On Tue, 11 Mar 2014 20:02:49 -0700, Michael Black wrote:

...snip...

That's the question most of us missed. I could have sworn he said "rf
interference", but when I looked back after posting, realized he hadn't
indicated any specific bit of interference.

...snip....

I assumed stuff like nearby motor drivers, or worse, from those new solid
state ballasts. The cheap ones can get VERY energetic. and coupling
magnetic into audio is not an easy item to stop.

People are so used to 'shielding' with electrostatic shields and lowering
the impedance to reduce high impedance interference, that they are
completely caught off guard when confronted with low impedance noise
sources, like magnetic field. May have noticed that lowering the impedance
makes little difference.

Sometimes, high impedance noise gets 'converted' somewhere to a low
impedance source which is really mindboggling, since you recognize the
noise and expect it to be coming from high impedance.

Best way,...balanced pair, shielded. And I MEAN balanced over the WHOLE
spectrum! DC to GHz, 10GHz if you can do it. That includes termination
circuitry, Tx/Rx, and high quality cabling with very symmetrical conductor
placement and spacings within that cable. That is, if you're goal is 120
dBV, or better 140dBV, down noise floors you had better pay attention to
everything.