If you’ve ever done home recording, you’ll probably have met the beastie known as a ground loop.
Here’s a scenario: you turn up somewhere to record or broadcast a gig. You’re set up in some quiet corner of the venue. You get your balanced audio feed from the FOH mixer, run it to your studio and plug it into your kit. Immediately you hear a tremendous humming noise coming from the channel. You check and double check your cables, you check the inputs. You tear your hair out. You have encountered a Ground Loop.
In this instance, the kit in the studio is connected to a different power circuit to the kit which you’re getting the FOH feed from. In the schematic above, you can see that the ground connection between the two equipment is separated by a significant amount of resistance, caused by the cabling and other equipment between your power supply and the power supply used by the event. The two grounds do not have the same potential – and therefore a current will flow between them.
When you connect up the audio cables, you are completing a circuit between the two differing potentials through the ground connection (the shield in the cable). Thus, a current will flow through this connection, which will affect the reference ground on the receiver, which will in turn affect the signal coming into the input (shifting phase and so on). Since the power supply is AC, this is characterised by a hum at 50Hz (in the UK) – possibly with other harmonics too. Phase shifting will make it even weirder.
Here are a few solutions that we’ve used in the past:
Solution 1 – Use a Ground Lift cable
A ground lift cable is one in which the shield (or screen) is only connected on one side. Blue Room has a nice wiki entry on ground lift. There are disadvantages to this approach – in that it will reduce the shielding ability of the cable, since it is only grounded at one end. This is particularly problematic on long runs, where interference is more likely.
A DI may also have a ground lift switch. I’m going to assume that you’re running a balanced feed anyway (because you damn well should be!).
Solution 2 – Run from the same power
If you’re running from the same power source (same phase), it’s far less likely you’ll get any loops occurring. However, you need to take care that your power runs aren’t too long, or you’ll end up making the same problem all over again. You can minimise this by using the fattest power cable you can get your hands on (i.e. less resistance).
Solution 3 – Don’t use a cable at all
If there is no return path (i.e. no physical connection between the two points) then there is no way for a ground loop to occur. By what magic might you achieve this you ask?! A radio link. CSR has previously used a belt pack microphone transmitter and receiver … but since the belt pack is battery powered this poses a problem for long broadcasts (“The link has gone down!” “OH shi-*runs*”). We alleviated this problem by making a “fake” battery, connected to a regulated wall-wart DC power supply. However another nice trick would be to use an IEM (In Ear Monitor) transmitter tuned into the same frequency as a radio mic receiver (never tried it, but recommended by a soundee at an event we were broadcasting)
Your limitations here are in the range and quality of the radio link. However, with decent kit, you shouldn’t have any issues. We’ve been using a Sennheiser ew100 receiver, and with a pair of A1031 antennas we manage to get the link to go almost 300m in an open air test. Given that the transmitter is stationary you can get away with just one antenna, diversity is only really useful if you’re moving around and could walk into a dead spot. If your transmitter is in a dead spot, you can move it and after that it won’t go anywhere (or at least, it shouldn’t!).
So there you are. I’m sure there are other (probably better) solutions, but those are the ones that we’ve tried and tested.
