Ground and hum.....???? Explain

[aron]

I've been bothered by not understanding this simple concept: ground..... and hum.

I was working on a guitar and after carefully soldering everything (new pickup, new pots etc) it HUM. BADLY.
OF COURSE I assumed that it was some sort of grounding problem and was due to my wiring. But guitar wiring is so darn simple, how could it be? I started to debug and to make a long story short, it was the cheap jack!

1st case:
Can anyone explain to me why it would have pronounced HUM if the jack was somehow not contacting very well?
Follow me on this.. If I plug in my cord and don't plug it into the guitar - it hums, but in this case not too loudly.
If I plug in the guitar - the hum is increased rather loudly. Is it because the guitar is acting as an antenna? Wouldn't the jack have _some_ contact with the cord, at least a little? Why would it hum louder?

2nd case:

I had an amp I modded and I remember a section of it humming badly after the mod. Someone suggested that one of my pots might not be grounded properly. Now why would this cause hum? If a pot is not connected to ground - it wouldn't add hum would it? Wouldn't a bad connection to ground act like a high value resistance?

It seems like these two cases might be related, but my guess is that the first case, the guitar is acting like an antenna. The 2nd case, I don't know. I would seem to me that lifting or creating resistance in the circuit would be similar to taking it out of circuit. Perhaps it's _where_ the pot is that might cause the hum by increasing the impedance of a input section.

hmmmm.....

[zpyder]

an excellent question... can't wait to read some educated responses.

cheerz,
spyder

[aron]

hmmmm or should it be hummm......

I think my antenna guess is correct. Check this out:

http://www.neelyguitars.com/Pickups.html

But why the second case?

[aron]

In my 2nd case, the amp would hum when the pot was turned down but humming would stop when the pot was wide open. Very strange.

[Gilles C]

Maybe the path to the ground (even through the Vcc) is shorter with the pot at max (through caps and the other components, and that you are adding resistance when turning  the pot to level 0 toward pin 1, which is not connected to the ground. So the input picking hum is more isolated from ground than before.

Gilles

[Rob Strand]

While it is techinically correct to think of the guitar to be antenna it doesn't actually give much intuition into what is going on.  For example in case one the small exposed tip of the jack is already acting as antenna, so why should the guitar be a better one?

The key to understanding this stuff is that, at low frequencies, interferrence (noise, humm) gets in by a capacitive coupling, or inductive coupling (like a transformer).  The cases you are talking about are caused by capacitive coupling.  In any environment there are noise generators which includes mains wiring, fluorescent lighting and switchmode power supplies.  The metal surface of these forms one plate of a capacitor - think of a noise voltage connected to one end of a capacitor.  The plate of the capacitor is the terminal that connects to the input of the amplifier (jack tip or guitar). 

From physics you know the size of the capacitor get bigger as the areas of either plates increase or the distance decreases.  In case 1 the jack tip has a small area and the capacitance is small.  When you connect the jack to your guitar it connects to tip to the wiring, pickups, and metal parts of the guitar.  Normally there are shields in the guitar and these connect to the ground terminal of the input.  Most of the capacitive coupled interferrence hits the shield and conduct down to earth through the cable ground, and is diverted away from the signal lines.  If however the guitar ground isn't connecting the whole of the guitars conductive parts channel the capacitively coupled interferrence to the input, there is nowhere else for it to go.  The capacitance is much larger here, because the area is large, and so the interference signal is stronger.  In that mode the strength of the interferrence boils down to the area of the "receiving" terminal.

Suppose now you touch the strings.  In most cases you will find it makes the interference stronger.  That is because your large body area increases the of receiving plate area.  However it can get a bit more complex.  If you were close to a grounded metal object there is a capacitance between yourself + the guitar to ground.  So there are three termimal involved the interference plate, the plate connecting to the amp input and then the external capacitance connecting to ground.  If you draw a picture it forms a capacitive voltage divider.  As the capacitance between yourself and ground gets larger it shunts away more signal and the interferrence level drops.  The best example is case 1, you touch the strings and the noise get louder, but you then touch something grounded and that configuration will usually have the least noise.

Once you understand how the capacitors are connected case 2 is very easy to explain.  When you dial down the volume control the volume control pot is like two resistors R1 which connects the guitar pickups to the amp input terminal and R2 connecting the input terminal to ground.    The noise source is connected via a relatively small capacitance, through space, to the guitar side of R1.   Circuit therefore is a interference voltage source, then that connects to the capacitance, that is in series with R1, the other side of R1 goes to the amp input, then R2 connects to ground.   Usually the impedance of the capacitance is larger than R1 so you can ignore R1 and that simplifies things to:  interference voltage source, capacitance, amp input, then  R2 to ground.

