In rec.aviation.owning Jim Burns wrote:
There's a note in the install manual that says if you need to shorten
the twisted pair, to cut the pair coming from the shunt, do not cut
the pair that is attached to the indicator. We didn't cut either, but
what difference would it make?

I don't know for sure. Possibly the connector used at the indicator end
is harder to take apart and reassemble than the connectors used at the
shunt end. Or maybe they prefer for the wires to be untwisted or more
loosely twisted at the shunt end, rather than the indicator end.

It also mentions routing the twisted pair away from other wire bundles
"if possible".

Mainly this is to prevent other wires from inducing a current into the
twisted pair. The signal in the twisted pair isn't very large; at 60 A
charge or discharge it will only be 50 millivolts (0.05 V) at maybe a
milliamp or two. This is comparable to microphone level audio, and you
might be familiar with how easily that picks up electrical noise.

I would think that interference problems would be consistent and
repeatable.

Probably true. The catch is that to create interference, you'd need to
turn a big electrical load on or off, which will change the ammeter
display anyway even if things are working right.

One way to check for interference would be to disconnect the twisted
pair wires at the shunt and connect them directly to each other and to
nothing else. In this case the ammeter should always read 0.0 A; any
other reading indicates interference.

This appears totally random and either -0.0 or accurate.

I agree that this isn't consistent with interference; it's more like a
broken wire or loose connector that's either making contact or not.

Matt Roberds

Disclaimer: This is based on experience with ground vehicles. I don't
have an A&P; I don't even have a TG&Y. Some of this may not be
allowable owner maintenance. Your mileage may vary.