Trying to understand grounding / shielding

[moosie]

Hi folks. My first post here. I've read a lot of stuff here and elsewhere about grounding and/or shielding guitars (including QTB), and am still confused.

I'm not trivializing the shock hazard, but would like to separate the issues of shock protection and making the instrument quiet, if possible. Initially I'm after quiet, playing only in my house using a three-prong outlet.

My partial understanding goes like this:

The guitar (a tremolo Strat) can be quieted by essentially constructing a Faraday cage around the electronics. Using conductive-backed copper tape, I'll completely cover the pickguard, and the control and pickup routs. The cavity foil will overlap the face of the body, mating with the pickguard foil when screwed down.

So, a copper box.

Then, conceptually extend the cage to the output jack via the braided shield of a multi-conductor wire connecting pot to jack, but not necessarily needing to foil-line the jack rout.

What next? I'll make sure all the copper connects to itself. But I've still got the trem wire, the braided shield from the output jack, pot shells, and "signal return".

Is ALL of this considered one common ground, aka signal return? Seems so, otherwise we'd have to insulate the pots where they touch the pickguard...

What should connect to what, and what should NOT touch the shielding?

After reading John H's post about guitar ground loops having negligible effect, and witnessing Fender continue this "bad" practice even in Custom Shop guitars, maybe I don't have to tear my hair out over ground loops...?

All else equal, I like the idea of soldering the pot shell as a more reliable connection (over time and oxidation) than a lug on the bushing.

So, I'm probably leaning to Fender's vol pot mostly-star-ground. Problems? Trem wire ... where? any convenient place, like a lug screwed through foil into wood? Output jack shield connects to back of vol pot (which is essentially the same as connecting it to the signal return wire inside the same cable, yes? 'Cause they both connect to the ground side of the pot. Right?

I apologize if this has been hashed over too many times. I'm a newbie to all this, and have done my reading, but just need a little guidance to "get it". Any help is appreciated. Thank you.

[newey]

moosie-

Hello and Welcome to G-Nutz2!

You've got the basics of the shielding scheme down, minus the "blocking capacitor". I haven't used the capacitor in any of my builds, as the risk is minimal if one uses SS amps and plays only at home, through a known-good 3-prong outlet.

Using the capacitor forces one to do a "star-grounding" scheme; without the cap you can certainly use the back of the volume pot as a grounding point. Don't sweat the "ground loops".

Is ALL of this considered one common ground, aka signal return? Seems so, otherwise we'd have to insulate the pots where they touch the pickguard...

The shielding and the signal have to run together at some point, since your amp only has two connections going into it. The idea is that the point where the two connect should be as close to the output jack negative as possible, either to a star-grounding point which is then connected straight to the output negative, or to the back of a pot which is likewise connected to the output negative.

You are correct that both the cable shield and the output jack neg would go to the back of the pot, if that's what you use as the grounding point.

Even without the capacitor, I usually use a star ground point, just because it's difficult to get solder to stick to the back of a pot without risking too much heat, which can damage the pot. If the pot has been previously soldered onto, it is easier.

But I find that a washer, with a screw holding it to the cavity shielding on the side of the cavity, works pretty well.

You are correct that shielding the output jack cavity is unnecessary. I've had problems the two times I tried it, because there is very little clearance, and the layer of foil reduces it further. It's easy to have the tip of the plug touch the shielding, shorting the output, unless everything is aligned perfectly.

Trem wire ... where? any convenient place, like a lug screwed through foil into wood?

Yes. When you check all the shielding for continuity to ground, also check this connection.

You mentioned "insulating the pot shells" from the shielding. With the star-grounding scheme (whether one uses a washer over a pot or a screw into the cavity), you want the pot shells to be grounded to the shielding.

If you're using the back of a pot as a grounding point, and the other pots are connected to that point, then having the pot shells contact the shielding does result in added ground loops. These shouldn't matter, as you have read elsewhere here.

But if you want to avoid those, the best way is to just trim the foil back a bit on the pickguard where the pot shells would contact the pickguard. The slight holes in your Faraday cage probably won't make any difference to the efficacy of your shielding job.

However, if you do this, then a separate ground would need to be run from the cavity shielding to the back of the pot, since the cavity wouldn't be grounded through the pot shell/shaft.

[moosie]

Hi newey,

Thanks for the response. Let's see if I understand...

Even without the capacitor, I usually use a star ground point, just because it's difficult to get solder to stick to the back of a pot without risking too much heat, which can damage the pot.

But I find that a washer, with a screw holding it to the cavity shielding on the side of the cavity, works pretty well.

One of my goals was to avoid soldering all those wires (and subsequent modifications) to the pot for the reasons you state.

So, in this scenario, you collect (and solder) ALL the ground wires on this washer, and then "going out the other side" of the washer is the single wire to output jack. Yes? In fact, this washer could be the terminus of the trem ground wire, too, right?

With the star-grounding scheme (whether one uses a washer over a pot or a screw into the cavity), you want the pot shells to be grounded to the shielding.

To the shielding...? Or to the separate star ground washer? Same-ish difference, but seems like the goal is to bring ALL commons directly to the star, by their own routes, right?

And if the goal is to have the shield join the common as late as possible (which answers a lot of my confusion, BTW), this should be the only place the shield touches the circuitry, including pot shells. Yes? So I could trim the foil at the pots not for ground loop, but to achieve the shield only connecting the circuit at the latest point. ??

When you check all the shielding for continuity to ground, also check this connection.

Uh, yeah. How do I do that again? My thinking hadn't got that far. I do have a standard digital multimeter, but not a lot of experience with it. Set to lowest ohms setting, and see if there's any tiny value between two points of interest?

Am I getting warmer or colder?

