Grounding high frequencies 2

[testtube44]

I'm having trouble understanding how to shield high frequencies, if the frequencies are so high, that the ground cable would act like an antenna too. Would it be a bad idea to shield the ground cable ( ) and connect that shield to ground in a slightly different spot? This should create canceling fields and prevent the ground cable from emitting more radiation. Is this correct, or would it just cause more problems?

[HB9EVI]

Often the screen of a RF signal cable actually acts like an antenna; that always happens, when the current flow is not fully differential; that's a so called common mode signal with makes the shield of a coax cable radiating RF; the origin is often feeding a balanced system with an unbalanced feed line.

Those common mode signals can be radiated also over the supply lines of an RF circuit - the solution, whereever it might happen are common mode chokes - that's all it needs.

[testtube44]

What if you're not allowed to change the frequency content of the line? Is there any other way to create a canceling field? I think differential pairs (one opposite the other) would do just that.

[David Hess]

Sometimes a tuned ground counterpoise is used but that does not stop the ground from radiating.  Some improvement can be made by using a ground connection which has lower inductance like a braid or strap which is like using the shield of a coaxial cable.

[HB9EVI]

What if you're not allowed to change the frequency content of the line? Is there any other way to create a canceling field? I think differential pairs (one opposite the other) would do just that.

a common mode choke doesn't alter or attenuate the frequency content on a signal line, since those are differential currents; if you hook a snapon ferrite onto a usb or an ethernet cable, you are actually filtering common mode signals; the data signals are not affected by the inductance the ferrite forms on the wires.

if you think of differential pairs as a balanced signal line, you have to watch out as well to not end up with unsymmetries resulting in radating signals from the wires. of course this issue is no one-way problem only; missing or incomplete symmetry can cause the signal wire acting as a receiving antenna; that's a considerable topic when designing balanced audio gears, it's all about the common mode rejection ratio