Well, today seems to be one of those days for shielding topics. I've usually stayed mostly quiet about such, but I'll chime in here. Those that know me will know to expect a long post.
SoundAsAnOldEngineer, you sound to me like you've dealt with such things before, albeit perhaps in different situations? Especially given your title, it sounds as though you've had some formal training in the field. Not that there aren't probably several other forum members with similar background.
There does seem to be a lot of builds here with fully shielded cavities, most often using copper foil. I agree with Cagey (and others) that for standard-issue single coils, shielding the cavity is an exercise in futility. Because the pickup will happily grab most any RF and marry it to your tone, forever after inseverable. While perfect cavity shielding could theoretically prevent any more interference, it's just a case that the bucket is already full, or if you like the horse is already gone - you've already got as much as it is possible to get, so don't bother. Actually, at that point, and for the same reason, there's not really much point in shielded cable either.
So what about humbuckers or "noiseless single coils" (another flavor of humbuckers). There we may be able to do something. All good humbuckers I've seen have shielded cable. That's a good thing. It'll be grounded at the pickup end, and should be grounded at the business end too. So now we have a nice little conductive grounded sheath for it to live in. Some advocate that the point-to-pint wiring in the cavities should also use shielded wiring. And I agree that'd be good, for the same reason - the shield will absorb the RF, so it never gets to the wire. Naturally there's a non-conductive (insulating) layer between the shield and the wire.
But what if that insulating layer were a bit thicker, like a coax? Well, that'd work the same way, wouldn't it? And could you have two conductors inside the same shield? Sure, why not, so long as they're insulated from each other. And what if you made the insulating layer even thicker? Yeah, sure, we already figgered that'd be all the same thing. What about, like, a lot bigger? Like maybe the size of the control cavity? And now the insulating layer is air? Yes, I mean shielding the cavity. Isn't that electrically exactly the same as a multi-conductor shielded wire? Sure it is, so long as we ground the conductive shield. So shielding the cavity is in a practical sense exactly the same as using shielded-wire point-to-point wiring. Except that the shielded cavity is done once, and then all the internal wiring is simple insulated wire, and a while lot easier to do (especially grounding), and neater too.
So what about those holes that SoundAsAnEngineer brought up? Well, even in shielded wire, there are holes in the shield. Even those that aren't woven must have holes at both ends! And the signal-conducting wire extends beyond the shield at every such end (but only a tiny bit). So.... are holes bad?
Yes, they are. Just as bad as antennae that penetrate the shield. Because they let the noise in. Let me give an example. Every day at work I deal with a large Faraday cage - bigger than a two-car garage. And that thing's gotta be sealed perfectly. The walls, ceiling and floor are copper-clad. The only door is pneumatically closed and has conductive seals. All auxiliary power lines in are conditioned DC. All other signal lines must go through solid-state filters embedded in a special conductive panel that is integral with the shield wall. Anything that could conduct corrupting signal into the cage is relentlessly hunted down and killed (usually not a fun or easy exercise). We do pipe in fluids and gases, and use fiber optics a lot, but all that's OK - no 'lectricity. That sucker is as tight to RF as a bathysphere is to water. And it's very seriously grounded. So what we do inside is done in an absolutely purified electromagnetic environment.
And we also have a couple of holes in the wall, into which are welded short pipes, about 3" in diameter, and through these we can pass various lines (e.g., fiber optic) from an assortment of doodads. Hey! What? Holes in the wall?? Doesn't that defeat the whole purpose of the Faraday cage? Like a screen door in a submarine? (thanks for that one Cagey) No, for the reason that SoundAsAnEngineer mentioned. RF cannot pass through an aperture smaller than half (a quarter, to be safe) of its wavelength. So as he said, a hole of three inches is as good as a steel wall to RF below a Gig.
