How Can Changing AC Sockets Cure Hum?

Mostly right except for wondering about ground. It's a safety thing and therefore critical in my book.

Theory says that a completely unloaded ground will/must have a very low impedance to earth ground. So when (not if) a fault occurs there is a very brief but large current to ground via a safe path. Large you say?? Yes, large enough to trip whatever breaker/fuse 'over current protection' and shut down the offending circuit. That's why the fault is brief. If anything gets in the way of that fault current it may not be large enough to make it brief enough to prevent damage. This is why GFIC and Arc-fault devices are still a good idea. Breakers do not work fast enough to protect people.
 
I see Aaron chimed in just before I did. A different way of looking at it is always good.
 
Neutral - in non-scientific terms (electrons and atoms), the neutral wire is the "return" path for the electrical current serving any given device.

Thanks. I did somre more reading on it and this closely matches that. I might say that the neutral wire complets the circuit, rather than that it is a return path, since this is AC we're talking about. Current flows both ways through the hot and neutral wires. What I didn't take into account is that, when all is as it should be, no current flows through the ground wire, even though it is electrically connected to the neutral wire (but not in my mixer, be that audio or kitchen).

The purpose of a ground path to earth (ground) (in layman's terms) is if somewhere in the electrical path/device/wire/etc electricity is getting lost somewhere-and not making it's way back to the neutral/ground bus bar at the tie in, breaker panel, etc. It provides a safe path for electrons instead of taking the shortest path to ground, which is what, by nature (science, actually) they want to do. The shortest path could be a priest, performer, or technician, as mentioned above.

I see. So, if the hot and neutral wires were always intact, and a complete path from one, through a device, to the other were always to exist (or, were not to exist because the path were to be safely interrupted, with, say, an "off" switch), the ground wire would never be needed. But, if the path is interrupted, and the hot wire makes contact with the chassis of a device, the ground wilre completes the path back to ground, and in a very low-impedance way, which: 1) means very little current will flow throw anyone touching the chassis and also in contact with another low-impedance path to ground, and 2) will pull so much current that it will trip a breaker.

From Richard Cadena's article, I gather that a GFCI adds two more protections: first, if the ground path itself is not effective, either by reason of it being open, or not low-impedance, a lot of current might flow through a person instead of back to either the neutral or hot wire (depending on which phase of the AC cycle we were in), it will detect this and open the circuit, something a breaker might not do if the total current being drawn is still below the breaker's limit and, second, will do it faster than a breaker would anyway.

PSS- the fact that the lights being at 50% created a louder hum could be a variety of issues. Grounding, unshielded lines run in long, close, proximity to AC lines or lighting lines, etc. Hard to diagnose without knowing more details.

Yeah, one step at a time :). Now that I know a bit more about what a ground wire does, it's probably time to gird my loins and study up on ground loops. The hum we're getting may be from one of those, or something else (or both). Our performance mixer plugs into a channel on a rack mixer that is quite close to the dimmers. I wouldn't be surprised if having audio lines that close the dimmers isn't, itself, a bad idea.

I read up recently on MADI, a synchronous digital audio protocol. Gosh, I'd love to work in a house that used that. I can see how it wouldn't cure every hum problem one might face, but it sure does seem to solve a lot of problems, hum and others, and has a certain logical elegance I like. Also, as it is a digital protocol, not an analog contraption, I can understand it better ;) .
 
Looks like you did some research and found some matching answers! That's always good.

I see. So, if the hot and neutral wires were always intact, and a complete path from one, through a device, to the other were always to exist (or, were not to exist because the path were to be safely interrupted, with, say, an "off" switch), the ground wire would never be needed.
.

Correct.

And yes, you are correct on the GFCI and circuit breaker protection, which *should* provide quick enough protection (fractions of a second, when dealing with a GFCI) to save a person from getting electrocuted.

