N.icholas wrote:I read the rane artice and some others and this is what led to a lot of my confusion e.g one AES standard says no and others standards/conventions say yes.
If you understand the science behind it all, the different approaches make a lot more sense, but there is no absolue right or wrong answer at the present time.
The modern approach is that all of the audio circuitry should be screened -- and that includes the signal wires in cables. So the cable screen should connect to the XLR shell and the equipment case -- and that should all be tied to the mains safety ground too. Essentially, the idea is to maintain an complete faraday cage around all of the audio electronics and interconnects. And, in principle, it makes a huge amount of good sense.
For that reason a lot of commercially made XLR cables will have the link in place between pin 1 and the XLR shell.
However, in times gone by many manufacturers took the approach that there should be two grounds: the clean audio ground (pin 1) and the dirty chassis (the metal box around the equipment that is tied directly to mains safety earth).
To minimise problems with ground loops through the mains safety earthing of separate equipment (in poorly designed products) many designers therefore contrived ways to separate the audio and chassis grounds (at least at audio frequencies).
In this kind of equipment, linking pin 1 to the shell of XLRs will obviously bridge the two grounds and potentially open a whole world or ground loop worms.
So if you suspect you might be using less than perfectly designed equipment that might be susceptible to ground loop issues, then the best advice is NOT to link pin 1 and the XLR shell.
Full screening will still be in place because the XLR shell is grounded to the chassis once plugged into the socket anyway.
Where the screening breaks down is when joining two XLR cables. In this situation, the short bit of exposed wiring inside the connectors (but outside the cable screen) is potentially open to interference becuase the metal XLR shells are floating (not grounded to anything).
In practice, this is unlikely to be a major problem and I've certainly never had problems with it... but you can get the best of both worlds by using a small capacitor to link between pin 1 and the XLR shell. Something like a 0.1uF ceramic disc capacitor will provide a very low-impedance path to ground for RF interference, while maintainig a very high-impedance path at audio frequencies -- so the XLR shell acts as a proper RF screen but won't bridge the audio and chassis grounds at audio frequencies, and so there is no risk of ground loops.
Hope that makes sense.
The other possibly strange thing is that the supplied XLR male cases are black metal and appear to be non-conductive (!)
They are anodised -- you'd need to scrape away the anodising to get back to bare metal to check continuity. I wouldn't bother!
Could getting some ferrite rings help?
Yes.
also fogot to say if I hold the mike (sm58) and touch the desk the level of interference drops ( am I acting as some sort of capacitance). The rf is also there with no mike attached but increases with the mike attached. Also there with no cable coming out of the stage box end. Plus not there if don't go through the snake.
The question is, where is the interference getting in? Is it being picked up by the multi and the mixer is failing to reject it, or is it being picked up by the multi and demodulated also in the multi, so that it's presented to the mixer as an audio signal? It could be either!
Either way, clearly there is poor RF screening in some lines through your multicores. Ferrite beads on the XLR pins inside the connectors at the mixer end will probably help.
But I'd also remake and resolder all the individual core screens to their respective pin 1s. If there has been some corrosion in your unused multicore while it was stored then it could be that some of the connections have started to become high-impedance ata RF and/or semi-conducting. A normal cable-test or continuity test won't reveal this kind of problem.
Does the problem remain no matter what channels of the mixer are used? What is the mixer, by the way?
Hugh