If you draw that out you basically have an adjustable high pass filter where the R2 part varies with the volume control position.  As you dial the volume control down the high pass filter cut-off frequency is increased and that causes the low frequencies to be attenuated more and hence the interference level drops with volume control position.

There's one other thing which can be explained by the high-pass filter view.  If you could listen to the mains signal it has a very strong low frequency content and the buzzy part is very low.  When you listen to capacitively coupled interference it sounds buzzy.  The buzzy parts are mostly high frequency harmonics.  When  the mains is filtered through a high pass filter the low frequencies get attenuated more and what this does is accentuates the buzzy high frequencies - like a treble booster.

[idlechatterbox]

From an engineering/design standpoint, I've never understood why there isn't more research into a non-humming amp (or pedal, guitar, etc). For instance, one of the most often mentioned sources of hum is the 120v house wiring, or a flourescent bulb. Since we're not going to stop using 120v 60hz any time soon, why is a known design problem (as in, known since the days of Thomas Edison) still causing hum? I guess I'm trying to think of analogies to other areas of design and problem-solving, and in most of those cases it seems that the "barriers" to solving the problem are things like strength-to-weight ratio, limitations at the molecular level (e.g., heat or expansion problems), the principle of diminishing returns (you could make a computer the size of an eraser, but it would be hard to type on it), and so on. With the problem of humming amps and other audio gear the response seems to be that we'll just have to get used to it, and that's what you read in an electronics text from 1950s with a dude in a crew cut on the cover. If the amp and guitar-pickup designs are inherently hum-worthy, why don't the designers just toss out the baby and the bathwater and design amps "for the person who uses 120v house wiring"? Or is there such research underway and I'm just not aware of it?

Next election, I'm voting for the candidate who can promise an end to hum in my lifetime. Haven't we all suffered enough???

I've got a few things going on right now, but I'd like some really smart people over in Cambridge to solve this

[R.G.]

Good explanation Rob.

Just for completion -

Hum is with us always because we use 50 or 60Hz signals in our power lines and surround ourselves with it. While the 120/240V stuff in our houses is bad enough, the nearest transmission line is running typically at a couple of thousand volts up to many kV for the big towers. We also use a lot of it, so the currents are big, leaking big fields of magnetic flux at the power line frequency.

Hum gets into audio three main ways and conbinations of them: capacitively, inductively and conductively. The only way we don't get power line is by RF emission, and that's only because the impedance of free space is so high at 50/60Hz, and so the coupling is very poor.

Rob has done a good job of describing the capacitive path. What most people are not aware of is that the space we live in is filled by high voltage fields. There is a substantial voltage difference (in the 100V range, can't remember exactly)  between your head and your feet simply because the atmosphere is electrically charged even without power line radiation. You don't get shocked because the impedance of the atmosphere is so high that your body "shorts" the voltage. But high impedance impedance inputs can and do detect it. Rob's not is a good explanation.

You also get hum inductively, because there is a magnetic component of power line emission. You can see this by connecting up a loop of wire to an oscilloscipe probe. The bigger the loop, the more power line voltage you will see around the loop. It gets even bigger as you move it near a hum emitter, like a transformer. One trick used to place transfomers for minimum hum is to use a loop on a scope to "map" the emissions from a power transformer to put sensitive stuff in a low hum place. This loop pickup is one way you get ground loops. The other is conductive.

Conductive hum pickup happens because power line currents travel through real wires, which are really low value resistors. This leaves a voltage equal to the current times the resistance across the wire, and if you thought the wire was "ground" and used it to reference an input to, the input will amplify the power line frequency. You can also get power line buzz conductively, as things like fluorescent lights and dimmers make spikes every power line cycle, and these spikes ride along on the lower frequency power line wires and pass through the capacitances of the transformers and rectifiers into our circuits. Our ears note that there's a spike happening every 120th or 60th of a second and tell our brains that there's 60 or 120Hz in there.

Notice that all of these effects happen both in free air and inside boxes with any AC power going into them. A tube amp with a big fat power transformer has its own hum factory stuffed in there.

From an engineering/design standpoint, I've never understood why there isn't more research into a non-humming amp (or pedal, guitar, etc). ...  Or is there such research underway and I'm just not aware of it?