In an ideal world, audio and lighting power would be pulled from different "phases" and be isolated, which resolves a lot of problems in the professional arena, or even at the college/high school level. This may not be the case in the facility you are helping at. In general, I tend to keep lighting dimmers and sound rigs far, far away. :)

Digital protocols are a good protection, but I wouldn't substitute them for a good ground system that is designed and implemented correctly. I would say that if the "marked" outlet in your case is providing you hum-free results, that's a good option, IF (and only if) the ground is still intact. (Have you purchased a tested from the local hardware store yet?)
 
One thing to note apart from the fault shunting properties of the earthing system is the around metal chassised devices...
The Neutral will in normal operation be at a small but real voltage above ground, because it has finite impedance, and so with current flowing in it, will have a voltage drop across it.
If we bonded chassis to neutral, and you had say 2 circuits in a kitchen, you could have say 2 toasters with different voltage potentials on each and get a tingling touching one in each hand.

There is also a benefit for dissipation of EMF and static to ground through the chassis.

Talking of breakers and humans...
Let's be clear on what breakers are there to do, it's to stop the power flowing before the wire gets so hot it causes an electrical fire. They are NOT effective against shock protection in most circumstances.
It takes roughly 30 - 50 mA across the chest to cause ventral fibrillation, when that happens you need a hospital, let it continue for long and you'll need the morgue instead. Even 100 mA is NEVER going to be the different between a breaker tripping and holding.
The body has an impedance in the order of 100 k ohms with dry skin, but that can quickly drop to 1k or so with wet skin, so mostly we sit somewhere in between. That means normal mains voltages are enough to generate a problematic level of current.

I can't speak to US GFCIs, but I can talk about RCDs. If the vector sum of all current carrying conductors on a circuit is >30mA, that sucker will trip within 100 mS to meet standards. In certain applications, it reduces to 10mA to trip (healthcare areas being one such example). There is ONE fault that an RCD won't save you from... Making a circuit between any of the current carrying conductors. If you end up with active in one hand and neutral in the other with rubber shoes, no RCD will save you, it sees the currents coming back just fine. Ditto between actives in a 3 phase system. About now's where the reminder about one hand behind your back when dealing with live anything comes into play...

If you want to talk MADI vs analog audio, let's do that in the sound forum, if only it was as simple and nice as you like to think it is...
 
It also possible to avoid hum issues in analog interconnections without doing anything special with AC grounds or power "cleanliness." Whole books have been written about this, but here are a few tips for making connections between equipment separated by significant distance or where both ends are powered from different AC sources:

1. Make sure that the outputs and inputs are truly balanced. Unbalanced audio and impedance balanced inputs and outputs need to be avoided. Manufacturers who make such equipment should be shot.

2. On line-level circuits, disconnect and isolate the shield at one end of the cable. This is typically done at the input side. Commercially made, XLR ground lifters should be in every audio kit.

3. An audio transformer, with a Faraday shield, can provide extra ground current isolation and exceptional common mode noise rejection. The transformer usually works best at the input.

Consider that the public telephone network has made noise free audio connections over miles of distances using balanced audio pairs and audio transformers, and they don't use any shielding. Not only that, but their cables run alongside power lines. These two concepts are very powerful.
 
2. On line-level circuits, disconnect and isolate the shield at one end of the cable. This is typically done at the input side. Commercially made, XLR ground lifters should be in every audio kit.

Truth. This is actually more common than people think. I've come across several installs where the DSP was located near the mix location (usually during the initial install), and the ground was disconnected at the DSP from every run from there to the amp rack, backstage in a loft. The system is quiet, all things considered, and the power likely isn't isolated or split on a specific phase.

Good points.
 
One thing to note apart from the fault shunting properties of the earthing system is the around metal chassised devices...
The Neutral will in normal operation be at a small but real voltage above ground, because it has finite impedance, and so with current flowing in it, will have a voltage drop across it.
If we bonded chassis to neutral, and you had say 2 circuits in a kitchen, you could have say 2 toasters with different voltage potentials on each and get a tingling touching one in each hand.

Ah, maybe this is one reason why there aren't just two neutrals (or two grounds) in parallel, each serving as a backup to the other.

Such an approach (or just bonding chassis to neutral) also has current flowing through the chassis, doesn't it? If it were possible to disassemble a chassis, one might open the circuit it was a part of, with one's left hand on one side of that circuit, and one's right hand on the other...
 
