In that case - using the power supply 'ground' as the reference for the input signal - then it is not a balanced input as referred to in the OP's question.
That cable has no connectors to make any use of a balanced input. It simply makes the balanced input into a single ended input. Thus losing any CMRR advantage of a balanced input.
It works for my purposes, and 95% of my system is single ended RCA input anyway... CMRR is great if you have noise... I made a great effort to eliminate that type of noise from my system so balanced I/O is pointless here.
It really depends on whether the balanced input is properly implemented with pin 1 going direct to chassis (or in any case not being part of the audio signal circuitry) or whether it has the "Pin 1 Problem". If it does not have the pin 1 problem then it is preferable to connect the screen at the unbalanced output.
In this way the balanced input (XLR) reads the differential signal from the RCA and good shielding is maintained. CMRR can be improved if you impedance balance the line to XLR pin 3.
If there is a pin 1 problem then a parallel RC network - eg 100R//100n - in the screen connection is usually effective at minimising 'hum' whilst improving shielding compared to no screen connection.
Maybe this could be a good compromise?
The capacitor improves the screen at HF and the resistor helps to balance the impedance?
CMRR of a solid state differential input not being an instrumentation amplifier is rather poor in most cases I guess.
Dealing with equipment that suffers from the "pin 1" problem is rather challenging.
The capacitor improves the screen at HF and the resistor helps to balance the impedance?
CMRR of a solid state differential input not being an instrumentation amplifier is rather poor in most cases I guess.
Dealing with equipment that suffers from the "pin 1" problem is rather challenging.
