Does impedance balancing in the far end make no sense?

Question

Below shows a conversion of unbalanced line to a balanced one:

My question is about: if the unbalanced line is very long which means the signal source and the isolation amplifier is apart from each other 100 meters for instance.

In that case, if the balancing isolation transformer is at the far end i.e. right before the difference amplifier, and the unbalanced line is very long what happens to the CM noise coupled on the unbalanced line?

Will it be rejected at the isolation amplifier? If no, does that mean the isolation amplifier should be right after the signal source?

Answer 1

It's complicated. If there's truly just common mode noise on the unbalanced line, then it'll get rejected -- or not -- by the transformer. A transformer that's actually built in a way that reflects your schematic symbol would probably do pretty good, but there's no general statements that can be made. Unless the unbalanced line is exceptionally well isolated, however, chances are high that common-mode noise will leak into the signal its carrying, and that'll get in to the signal seen by the differential amplifier. In particular, if it's grounded at both the near and the far end, the scheme probably won't work well.

Note that a place I worked at did something kinda like this -- it used 75-ohm coax video lines in a noisy environment, with both the center tap and the coax shield going to a differential amp at the receiving end. But the guys designing the wiring had to take care that the coax shield was isolated from system ground all the way to the receiving board.