Ground Loops????????????

[JohnH]

Very sorry but this has been bugging me for two years, so I decided to raise it again.

One of our tenets in guitar wiring design is to avoid 'ground loops' in our layouts - Star Grounding rules etc. The reason to avoid ground loops would be to minimise a source of hum/noise in the guitar, by making sure each grounded point only has a single route to the ground.

I have had several guitars with inherent ground loops which had no less hum when this was fixed

If ground loops in a guitar are significant, it ought to be possible to invent an analogous test, with a humongous ground loop that clearly demonstrates the effect. So I had a go, across about 10' of living room floor:



The essential characteristics which represent a guitar situation are:

1. No mains or powered amplification within the test area
2. final ground is via the amp cord, there are no other direct grounds.
3. At the jack tip is a resistor (I used 68k), which represents the resistance of the guitar circuitry, which is to be grounded at the other end.
4. The far end of the resistor is grounded by two routes, forming a massive ground loop over several feet.

Why might this hum? The low resistance loop could be induced with signals from electromagnetism in the room, resulting in small but significant voltages across different parts of the loop which are applied to the jack.

But it doesn't. There's no noticeable change in hum with or without the loop. Any such effect in a guitar would be less significant again.
Once the wire is ungrounded however - very noisey indeed.

So who thinks ground loops occur in a guitar, and who can describe a setup that will positively create one?

I'm not talking about other sorts of ground loops which do certainly exist, such as when powered audio equipment is grounded to different mains sockets, or within poorly designed power audio equipment. In these case, currents and hence voltage differentials in the ground circuit are much greater between different points.

(and BTW - shielding is good, and avoid hanging from hot - those two are easy to prove)

Any takers?

cheers

John

[sumgai]

John,

Love your idea, it's the old "put up or shut up" argument all rolled into a tidy little package, ready for testing by anyone who might care to do so.

Personally, I never give small area ground loops a second thought, I consider them to be too insigficant to fuss over. Where they might come into play in the guitar world would be if one side of a loop were to effectively bypass the safety capacitor that goes between the touchable metal parts and ground. It's not exactly impossible for a newfer to do such a thing, I'm sure you'll agree. Blindly following the "no ground loops" tenet helps to prevent that kind of thing, I'm sure.

What size wire did you use? Did you take resistance measurements for each portion of the loop? And 68kΩ seems kind of high to me, no?


sumgai

[JohnH]

Sumgai - thats fair enough. I have not been putting in the safety caps in my builds, but I respect the reasons to do so and it would be unfortunate to waste their benefit by looping past them.

My test loop was made using spare guitar cords and aligator clips. I just connected up with the outer braids. Normally this is the screen for the inner wire, but it this use, its effectively just acting as a bare wire. The resistance of the wires is very low. I would not trust my meter to measure them. But the induced voltages due to 'hum' currents should be approximately independent of this since lower resistance with a given flux would give higher currents.

I might try with a much finer wire.

The 68k represents a guitar with its volume pot slightly reduced. Clearly directly connected pups would be much less, but this is based on the thought that the lower this value, the less hum I would get so a higher value is a better test. It could be worth trying with much lower or zero however.

Thanks for commenting

John

[sumgai]

John

It seems to me that if you're trying to duplicate the conditions under which a ground loop might become harmful within a guitar, then you should use the same kinds of material, no? And to my feeble mind, if you insert any more resistance/reactance than the minimum found within that same guitar, then you risk reducing the percentage of difference between the two portions of the loop. The less resistance overall, the greater the difference will be, and thus, probably more observable.

(QED: 68KΩ + 0.01Ω versus 68KΩ + 0.1Ω ...... that would take a helluva sensitive meter to spot a difference, eh?)

Not to mention, in the average guitar cord, whether it be the center conductor or the shield, the resistance per foot is on the order of less than 1/100th of an ohm (assuming 20 ga. stranded), but when you get down to 22 or 24 gauge solid hookup wire, the ohms-per-foot increases about 5 or 6 times. Again, making it easier to spot a difference, I should think.