Go look in your microwave. Don't open it, just look through the window. If you can see in there, then can't the microwaves get out? And if I can't put any metal in there, then how does that light bulb work when the thing's turned on? Because both you and the light bulb are protected from the microwaves by a grid of little holes. The microwaves are really just "too big" to get through the holes. Or the crack around the door. The box is a Faraday cage, but the holes are OK, given that their size matches your cutoff frequency. You can make a Faraday cage from chicken wire. Or screen door mesh, so long as it's conductive. Of course, the cage must be grounded in any case.
Coming back to guitars now, then isn't your typical shielding job plenty good enough? I mean, the pickup (humbucker!) wires and ground wire come through tiny holes. The pot shafts need holes a bit bigger, but not by much. So, all good, right? Not quite. Actually, a lot of cavity shielding jobs I see forget all about the door. I mean the control cavity cover. The little land (recess) the cover sits in often gets little or no shielding. And the back (inside) of the cover, if it gets any shielding at all (duh!), often doesn't extend to the full surface. What you need to make a tight cavity is the entire aperture of the cavity to be completely sealed, conductor-to-conductor, around its entire periphery. Gaps are bad. Close the door!
Hey, wait, wouldn't a gap just be a little crack? Yeah, it would. But electromagnetic waves are not really like the simple single squiggly lines you see in drawings. They don't travel just point-to-point. Well, each one does, but there's a lot of them. They're pervasive ... in other words, everywhere. They may be stronger "from" or "to" a particular direction, but they bounce around and you can often consider them to be omni-directional. And they are not polarized. Which means that they vibrate in any and every possible way, perpendicularly to their direction of travel. What all this means is that if you have a crack in your shield, RF of the right size will pour in. A hole in your shield doesn't have to be round. A shield aperture of any size or shape is equivalent to a hole of a diameter equal to the largest extent of the aperture. (Don't ask me about thin curved apertures - the math is thick). So the contact between your copper-clad cavity cover and your clad cover mounting land must not have any single gap longer than the specified maximum.
So will it help? Not necessarily. I've seen guitars where shielding worked. Or at least, they didn't get RFI. But then, they might not have without their shielding either - dunno. And other guitars, even with shielding that looked done well, that were still susceptible to noise. Finicky things, these axes. There's a lot of variables, as each guitar and rig (and environment) is a little different.
So should you shield? My philosophy is ... if you have humbuckers, go ahead if you like, it can't hurt. And it may help. Actually, it probably will help some, but to what extent can't be predicted. And the benefit will probably vary with the environment. But as I said earlier, it could make wiring and grounding simpler and easier. And it's prettier, if that matters to you.
But if you're gonna go to all the trouble and expense to do it, you may as well do it right, or not bother. Make all your holes small. Make sure that nothing penetrates the cage except the pickup leads and the ground. Actually, you can just connect ground to the outside of the cage without bothering to penetrate it. And the pickup shielding should preferably stop at (and be connected to) the shield. Get the shielding foil right up to the pot holes - heck, cover 'em with foil and poke holes in it after with a pencil. Then screwing down the pots onto the copper grounds 'em. Anything that goes out of the cage (i.e. to jack) must use shielded wire, with its shield grounded at the cage wall. Or if the jack is in a separate cavity, the cage can be built to extend to the output, thus also shielding the jack! And the entire back of the cavity cover and its mounting land must be fully shielded and as much fully in contact as possible. Pretend you're building a fishtank - consider that a good shield would be close to waterproof.
I have one unshielded single-coil guitar. It's a little noisy. Shielding won't help it. I have a couple unshielded humbucker guitars. One's a little noisy. But not enough to motivate me to go to the trouble of shielding in the hopes of making it better. And I do have one build in the works with humbucking pups ("noiseless single coils"), and it's getting shielded. Not because I want to shield the pots and caps, but because it's gonna have a whack of active electronics. And THAT is a grrreat reason to shield. I've had breadboarded boosters and preamps that picked up radio stations. But get that on a circuit board, put it in a cage, and goodbye RFI.