There is, and you're just not aware of it. Also, it's not on a par with the search for a cure to cancer.

Every engineer has a boss on his back to make things cheap and to quit designing when things are good enough. That almost always involves the stuff working to past the warranty period. Some engineers sneak through things that survive past the warranty period, but the bosses don't usually know that. When you get down to where the hum isn't noticeable, there is a powerful motivator - your paycheck - to quit making it better, except in a few cases.

The situation is also not particularly in a designer's hands. Every engineer knows how to solve the hum problem at one fell swoop - use only DC power. We'd still have ticks and dips as things start up, but no hum. But that involves replacing every single piece of electrical equipment in the world. We're not going there. We also know how to make hum much better in things like guitar amps - get off that silly coax shielded wire you use for an input cord and use shielded twisted pair in a balanced configuration and make the setup immune to hum. But we can't even get people to do that. A good audio engineer could design a setup which rewired a guitar and amp with STP and used a second tube section in the input as a diffamp that would make the guitar cord immune to hum pickup. You can guess what an effort rewiring every guitar and amp front end on the planet would be. Can you guess how many guitars or amps you would sell that did not have a single shielded coax cable jack, only a shielded twisted pair jack? In this case, it's not the technology, it's the PEOPLE who won't buy better amps and guitars.

[Paul Marossy]

Hum is with us always because we use 50 or 60Hz signals in our power lines and surround ourselves with it. While the 120/240V stuff in our houses is bad enough, the nearest transmission line is running typically at a couple of thousand volts up to many kV for the big towers. We also use a lot of it, so the currents are big, leaking big fields of magnetic flux at the power line frequency.

I wonder if part of the reason why I generally never experience many problems with EMI in general is because most of the power lines in the city where I live are buried?

Conductive hum pickup happens because power line currents travel through real wires, which are really low value resistors. This leaves a voltage equal to the current times the resistance across the wire, and if you thought the wire was "ground" and used it to reference an input to, the input will amplify the power line frequency.

I never really thought about it that way, but that's right - each wire has a certain amount of resistance from one end to the other. I used to think that in tube amps and what not, the wires are all so short that ground loops weren't really possible. But, after building three tube amps and converting another one to a star grounded scheme, I can see that there is something to this after all - not that I ever doubted the stuff I read at GEO. I work in a consulting engineering office, and voltage drop is a concern, but we are talking about much greater lengths of wire than 10 or 12 inches. I just never really thought about it on a smaller scale before.

A tube amp with a big fat power transformer has its own hum factory stuffed in there.

Amen to that! 

[stumper1]

Why do some guitars buzz (Hum?) until you touch the strings and some don't? I have seen several Les Pauls with this problem while the majority are just fine. I've asked this question before - not here - and only got , "just 'cuz".

[Rob Strand]

> Why do some guitars buzz
It's quite often the lack of shielding in the control cavity, around wiring and arounds the pickups.

[Paul Marossy]

I have a bass with active pickups that hums no matter what I do to it. I've made sure that everything is properly grounded, shielded the control cavity, etc. No matter what I do, the darn thing buzzes. It's really annoying and it shouldn't be happening, but it is. 

[idlechatterbox]

QUOTE:
"Every engineer has a boss on his back to make things cheap and to quit designing when things are good enough. . . . . Can you guess how many guitars or amps you would sell that did not have a single shielded coax cable jack, only a shielded twisted pair jack? In this case, it's not the technology, it's the PEOPLE who won't buy better amps and guitars."

Thanks for the nice response. I've passing familiarity with the day-to-day life of an engineer    But I really wasn't getting at an explanation that had to do with predictions about what people would or wouldn't buy, or what an employer would decide is cost-effective enough to let employees work on the design. In my view, the prohibitive cost of a hum-less amp might explain why you won't see them in Target stores, but it doesn't explain why people who demand top performance in every other regard, and aren't squeamish about paying for things like 300 dollar pickups, boutique pedals, etc, wouldn't be willing to pop for it.

Either way, that's not to say that you're wrong, and we can agree to disagree. You're surely correct about DC power. Like you say, to the extent that amp-hum is a necessary by-product of the way juice gets to the outlet, a hum-less amp would require a different world, not just a few circuit changes.

In some ways, I'm screwed. I live within a baseball's throw from a series of electrical towers. The kind that you can stand under with a flourescent bulb. Added to that, I live in an apartment complex, so even if I'm not running my vacuum (and I'm usually not), my neighbors are. I need to learn to embrace hum, but it's hard 

thanks again

[R.G.]

