Shielding a Semi- Hollowbody For HBD Wiring

[sonosonny]

I am planning an HBD / PRails wiring project on a Semi-Hollowbody. My initial concern is the shielding of ALL wiring. Plans are to use copper braid shielded 4 and/or 2 and/or 1 conductor wire (whatever requirements are needed) for ALL connections to the 4PDT switches, P/P pots, 3 way Pup Selector, and to the output jack. I am going to use Fralin 4 conductor braid shield wire and some 2 and single conductor braid shield from my in house stock. With all wiring being shielded and of course grounded would this be the absolute best way to shield the Semi?

Rune had suggested I encapsulate the entire harness in very fine copper screen and ground it thus creating a grounding pocket environment similar to a Str@t pickguard/wiring cavity being fully lined with foil copper tape.... I was balking at this due to grounded screening accidentally touching 'hot' on one of the 4PDT/PP switches = short. Would be a bad dream to get this harness into the Semi and have a ground screen to hot short issue, take it back out , re-test, fix short/tape encapsulate the 4PDT/PP switches, etc, reinstall.

I have worked on a number of Semi's (challenging, but rewarding) and have confidence the HBD can be shoe horned into the wiring cavity with pre wiring on a cardboard/fiber board pattern/trmplate with the thought of how the harness with be routed into the "F" holes. I usually use the string method to pull the pots and switches into the top holes.

I plan on documenting this entire process for everyone and posting on Gnutz2.

So my questions to all Nutz are:
1. Fully shielded wire throughout Semi-HB would yield acceptable/quietest shielding for the single coils per HBD switching options.
2. Are the exposed lugs on the 4PDT's,P/P DPDT's, 3 way PUP selector susceptible to introducing unwanted noise? Solution ?
3. Should jumpers on switches be shielded as well or = overkill?
4. Are there any better/easier ways to shield a Semi-Hollow than my proposed plan?

Thanks to all for any input/suggestions/advice.

Regards,

Son-O-Sonny
(Runewalker consultant /spiritual advisor)

[ashcatlt]

Wouldn't it have been nice if they'd sprayed the insides with conductive paint for you before assembly?

Ah well. It is true that any break in the shield is a place for noise to enter the system. Yes, this includes switch and pot lugs. I'd venture to say that a short jumper between two exposed terminals won't make things much worse. Remember also that the switch frame and pot bodies themselves act as (at least partial) shields for the guts of the components.

I think it would be possible to shield those lugs by just wrapping the thing in electrical tape and then like a ball of foil over that. Doesn't add much size to a pot or switch, and still allows you to fish each one individually.

[newey]

I hate to disagree with RW, but I see the copper mesh idea as a non-starter, for the same reason mentioned by Son-o: How you gonna get it through the f-holes, without twisting, tangling and potentially shorting stuff out?

Ash is right, any component in the signal chain could, in theory, be responsible for introducing some noise. But I wouldn't worry about shielding switches and pots; your plan to use shielded cabling throughout ought to be enough. Longer runs of cable are much more likely to be "dialin' in that noise" than a lug on a pot.

The pickups are really where the (noise) action is, so shielding around the back of the pickups might yield a greater benefit than all the shielded cable. But, as we've discussed before, wrapping the pickup too closely risks altering the tonal characteristics of the pickup (and, probably not in ways that you'd like). You might, though, try a copper mesh "box" that would mount behind the pickup, if you can get it and the pickup to drop in through the hole. I envision this sitting well off the back and sides of the pickup.

And, please do document the build for us. The whole string method thing has intrigued me for years, never done it myself or seen it done.

[stevewf]

So now I'm also wondering if I could improve the shielding in my hollow-body guitar. It's built from the "Rockabilly: kit by TheFretWire. The controls can go in pretty easily, via either of the HB-sized pickup holes, as the body's hollow except for a pillar under the bridge - doesn't have that long center block of a semi-hollow.

However, the body arrives completed and unshielded, so there's a risk of noise that comes from external sources other than my fingers and pick. I've taken some steps to shield the signal wires, but I'm sure I could improve it, having now read some QTB and other Nutzy posts.

I gotta say, this guitar is surprisingly quiet already, but heck, let's see if I can get a hollow-body to be the hum-free envy of players with their lead-lined [control] cavities  ! Oh yeah, while minimizing the sacrifice of high frequency tones  .

