Chasing Electrical Ghosts
March 28, 2020 9:28 AM   Subscribe

What's the best way to solve ground loop/EMI issues at home?

I have an audio recording setup that is suffering from (what I think is) a grounding issue. I have not had a ton of success sorting it out. How do you troubleshoot something like this at home?

Some of the electronics in the room include:
-Desktop PC sitting on a wooden floor
-USB audio interface
-Pair of studio monitors
-Electric guitars (these give off a low hum when plugged in to the audio interface)
-Condenser mic (much more noticeable hum)

Everything is plugged in to the wall using their original power cables (not using any power strips). I have noticed that some of the interference gets worse when the ceiling light is on (I think the CFL may be emitting some bad EMI). What else should I be looking at?
posted by backseatpilot to Technology (10 answers total) 1 user marked this as a favorite
 
-Desktop PC sitting on a wooden floor
-Pair of studio monitors

Are these using grounded power cords? I'd expect a yes here.

A ground loop is created when any two grounded devices have their chassis connected, via for instance a shielded signal cable. For that loop to be broken you need to put a device, very appropriately called a ground loop isolator, in the signal line(s) between those devices. Those are essentially 1:1 transformers in a nice case fitted with jack/RCA/XLR plugs or sockets.

-USB audio interface
-Electric guitars (these give off a low hum when plugged in to the audio interface)
-Condenser mic (much more noticeable hum)

For these it sounds like bad, or at least insufficient shielding. Different cause, different solution required, like balanced cables and improved shielding.

(I think the CFL may be emitting some bad EMI).
If it's an early type, comparatively heavy, then it's quite likely equipped with a standard current limiting choke which indeed throws out a lot of EMI. Replace with a standard filament bulb if you happen to still have one, and see if things get better. Or get a different CFL or a LED bulb; a decent brand, not some bottom-shelf cheapo.
posted by Stoneshop at 10:11 AM on March 28, 2020 [1 favorite]


Response by poster: Oh, I forgot to mention - one of the reasons I’m suspecting a grounding problem is that the noise tends to go away when I touch some of the components. The mic especially will stop humming if I hold it in my hand.
posted by backseatpilot at 10:34 AM on March 28, 2020


You can get relatively inexpensive ground loop isolators on Amazon thanks to the fact lots of cars introduce a crap ton of alternator whine when trying to charge a phone while using it to drive said car's aux input.

Unless there is an obvious issue like different audio components being plugged into different household outlets, the ground loop isolator is the thing you use to keep from tearing your hair out for months trying to chase it down.

(And yes, the hum going away when you touch things also makes me think ground loop)
posted by wierdo at 10:48 AM on March 28, 2020 [1 favorite]


How do you troubleshoot something like this at home?

By modelling the setup until I understand it and then working methodically toward fixing it.

Here's a really good backgrounder with a pile of useful advice in it.

For what it's worth, I often find that ground-related noise gets better, not worse, if I run all the equipment off the same (grounded!) power strip rather than plugging it into multiple wall outlets, especially if I would otherwise need to use several groups of these some distance apart within the room. Getting all the chassis grounds wired back to a common connection point at the power strip is helpful because the safety ground wires inside the walls will usually have some degree of noise induced onto them by the adjacent current-carrying mains wires. But the main thing is finding the right places to put decent signal transformer isolators.
posted by flabdablet at 11:47 AM on March 28, 2020 [1 favorite]


An easy thing to do is a buy a ground-removing 'cheater' socket and try using it on individual 3-pronged pieces of equipment to see if the hum disappears.
posted by Dmenet at 12:37 PM on March 30, 2020


As stated repeatedly and emphatically in the backgrounder I linked above, this is not safe.

If you disconnect the mains safety ground wire using a cheater socket, then any ground current caused by the kind of electrical fault that the safety ground is supposed to be protective against will instead flow in the shields of signal cables connected to the cheated equipment.

Signal cable shields are absolutely not rated to carry that kind of current, so if such a fault does occur, forcing them to do so risks electrocution and/or fire.
posted by flabdablet at 12:26 PM on March 31, 2020


Response by poster: I tried plugging everything in to a (grounded!) surge protector, and that seems to have helped quite a bit. Is there any value in buying one that includes noise attenuation/power condition?

Of course all the work equipment that I had to bring home is now generating its own fun spectrum of noise, but that's a different problem.
posted by backseatpilot at 1:20 PM on March 31, 2020


Is there any value in buying one that includes noise attenuation/power condition?

That depends pretty sensitively on how the noise you're hearing is actually getting into the equipment. If, for example, it's mostly switching noise from the PC power supply being capacitively coupled via that power supply's transformer windings, then adding power line filtering upstream of that PC is not going to make a whole lot of difference.

The fastest and least expensive way to deal with this thing is going to involve sketching out a model of your setup and working though it methodically, testing it for noise sources as you go, as outlined above. The slowest and most expensive way is going to involve throwing random fixes at it until you either luck onto a solution or run out of patience or budget.
posted by flabdablet at 6:37 AM on April 1, 2020


This thread might be dead and solved, but if you use a cheater adapter on the single piece of equipment, and it eliminates a hum, you have removed a ground loop and therefore that piece of equipment does indeed have a route to ground and is therefore grounded. You need to be connected to ground at two points to have a loop. There is a danger in that that piece of equipment is now grounding through the interconnects, which can be remedied by breaking the ground on the interconnects and re-grounding that piece of equipment.
posted by Dmenet at 11:26 AM on April 2, 2020


Breaking the interconnects will tell you the same thing with zero safety risk. If instead of buying a cheater you buy the dummy connectors recommended in the backgrounder I linked above, and use them according to the procedure described on page 12 onward, you will also get better information than a cheater can give you about exactly where and how the noise is getting in.

The amount of noise current induced into a ground loop by stray magnetic fields is proportional to the area bounded by the loop, so another way to check whether ground loops are the source of your noise is to reduce that area as much as possible. This can be done without buying anything new at all.

If you have two pieces of grounded equipment with an interconnect between them, try siting them physically close to each other. If that doesn't increase your hum, then try loosely twisting their grounded mains cables together for most of their length and feeding them from the same power strip, and getting the interconnect as short as you can by winding it up into figure-8 loops (not ordinary round-and-round ones!).

If doing those things cuts the hum some, then you have reason to be confident that sticking a decent isolating transformer on the receiving end of that interconnect is going to cut it even more.
posted by flabdablet at 11:29 PM on April 2, 2020


« Older Butter-free desserts that aren't good for you   |   Texts to my personal number are showing up on my... Newer »
This thread is closed to new comments.