I'm sure that there exist people who would pay for a hum-free amp. After all, there are speaker cables that cost multiple thousands of dollars - just for the cables - and more for the rest of the system. I was just trying to say that there are many reasons why such stuff doesn't exist, and they're not all just the poor design of the equipment.

It's an interesting thought. I certainly could build a no-compromises amp. Separate power box, toroidal transformers, all balanced wiring and amplification stages, rare metals wiring, rewired and shielded guitar with balanced outputs and balanced cord, that kind of thing. I'd price it at about $18K, and if I sold one, I'd make a second one, I guess, and so on until I ran out of chump... er, clients. That's how the esoteric audio market got started, after all. I don't know if I have the stomach to do the advertising necessary.

You do have a big problem with where you live. There is a hum eliminator that you could do. It's pretty simple, too. You take the AC power line voltage and half wave rectify it. Then you run it through four bandpass filters, one each at 60Hz, 120Hz, 180Hz and 240Hz. Each of those goes into its own two-stage phase shifter, and the outputs are all summed at variable volume with the audio you're listening to.  The result is the ability to fine tune out 60Hz hum by diddling the volume and phase of the 60Hz filter, and likewise, its first three harmonics. This will get the majority of the power in any power line hum.

This hum canceller is not a general solution, because you have to tune it to exactly the hum caracteristics you have RIGHT NOW, and it will de-tune if any of the hum changes. But it might get you past your immediate problem.

I live 17 miles out of town and my power comes from a single transformer on a pole 200 feet from my house and 3000 feet from the main feeder line. That helps a lot. But my house itself still makes hum loops happen.

[moosapotamus]

...get off that silly coax shielded wire you use for an input cord and use shielded twisted pair in a balanced configuration and make the setup immune to hum...

What about keeping the silly coax and putting the differential preamp in the guitar instead of the amp?
Someone from another forum sent me this a while ago, but I have yet to try it...

Differential Input Preamp

Any merit to an idea like that?

~ Charlie

[R.G.]

putting the differential preamp in the guitar instead of the amp

It's better than what we have now.

If you get the impedance on the cable run down under 1K, things get a lot better. You still have to worry about hum pickup directly by the pickups and guitar shielding, but it's better. Actually, you can do most of this good with even single ended preamp on the guitar, or even a cable driving buffer.

[Rob Strand]

>active pickups that hums no matter

It's probably the pickups themselves.  Sometimes you can reduce humm/buzz by using copper tape in the pickup cavity and making sure it come all the way up the sides of the control cavity (most people do this).   Other times it's the 'hot' side of the windings being too close to ungrounded pole pieces.  You can get an idea of pickup problems by laying you finger over the centre of pickups and without touching any grounds like the strings - try it next time you change the strings.  Move your finger to the sides to see if it's getting in the sides.

For passive pickups sometimes you can reverse the connections of *all* the pickups in the hope of moving the hot side of the winding to a quieter place.  This is best as a last resort after you have fully shielded to pickup cavity.

For actives you can't get inside and usually there is just a coax hanging out so you can't flip the winding wires.   There's not much you can do.

It could be something else!
 

[tubescreamer1010]

doesnt grounding stuff shield it from external interferences so of there is resistance on a grounding lug of a pot then there isn't enough ground on that pot and it won't be shielded as well.

[billings]

So how would I hook up my Fuzz Face for twisted pair ins and outs?

[Paul Marossy]

It's probably the pickups themselves.  Sometimes you can reduce humm/buzz by using copper tape in the pickup cavity and making sure it come all the way up the sides of the control cavity (most people do this).   Other times it's the 'hot' side of the windings being too close to ungrounded pole pieces.  You can get an idea of pickup problems by laying you finger over the centre of pickups and without touching any grounds like the strings - try it next time you change the strings.  Move your finger to the sides to see if it's getting in the sides.

I think it is the pickups themselves, although, I haven't tried shielding the pickup cavities. That might be of some help.

For passive pickups sometimes you can reverse the connections of *all* the pickups in the hope of moving the hot side of the winding to a quieter place.  This is best as a last resort after you have fully shielded to pickup cavity.

That's a trick that I hadn't heard of before.

For actives you can't get inside and usually there is just a coax hanging out so you can't flip the winding wires.   There's not much you can do.

Yeah. Bummer.

It could be something else!

Yeah, those darn hum gnomes! Argh.