Thanks for your help.
-joe

[newey]

So, in this scenario, you collect (and solder) ALL the ground wires on this washer, and then "going out the other side" of the washer is the single wire to output jack. Yes? In fact, this washer could be the terminus of the trem ground wire, too, right?

Yes. And yes.

but seems like the goal is to bring ALL commons directly to the star, by their own routes, right?

Well, not exactly. Every connection is a potential failure point. You don't want to be making more than are necessary. If you're using a star ground (apart from a pot shell, which can also be a star ground) then using the contact between the shield and the pot is fine, since the shielding will be connected to the star point anyway.

So I could trim the foil at the pots not for ground loop, but to achieve the shield only connecting the circuit at the latest point. ??

Don't over-think all this. Again, the ground loop probably won't matter. And, if you're using the back of a pot as your grounding point, then the shield making its connection to the shaft of the pot is, effectively, the "latest point".

I do have a standard digital multimeter, but not a lot of experience with it. Set to lowest ohms setting, and see if there's any tiny value between two points of interest?

Yes. You should have no more than a few ohms resistance between any two points on your shielding, or between the shielding and the jack negative, or any other ground point.

[moosie]

Thanks again, newey! I think I've got it!

Preparing to build JohnH's SP Strat hss circuit.

guitarnuts2.proboards.com/index.cgi?board=schem&action=display&thread=5624

[newey]

JohnH's Strat SP is a great design. Please start a separate thread here for any wiring/building issues you encounter.

And please let us know your results and impressions. So far as we know, treguiars is the sole build of that design, so more data will always be welcome!

[ashcatlt]

I really only skimmed this, but it seems like you've got some confusion re:the pot shells. You're aware that the shell of the pot isn't actually connected to its internals, no? It doesn't carry signal unless you force it to - like by bending one of the lugs over and slodering it to the case. Otherwise, it is a shield just like the copper foil.

[JFrankParnell]

Jun 23, 2012 20:45:05 GMT -5 newey said:

moosie-

Hello and Welcome to G-Nutz2!

Yeah, hey man

Jun 23, 2012 20:45:05 GMT -5 newey said:

The slight holes in your Faraday cage probably won't make any difference to the efficacy of your shielding job.

Wait, is this true? I mean, electrons are skinny, they can get thru. I guess i thought you hadda have a perfect cage.

[newey]

Electrons are skinny enough to fit between the molecules of a copper sheet, too. Luckily for us, our guitars are not likely to encounter any electron beams in most of the places we're playing.

We're trying to shield our guitars from electromagnetic energy which propagates through space as photons, not electrons.

But, when talking about these things, wave-particle duality enters the picture. It's not very helpful in this instance to talk of particles like electrons and photons, the wave model is better conceptually in this circumstance. So, think of waves.

Faraday's original experiments used a wire mesh to construct the cage, not a solid sheet of copper. Waves that would fit between the molecules of the copper are short enough in wavelength to also harm you- microwaves are an example.

Now, there are numerous different wavelengths we want to block- "noise" is, by definition, a bunch of frequencies. Some are shorter, some longer. But none of it is short enough that we need to worry too much about slight gaps in our copper.

And, of course, our Faraday cage can't be perfect since the pickups, the main things that will pick up noise, have to protrude outside of the cage in order to sense the strings.

Standard 60Hz AC hum, for example, has a wavelength of about 18 feet, in a vacuum. Traveling through a medium like air, however, the wavelength shortens, depending on the density of the medium.

Waves with a frequency greater than about 20K Hz can't be heard by us anyway. A wave of 20K Hz is about 17mm long (again, as measured in a vacuum, shorter through the air.)

[ChristoMephisto]

I used some fabric like material from Laird used to block RF signals.
Easy to work with, and creates contact when you layer them.
Just can't solder to it.

When i was shielding my tele and used a brass tack for the 'grounding screw'
it ended up poking a hole through to the back
As a precautionary, you may want to measure your cavity depth and body thickness. You may not even have a problem with it, but better safe than sorry.

[newey]

itended up poking a hole through to the back

Yes, if using a screw with a washer, screw into the side of the cavity to avoid any such problems. The screw doesn't need to be very long, just enough to hold things down.

[moosie]

Wait, is this true? I mean, electrons are skinny, they can get thru.

Thanks for all the input guys (and all the neat Faraday info). I'm redrawing JohnH's Strat SP hss diagram for my use, and trying to incorporate *everything*, like shielding. Even though I'll start a new thread for the mod, I'll post the final diagram here for comments and a more "holistic" review.

I'm redrawing the whole thing because I have a tiny brain, and I learn better this way... Not speedy, but who cares.

[sumgai]

Jun 25, 2012 6:37:05 GMT -5 newey said:

Standard 60Hz AC hum, for example, has a wavelength of about 18 feet, in a vacuum. Traveling through a medium like air, however, the wavelength shortens, depending on the density of the medium.

Waves with a frequency greater than about 20K Hz can't be heard by us anyway. A wave of 20K Hz is about 17mm long (again, as measured in a vacuum, shorter through the air.)

I'm gonna throw ya a lifeline, even if you dinnt aks for it:

wiki.answers.com/Q/What_is_the_wavelength_of_a_60hz_power_line_frequency

Or for more theory behind the scenes:

www.comportco.com/~w5alt/antennas/notes/ant-notes.php?pg=2

(As any HAM operator knows, from having operated (legally) in bands ranging from 160 meters up to 120cm or thereabouts. (I say "up" as in upwards in frequency, even though I'm referring to "bands of frequencies" by the traditional nomenclature, that of their wavelength measurements.))


However, to be fair, your answer is correct for audio phenomena, meaning acoustic pressure/rarification of actual air.


HTH






sumgai,
The Mean Old Curmudgeon

[newey]

Lifeline accepted. I was short by a bit there . . .