Here's more about that signal chain. I've managed to make a pretty wire-conserving layout, meaning there's little back&forth inside the guitar's body. Most easily described by pictures. Meet the Red-Tailed Hawk:



(I sure hope that picture works; it's just a placeholder icon while I'm typing this. Here's a Flickr hyperlink, in case).
Most of the wire is single-conductor shielded, with the shielding braid exposed. The Neck-mounted P-90 (HB-sized) came with a single conductor, shielded, jacketed wire (black). "Jacketed" means the shielding is there, but it's not visible as it's under an insulating layer (plastic). The Bridge humbucker has a 4-conductor shielded jacketed wire (also black - sorry!). The bridge wire (single, insulated wire) has a quick-disconnect terminal that connects to its counterpart, soldered to the shielding of one of the single-conductor wires.

I definitely see some possible improvements, inspired by my recent Nutzy readings:

1) I could consider separating the shielding almost entirely from the signal return, per Nutz consensus. If I understand correctly, that means that the single-conductor wire ought to be replaced by 2-conductor. For this discussion, the white conductor will carry the "hot" signal; the black conductor will serve for the "ground" (signal return); the braid will be part of the shielding. The place where shielding and signal return are joined could be where the black conductor is connected to the output jack; aside from that, they never join.

This improvement would also include redoing the leads on the caps on the TCut and BCut controls, i.e. unsoldering the leg from the pot chassis (that's shielding, after all) and instead soldering it to the signal return path. There's probably a similar issue to be solved at the volume control, and at the rotary switch... really, wherever a signal has a connection to ground, I'm learning.

2) I could establish a physical layout that's mindful against building loops. Applies to both signal return and shielding. As for star, in this case I'd propose a hybrid line-segment & star layout, mostly because of the physical layout of the controls. I chose to put controls for Master Vol, Master TCut and Master BCut control on the bottom half of the guitar. That makes it easy for the signal's physical path to be more or less linear there, in fact all the way to the pickup selector in the body's upper bout. So the signal, signal return and shielding can all follow this path, where they can split into a star for the rotary and pickups. The bridge ground wire joins the braid along the line, as the exception that proves the rule   The shielding layout simplicity was not by accident; I was very happy to find this combination of controls and layout, given the need to shield up.

3) I'll need to pay attention to how the conductors leave the braid, and how to minimize their unshielded exposure while also ensuring that the braid can connect to control chassis where needed. Any tips? How do I even say the above better? "Conductor leads entering/exiting braided shielding"?

4) I could improve the shielding gloves that I've put over the controls (photo). Essentially, on each control, I electrical-taped over the leads and jumpers for insulation from the shield; then I used copper tape with conductive adhesive to wrap - almost completely - the control. The tape is stuck to the wire braiding and to the control chassis. Even the holes in the chassis are covered with shielding this way. Quick check with multimeter shows connectivity throughout. So, why improve? Well, it's a real pain to peel in case of update/repair. I have some ideas to make it easier to open/close as needed. I'll get too wordy here; when I actually redo it, I'll post. Note: I already had to redo the rotary switch, when I noticed that it wouldn't turn; I'd taped the wafer in place with the electrical tape. I had to rip off the shielding and redo it, which I did hastily. But, anyone see problems resulting from encasing the control like this?

5) Eddy currents... um, those are magnetic eddy currents, right? So in a discussion about shielding, eddy currents shouldn't enter except where shielding of the pickups is in question, like in this PG post). I ask this question so that I can feel comfortable when I more or less completely enclose my signal chain (and, incidentally, signal return) in braid & foil without gaps, except around the pickups. Anyone confirm?

6) Two conductor... sure?  Do I really need 2 conductors inside the braid at all points? Do I really need 2 conductors at all? Can it not be done well, using the braid as both shield and signal return? I have been reading Nutzy posts, and I recall that it's the unreliability of the shielding that makes it undesirable to use for signal return. Maybe I need to re-read, but does the fact that all joints of this 1-conductor shielded line -- signal and shielding alike -- are soldered (except that bridge wire) affect the answer?

6a) Bonus question: is the bridge/trem/string ground considered as part of the shielding? In terms of what part of the guitar's circuit the bridge/trem/strings wire should be attached to. Does it incorporate (literally) the player into the shielding scheme, is it a grounding thing, a safety thing?

7) Shielding the pickups. That P-90 in the neck has a chrome-plated ring around it, which is soldered at two points directly to the humbucker-sized base. The ring shields the sides of the aggregate HB, and not the top. Earlier, I wrote "except the pickups", but are there easy ways to improve its shielding without sacrificing high frequencies? I don't relish the idea of unsoldering the cover, as it seems to take lots of time to melt the solder enough (=cooking my pickup & burning my little fingers... hey! I play guitar with those!) Tips, anyone? If I did take off that cover, I could try to put on a proper 2-conductor wire (unless we've decided the guitar doesn't need it? Paragraph 6).