How much of their success might be due to band-pass filtration?
None. Broadcasters used wide-band telephone circuits for 70 years. Those circuits were conditioned to be flat from 30 Hz to 15 kHz, and they passed strict specs for noise. On plain old telephone circuits, the harmonics of 60 cycle hum is still readily apparent, so users would definitely complain.
 
While we are in the weeds of neutrals, grounding and safety it's worth noting that most power switches do not switch the neutral. Perhaps they should for safety, protecting from that minor voltage when other devices are in use. I suspect it's due to single pole switches being cheaper.
 
While we are in the weeds of neutrals, grounding and safety it's worth noting that most power switches do not switch the neutral. Perhaps they should for safety, protecting from that minor voltage when other devices are in use. I suspect it's due to single pole switches being cheaper.

Also important to note this is exactly why switches should not be used for power-disconnects when doing service work, especially in a somewhat unlikely yet very dangerous situation where the neutral is (incorrectly) switched instead of the hot side. Not to mention switch loops, three way circuits, and all sorts of other legitimate but potentially unsafe configurations.
 
'Forbids' with a few key exceptions, such as switching all other conductors simultaneously.

That's the only way I've seen it done.

Sent from my SM-G900V using Tapatalk

Indeed, a 'neutral' (grounded conductor) may only be switched or disconnected until ALL "hot" conductors are disconnected. They may also be disconnected or switched simultaneously.

How many times I have seen "switched neutrals", especially in lighting circuits or switched receptacles is quite disturbing however (and very dangerous). With the 'neutral' disconnected there is still power there. I've seen lights switched off glowing, or not turn off at all. Also seen someone turn off a switch and then BOOM goes the box when you go to work on it without verifying that power is truly disconnected.

This is why the MOST IMPORTANT tool for someone doing electrical work is their assorted METERS (mainly multimeter). Always verify.

I would take my multimeter to both the receptacle that gave the hum, and the one that does not and "poke around" to see what is what pretty much immediately. HONESTLY IMO you should ALWAYS meter your power BEFORE hooking up any audio equipment. Too many buildings have bad power/wiring.

Test Hot to Ground (in North America) = about 120v
Test Hot to Neutral ("") = about 120v, and similar as the above, if not the same.
Test Neutral to Ground = about 0v (less than a volt or so is fine, any more is a concern)


EDIT:

The meters I ALWAYS have in my car are:

True RMS Multi-meter (CAT III at least, mine is CAT IV)
Non-Contact Voltage Tester (NEVER trust it, but it is great for quick checks and such)
Receptacle Tester (most of mine have GFCI test functions, one with AFCI test)
Network Tester (doesn't apply to this, but I always have it handy)
XLR "Rat Sniffer" Tester (how often this saves your butt on some gigs)
 
Last edited:
I would take my multimeter to both the receptacle that gave the hum, and the one that does not and "poke around" to see what is what pretty much immediately. HONESTLY IMO you should ALWAYS meter your power BEFORE hooking up any audio equipment. Too many buildings have bad power/wiring.

Test Hot to Ground (in North America) = about 120v
Test Hot to Neutral ("") = about 120v, and similar as the above, if not the same.
Test Neutral to Ground = about 0v (less than a volt or so is fine, any more is a concern)

In the OP's case, I'd meter between the two receptacles - all combinations.
 
On a Saturday morning trip to Menards I ran into a local handyman. In my basket were about a dozen white 15 amp receptacles. The handyman asked, "Why are you buying them without ground?"
I replied, "They're going in to my mother-in-law's house. Cosmetic change... after 50 years she's tired of ivory."
Handyman sniffed and said, "Well you just hook the ground up to the neutral... they both go to the same place in the breaker box anyway"
I wish I had thought fast enough to ask, "What about the homeowner or the next handyman who flops back and white in an upstream box?" (I did fish new Romex to specific locations where she had been using a ground lifter. I remember how popular "cheater plugs" were when grounding became code in what was it, 68?)
 

Users who are viewing this thread

Back