But even after all of that, there's still the need to determine the actual hum perceived, not just a resistance difference, so I still like the way you're thinking. Let's hope your giant loop doesn't become a 1st order antenna for spurious powerline emissions in your neighborhood!


sumgai

[JohnH]

Further on this

I tried it with a loop made of two different several-foot lengths of a fine insulated hook-up wire, like I use within a guitar.

It makes no difference, and also no difference with large, small or zero resistances to the jack tip in that circuit. No hum from my ground loops!

John

[UnklMickey]

Ground Loops are a misnomer anyway. It's not about having a loop, per se.

It's about having current flowing through the wires, and the resistance of the wiring causing voltage drops. Then, the farther away you are from the true ground, the more the reference point is "modulated" by the current. How much current is there in a guitar. Almost nothing. How much resistance? Not much. So any point in the ground circuit, will be fine, unless there is a resistance introduced in the path to true ground. Sometimes this happens because of a poorly soldered joint somewhere, but that's rare.

Still... Good wiring practices are.

If doesn't take more effort to overkill, then do so. If it does, Keep It Simple, Stupid.

[JohnH]

What I think this means is that I'll carry on wiring as I do, but I won't be looking for invidious ground loops to explain unwanted noises, nor will I be ripping up perfectly good wiring on existing guitars that may have more than one ground route in the cavity.

John

[rumblejohn]

I am going to have to follow this thread. Last year I shielded and star grounded my Fender Jazz. I didn't cut any corners or skrimp, and it was significantly quieter, you have to lean it against the beer sign in the window to get it to hum. I didn't do the job in steps, so I don't know what had the most affect.

John

[warmstrat]

As I reported in this thread, long ago...

www.guitarnuts2.proboards45.com/v45index.cgi?board=wiring&action=display&thread=1598&page=3#15240

Removing the ground loops from my Squire Strat not only SIGNIFICANTLY reduced hum, but also cleared up the actual sound ("tone") of the guitar when plugged in. I have no recordings from when the guitar had ground loops, so I can't prove it, but it will take me a long time to forget that rushing sensation, only describable as the sudden realisation that what you thought was a bit of a crappy old guitar is actually a beautiful sounding old instrument.

(Apologies for my horrid writing style back there - I'm mortified!)

[D2o]

Apr 13, 2008 10:25:09 GMT -5 warmstrat said:

As I reported in this thread, long ago...

www.guitarnuts2.proboards45.com/v45index.cgi?board=wiring&action=display&thread=1598&page=3#15240

Removing the ground loops from my Squire Strat not only SIGNIFICANTLY reduced hum, but also cleared up the actual sound ("tone") of the guitar when plugged in. I have no recordings from when the guitar had ground loops, so I can't prove it, but it will take me a long time to forget that rushing sensation, only describable as the sudden realisation that what you thought was a bit of a crappy old guitar is actually a beautiful sounding old instrument.

(Apologies for my horrid writing style back there - I'm mortified!)

As you reported in that thread, long ago, in which you wrote

Well... i opened it up, measured the pots (they seem to be 500K log, which surprised me), and removed all the ground loops while i was at it. (i also switched the mid tone pot to work with the neck PU, but thats a minor detail) To cut a long story short, it's improved my sound VASTLY in terms of clarity and hum reduction, and i can't wait to do phase two of shielding it and modding my switching system. Hurrah for guitarnuts so far, this is fantastic!

I find this interesting:

and i can't wait to do phase two of shielding it

I shielded a friend's squire strat, and forgot to remove the wires from the backs of the pots, which created a ground loop.

What is funny is that when I realized my mistake and removed the ground loops, it made absolutely no difference.

Maybe that is the difference between your scenario and mine - I had already shielded the guitar, you had not.

Any ideas?

[mr_sooty]

So is it likely that the hum reduction acheived by the QTB Mod is down to the sheilding, and little to do with the star grounding? Because it certainly would be easier to just sheild and not worry about all that other stuff. Would we get the same affect? Because I know my '57 Strat is significantly quieter since I did the QTB mod.

