and whether they are terminated
with the cable characteristic impedance. For purposes of research, RF signals
were sent in opposite directions, to simulate the use of repeaters in the
system.
The simple cable is the twinax, that is, two coaxial cables side by side,
carried within a larger cable. We know from measurement that there is no far
end crosstalk because of the opposing direction of travel of the waves. The
near end cross talk is is generally low with such a twinax, but the total
crosstalk can be effected by the surrounding conductive sheath and its transfer
crosstalk.
It is interesting to note, that after all these years, if there is a problem
with crosstalk in high frequency coax, it is never measured directly that I am
aware of. The runs are not miles long, particularly those in high reliability
applications like aircraft and spacecraft, despite the cable bundles being
tightly grouped. Crosstalk measurements are still required for toll and local
area exchange cables which use twisted pairs, of varying lay lengths, to avoid
crosstalk. The military specifications do not require crosstalk measurements
on coax, but rather the cables are swept through out their useful frequency
range and checked for SRL. This is a modern day addition to the military grade
coax, which does find its way into amateur radio transmission quite often.
I had the task to design the first automated testing facility for toll and area
exchange cables. Up to that time in the late sixties, all these measurements
where taken by hand, in a limited sampling plan. Included in the primary
transmission parameter testing was far and near end crosstalk measurement. We
terminated both ends of the cable with fixtures we had built by Clinton
Instruments, the instruments were automatic leveling devices built by the
Siemens Corporation in Germany, and we used a PDP 11 DEC computer to drive the
system. Being able to record so much data, would lead to much improved
designs, and shorter development times for new cables.
Although our system took millions of measurements of crosstalk on twisted
pairs, I must say I never ever took even one measurement of crosstalk on
twinax, or quads. That was a research activity way back when, and I have no
experience with it.
73,
Mike
W2AJI
Former US Diplomat to the ISO and Subcommittee Chairman
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|