Bridge: I did unsolder the 'Tron-style cover from the humbucker (that's how I know about burnt fingers). The pickup cam with a 1-conductor shielded wire, but I really wanted four wires there for my sound scheme, so I uncovered, disassembled, re-leaded, reassembled and re-covered it. It should be easier to remove the cover now, as I used wire rather than simple solder blobs. This bridge pickup's shielding, too, is up for Nutzy consideration. Again, it has a HB-sized Filter'Tron styled, chromed-plated cover.

The chrome is not absolutely vital, but the rest of the guitar's trim is black & chrome, so keeping the chrome is desired. Mentioned here in context of chrome's potential effect on high-freq tones.

Plenty of room in the cavity, except where the neck pickup is right next to the set-in neck heel. Pet peeve against internal wires being visible to the audience.

[Yogi B]

Firstly, sorry for the delayed reply. Secondly, I've grouped my responses out of order, but in a way that's hopefully logical...

Oct 20, 2021 18:41:47 GMT -5 stevewf said:

1) I could consider separating the shielding almost entirely from the signal return, per Nutz consensus.
...
6) Two conductor... sure? Do I really need 2 conductors inside the braid at all points? Do I really need 2 conductors at all? I have been reading Nutzy posts, and I recall that it's the unreliability of the shielding that makes it undesirable to use for signal return. Maybe I need to re-read, but does the fact that all joints of this 1-conductor shielded line -- signal and shielding alike -- are soldered (except that bridge wire) affect the answer?

From what I also remember, your recollection is correct — at least personally I wouldn't want the integrity of the signal to rely upon possibly less than ideal connections between shielding.

As for joining or separating the signal return and shield: after traversing through the jack socket, the cable on the other side has combined signal return & shielding and that isn't a problem that people seem to care too much about. If they really did care, it wouldn't be difficult to make the 'star point' (where shielding & signal return meet) at the other (far) end of the cable, (or even, the two could never directly join if the signal is fed into a differential preamp). And since TS/TRS jacks/cables are the way they are, using a standard mono cable would just join the shield & signal return at the jack as per current practices (so lack of backwards compatibility with normal cables isn't a valid excuse for why this hasn't been done, there must be another reason...).

6a) Bonus question: is the bridge/trem/string ground considered as part of the shielding? In terms of what part of the guitar's circuit the bridge/trem/strings wire should be attached to. Does it incorporate (literally) the player into the shielding scheme, is it a grounding thing, a safety thing?

Though it is done for noise reduction reasons, it's not shielding per se. What it does is ensure that the player is at the voltage that is 'considered' to be ground (and thus silent) by the guitar.

It can be the exact opposite of "a safety thing" — potentially fatal — if your audio equipment or venue wiring is faulty. Even if it's not your guitar gear that's faulty thus you are actually grounded when touching the strings (or other metal parts), being grounded is a potentially dangerous. Say you also come into contact with something that is at substantially different voltage, say a microphone that's at mains voltage — now you're part of a completed circuit between that mic and ground, not good.

2) I could establish a physical layout that's mindful against building loops.

I'd say it's pretty hard to accidentally build a (wire) loop — you'd have to solder a completely unnecessary wire somewhere. Typically nobody cares that shielding tends to have lots of loops in it though...

...unless those loops are close enough to the pickups to start robbing tone. And On that note...

5) Eddy currents... um, those are magnetic eddy currents, right? So in a discussion about shielding, eddy currents shouldn't enter except where shielding of the pickups is in question

Correct, though side note: "magnetic currents" don't exist, or rather only exist in theory, but that's a technicality which is not strictly relevant. Aside from that limitation, magnetism & electricity are essentially 'orthogonal manifestations of the same thing' — this can be of benefit to us (magnetic guitar pickups wouldn't work at all if this weren't the case); but it can also be a hindrance, as in the case of eddy currents.

A voltage is induced across any conductor that is subjected to a varying magnetic field, additionally any (electric) current is also accompanied by a magnetic field. Those two things can happen simultaneously. A varying magnetic field (e.g. a moving magnetised string) can induce a voltage (across the coils of a pickup, or it's cover) which drives a current and thus creates its own (opposing) magnetic field.

For eddy currents to be worth worrying about there are two conditions that need to be satisfied: firstly, you need a substantively varying magnetic field to induce high enough voltages and cause sizable eddy currents; secondly, the opposing magnetic field caused by those eddy currents must be able to easily interact with the circuit in some way. Therefore in (most) guitars this is only really an issue local to the pickups, where both of the proceeding are true. Elsewhere in the wiring the current is so low (thus the strength of any accompanying magnetic field is similarly low) and the other components are low enough inductance (thus insensitive to changes in magnetic field) that there's no worry.