Oh, and I don't really care about the shock protection capacitor in the QTB. If I can get rid of the noise just by sheilding, then that's all I'll do!

[sumgai]

sooty,

A ground loop is merely a secondary path for the signal return. When two or more paths exist, any difference in the resistance of those paths will create a voltage difference, which will manifest itself on the signal. That voltage offset is sometimes able to cause noise in the signal, although it's usually too low to be audible.

I'm with John, I don't worry about secondary paths, they're too short to cause any voltage differences that might be troublesome. So like you've described, my guitar has no extra star ground, nor do I use the safety cap. Like ChrisK, I use the safest cable in the world, made by RF!

You have my blessings to do only the QtB mod, and nothing beyond that - you won't be sorry.

HTH



sumgai

[mr_sooty]

Apr 15, 2008 2:13:04 GMT -5 sumgai said:

You have my blessings to do only the QtB mod, and nothing beyond that - you won't be sorry.

But the QTB mod includes the Star Grounding. So what you mean is just sheild it, which is only part of the QTB mod - right? As long as the electronics are grounded to the sheilding (which they should be through the pots or something, shouldn't they?), that's all it really requires right?

[D2o]

But the QTB mod includes the Star Grounding. So what you mean is just sheild it, which is only part of the QTB mod - right?

I know you said "you", meaning sumgai, and I don't mean to be out of place here in answering "You got it!", but I actually wanted to comment on your next question.

As long as the electronics are grounded to the sheilding (which they should be through the pots or something, shouldn't they?), that's all it really requires right?

Mainly Yes, and a little no, in the form of a detail about the actual purpose of the shielding.

So: Yes, you achieve grounding by contacting the pots with the shielding ...

... BUT, you also must make sure that the shielding is done in such a way as to achieve electrical and magnetic continuity.

For example, if you use copper (preferable) or aluminum tape with non-conductive adhesive, and you sloppily overlap the tape so that you end up with adhesive from one strip contacting the aluminum from another strip, you will not have achieved (the full effect of) shielding. You would need to fold under the edge of one piece so that the aluminum side which you have just folded under contacts the aluminum from the other piece. See?

I will even go so far as to throw this out there: If you were to totally ignore the QTB instructions, and just properly shield your guitar, leaving the wiring bone stock, including wires between the backs of pots, you would have less hum than without the shielding, even though you now have ground loops (because of the grounding on the shielding chassis as well as the grounding via the wires between the backs of the pots).

Some guitar techs purposely include these redundant loops as a safeguard, in case a pot becomes loose without the owner noticing (apparently this is more likely to happen with non-guitarnuts than with guitarnuts ).

Since I have now used my daily quota of italics and underlining, I will conclude here.

[newey]

Sooty:

As long as the electronics are grounded to the sheilding (which they should be through the pots or something, shouldn't they?), that's all it really requires right?

The original QtB procedure uses a ring connector around the shaft of one of the pots so as to insure that the shield is grounded. The connector is then held in place when you tighten the pot down onto the pickguard.

That method is much more likely to give you a proper ground than relying on the shaft/body of the pot to make contact with your shielding. Lots of pots have plastic shafts anyway these days, some are mostly plastic, and even if yours are all metal you still want to insure a solid connection to ground for your shielding.

Your shielding just needs to be grounded somewhere, my preference is to run a ground wire with a small ring connector and attach it to the side of the cavity with a small screw, thereby contacting the shielding. You can certainly use other options.

[ChrisK]

Actually, the shafts have nothing to do with the conductivity of the pot back shell to the pick guard shield. Aside from the inside of the knobs, they come in contact with little.

It's the shaft bushing, connected to the back shell that makes the path.

In a passive electric guitar, the presence of gloops are fairly meaningless.

We all have heard that these are bad. They are, when they matter.

IMHO, the preferred method (at least by me 'cuz I use it) is to make the "0" terminal of the volume pot the system ground. I connect everything to it. This is easier if the terminal is soldered to the back shell. Since most pickups and pots have 'tousands of them Ohms (well, tens when fully CW), a few ""Ohms on the drains" won't matter.

