You asked for it, so don’t blame me for inducing your headache here brother 
(This is for when you don’t have balanced outputs obviously but amp does have balanced inputs).
From someone WAY smarter than me:
“First principles approach is there will always be some common mode ground noise currents flowing in your audio cables shield, due to low freq magnetic fields, RF from SMPS and cell phones, and from direct injection by the various products’ AC/DC power supplies.
And your audio cables all have non-zero ground impedance. So if this CM noise current flows in part of the cable that’s used for signal (e.g. RCA’s shield is also the signal negative wire) then it’ll corrupt the signal as hum, machine noise, hiss, etc according to V=IR where R is the RCA shield impedance.
But if this CM current is given a preferred pathway (e.g. separate shield wire in XLR cable) then it won’t corrupt the audio. Similarly if given a lower impedance path (e.g. very short RCA, or more-copper in the RCA shield) then it’ll corrupt less because the “R” term is smaller in V=IR.
An RCA>XLR cable, if built properly, will provide a mostly separate pathway for CM noise current to flow. It’s a half-way house.
Now the question is whether or not an amp can handle an unbalanced signal on its balanced input…. if the amp is designed right as a “differential” input stage, then no problem (as is the case with most product).
XLR pin 1 Zin is < 1 ohms to ground, and XLR pin 3 Zin is much higher around 1 meg ohms to ground. CM current can be modeled as a current source. According to Kirchoff’s Current Law and the ratio of these impedance pathways, a proper RCA to XLR-3 cable will force the vast majority of CM current to flow into pin 1. The pin 3 wire carries negligible CM current, so incurs no hum-inducing voltage drop.”
Built it myself. The recommended Belden1804 cable (star quad, silver plated) and shielding method came from John Swenson.