3) I'll need to pay attention to how the conductors leave the braid, and how to minimize their unshielded exposure
...
4) I could improve the shielding gloves that I've put over the controls

I'd say they're the same question, if your "glove" extends far enough then the inner wires won't leave the braid until they've entered the "glove".

Back before the mid-sixties Gibson used shielding "cans" to surround the pots on their (semi-)hollow guitars:

(Note that only the pots are shielded and even then only 3/4 of them, on some models there simply wasn't room for the can to placed around the final pot when situated within the body of the guitar.) I don't know of anything similar since then — as far as I know nobody's making reproductions, not even for Gibson's own reissues. So if you wanted a 'proper' solution be prepared to pay silly 1960s Gibson money.

As a DIY solution: small screw-top aluminium containers exist (example below), in which you could drill a hole in the base for the pot's/switch's shaft and one (or more) into the side for wires, although finding some that are of an appropriate size may prove tricky. 40mm diameter × 18mm height should be enough to comfortably fit your tone controls, but for the push/pull & the switches you probably need at least 30mm height.

7) Shielding the pickups. That P-90 in the neck has a chrome-plated ring around it, which is soldered at two points directly to the humbucker-sized base. The ring shields the sides of the aggregate HB, and not the top. Earlier, I wrote "except the pickups", but are there easy ways to improve its shielding without sacrificing high frequencies? I don't relish the idea of unsoldering the cover, as it seems to take lots of time to melt the solder enough (=cooking my pickup & burning my little fingers... hey! I play guitar with those!) Tips, anyone?

Since solder is relatively soft, you can usually cut through it with a craft knife or razor blade, obviously being careful — you don't want burnt and sliced fingers (or sliced wires).

Because your pickups have steel polepieces even full (nickel-silver) cover would have a lesser impact upon the resonant peak (compared to what it would upon typical single coils with AlNiCo rods), because there'll be a greater intrinsic loss of treble due to the larger eddy currents within the polepieces themselves. Nickel-silver being preferable as a cover material (as opposed to brass) because it has higher resistivity — thus, for the same induced voltage, the eddy currents will be smaller. In either case Filtertron style H-gate covers do a good job at reducing eddy currents when surrounding humbuckers (though why the standard has become that the gap on the humbucker sized versions is unnecessarily vastly wider than an actual Filtertron cover, who knows?).

Things are a little trickier for the P90, again you could probably get away with using a full nickel-silver cover, but ideally you'd want a break in continuity around the circumference of the cover like the tele neck covers as in stratotarts's Undercover Pickups article. That being said there'll always be continuity because of the attachment of the cover to the baseplate, so there'd be no benefit slotting all the way to the bottom — though: the closer to the base of the pickup the remaining path is, the lesser the eddy currents will be.

[stevewf]

Nice info, Yogi B, thanks!

My guitar has one of those hum buckers with too much gap in the Fak0'Tron cover. Again, it's the pickup that came with the kit, except that I four-wired it. Partly for $ and partly for fun, I wanted to keep it. Even though the solder was soft, jamming a hobby knife in there repeatedly was no picnic, and I was nervous about cutting a coil wire... and fingie-poos.

Interesting photo of '60s Gibson guts. Huh, they "can" (some of) the pots but use unshielded wires to the pickup... that is the pickup, yes, the squarish box?.

Given that the Gibson is a hollow-body (i.e without a shielded cavity), my question is: Does it make much difference whether the pickup leads are shielded or not? The classic Strat has unshielded wires -- but they're often twisted plus shielded in one direction by the pick guard. Should we shield our otherwise unshielded pickup leads?

[newey]

Nov 9, 2021 21:58:07 GMT -5 stevewf said:

Should we shield our otherwise unshielded pickup leads?

It's kind of one of those "diminishing returns" questions. The long lengths of wire in the pickup windings will be the largest contribution to noise, so we've always said here that shielding the pickup cavity is more crucial than the control cavity, where the short wire runs (and the pots and switches themselves) are unlikely to be a big source of noise. The pickup wires are longer runs, and therefore have more potential to introduce noise. And, since it is hard to test to see exactly where any noise is entering the system, best to err on the side of more shielding rather than less. If the guitar is going to be torn apart for rewiring anyway, whether it's a solid or hollow body, might as well shield everything you can rather than do half the job, and then find you have to go back in later.

Put more simply, it might be overkill, but we can't guarantee that it's overkill- so best to do it.

[stevewf]

This, by the way, was the inspiration for the shielding "glove" "sock": a chocolate truffle wrapper

After all, it's a flexible, conductive film of aluminum, a decent shielding material, insulated on one side.
In the end, electrical tape plus copper tape won out because it's easier to train into odd shapes.