One does have to mind the ground separation created with the separate safety cap if it is used.

[mr_sooty]

Which terminal is the Zero one? Would it be good to connect a wire to that terminal, then to the sheild via a ring terminal or similar screwed to the body cavity on top of the sheilding?

[ChrisK]

The one with the lowest resistance to the wiper terminal when the pot is at "0".

[mr_sooty]

So in this diagram, I would assume the outside terminal that's connected back to the volume pot would be the zero one, right? Also, there's already a wire running off the pot to a lug (part #19), so if this lug was screwed to the sheilded body cavity, would that be sufficeint to ground the sheilding? Or would it be better to move this wire from the pot to the zero terminal?

www.fender.com/support/diagrams/pdf_temp1/stratocaster/0107400_02C/SD0107400_02CPg2.pdf

[ChrisK]

So in this diagram, I would assume the outside terminal that's connected back to the volume pot would be the zero one, right?

Yep.

Also, there's already a wire running off the pot to a lug (part #19), so if this lug was screwed to the shielded body cavity, would that be sufficient to ground the shielding?

Yep.

[Or would it be better to move this wire from the pot to the zero terminal?

Nope.

In a passive electric guitar, the presence of gloops are fairly meaningless.

It makes no difference, which is why

I connect everything to it. This is easier if the terminal is soldered to the back shell.

I consider this to be good enough grounding practice. What I find problematic is the practice of using the shielding as the ground point, depending on conductivity throughout it for signal integrity. This isn't a ground loop, it's a bad ground and practice.

The best way to avoid ground loops or bad grounds is to use a star topology. With everything returned to this point, there can be no interaction of individual ground currents between other ground segments (which is what a ground loop is). By following this method, bad ground connections are easier to find, and to avoid.

If guitars were all that sensitive to ground loops, the method of using a stereo jack to switch the battery on in an active guitar (where the operational supply current flows thru the signal ground) would have fallen on its face once tried.

[mr_sooty]

Apr 20, 2008 23:31:31 GMT -5 ChrisK said:

So in this diagram, I would assume the outside terminal that's connected back to the volume pot would be the zero one, right?

Yep.


Yep.


Nope.

So all I really need to do (once my new '08 American Standard arrives this week!) is sheild it and screw that lug back to the sheilding. No need to use the soldering iron at all (apart from desoldering an resoldering the output jack to remove the pickguard), right? This would be enough to acheive the same sort of hum reduction I got on my '57 reissue when I did the full sheild and star ground , am I right?

[ChrisK]

So all I really need to do (once my new '08 American Standard arrives this week!) is shield it and screw that lug back to the shielding.

Since that lug exists in the first place, the guitar is already shielded.

No need to use the soldering iron at all (apart from desoldering an resoldering the output jack to remove the pickguard), right?

Don't know, it depends on what your (extra) shielding entails.

This would be enough to achieve the same sort of hum reduction I got on my '57 reissue when I did the full shield and star ground , am I right?

I don't know/can't say.

It depends on what a "full shield" is.

Since the ground point in these guitars is the back of the volume pot (with all other circuits connected directly and singularly to it), it's already a star ground (the theme herein).

[D2o]

This is not related to Chris' post above ... just an observation, FYI, FWIW (I better stop right there with the "F" words)

I just replaced the 3 -way switch in my guitar (SS, 3-way, 1V, 1T) and decided to add shielding:

- the shield has continuity
- pickup signal returns are soldered to the back of the volume pot
- a wire connects the back of the volume and tone pot
- a wire runs from the ground of the switch to the back of the volume pot

Despite redundant ground ("loops") there is NO additional hum.

D2o

[ChrisK]

Ground loops within a passive guitar are generally a misnomer.

The only time that a ground loop has any effect is if significant current passes thru the loop such that it generates an offset or error voltage that causes errors or noise relative to the signal of interest.

Most currents within a passive guitar are quite small and any internal ground loops are small as well.

A ground loop might also act as an antenna for EMI.


If star grounding practices are followed, there usually will not be issues from grounding (shielding is another topic).