Targeting inductance to achieve different tones?

[sumitagarwal]

The more I read the more I'm sold on the primary factors in pickups' sound being defined by inductance, geometry, and position (an extension of geometry, I guess).

So for a versatile axe I was thinking about how to hit:
*Bridge humbucker tone
*Strat-style notch tone
*Tele-style middle tone
*Relatively clear neck humbucker tone

Assuming an HH-configured guitar (the most common setup? Certainly my most common setup) and working according to inductance and placement this leads me the following layout:

/Neck humbucker outer coil - ~2.1H (Tele neck single coil)
\Neck humbucker inner coil - ~2.7H (Strat single coil)

/Bridge humbucker inner coil - ~2.7H (Strat single coil)
\Bridge humbucker outer coil - ~3.3H (Tele bridge single coil)

The neck humbucker coil mismatch and stronger inner coil should also help mitigate typical neck humbucker boominess. Conversely the bridge pickup in humbucker mode may be more shrill than desired.

Add to the above appropriate switching and capacitors to replicate 250K versus 500K pots.

Does anyone have any thoughts about this approach or what pickup/coils could best achieve this? I have both 22-fret and 24-fret guitars, and most tones would sound more "authentic" with the 22-fret pickup positioning but the 24-fret guitars should do a slightly better Strat notch tone by moving those coils closer together.

[antigua]

Fender did something similar with their Fireball humbuckers for the Meteora model, which has coil splitting guitarnuts2.proboards.com/thread/9926/fender-meteora-fireball-humbuckers-analysis , they made the coils imbalanced for better split tones with still decent series tones. 

Bridge - Measured L: 4.502H (bridge side: 2.077H, neck side: 1.939H)

Neck - Measured L: 4.058H (bridge side: 1.683H, neck side: 1.860H)



The neck and bridge are wired reverse of one another, so the neck splits neck side, bridge splits bridge side, and center position split is humbucking.

I think it sounds really good. If you go with higher inductance values, another thing worth trying IMO might be a ceramic or AlNiCo 8 magnet in the humbucker to give that sharper sound associated with Fender single coils. Since these pickups will have steel pole pieces no matter what, I'd stick with 500k pots and not bother faking the 250k pots.

Personally I like HH a lot better than HSS, even though I have some HSS Strats too. The humbucker bridge is always louder and softer, that's right I said softer, than the middle and neck single coils. It almost makes for a guitar that will do one or the other, but you can't switch between humbucker bridge and single coil neck without feeling some sense of let down, in either direction. With an HH, or an HH that splits both pickups at once, the neck and bridge tones are always more friendly with one another, and the HH guitar can have two distinct personalities depending on whether the two humbuckers are series or split.  

[sumitagarwal]

antigua this is great! And the Meteora does look interesting. Do you have an opinion on the 22 vs 24 fret neck pickup positioning? I figured 24 positioning is a bit more like an SG. I see the Meteora also leaves a gap at the neck, but not as much as an SG.

Going for the "full hot" approach to Strat and Tele inductance levels I think that means for sourcing coils I'd be looking for humbuckers with inductances of 6, 4.8, and 7.6.

As to the reverse wind, isn't it possible to simply flip one pickup's magnet and change which leads are treated as ground versus hot?

[antigua]

Feb 11, 2024 7:53:12 GMT -5 sumitagarwal said:

antigua this is great! And the Meteora does look interesting. Do you have an opinion on the 22 vs 24 fret neck pickup positioning? I figured 24 positioning is a bit more like an SG. I see the Meteora also leaves a gap at the neck, but not as much as an SG.

Going for the "full hot" approach to Strat and Tele inductance levels I think that means for sourcing coils I'd be looking for humbuckers with inductances of 6, 4.8, and 7.6.

As to the reverse wind, isn't it possible to simply flip one pickup's magnet and change which leads are treated as ground versus hot?

Yeah you can reverse the magnet and the hot and ground of the humbucker to get RW/RP with both outer or inner coils.

Based on the numbers, the total inductance of either coil alone is about 45% of both coils in series. So if you want two 2.7 henry coils, you'd have to part out a humbucker that's about 6 henries. 

[sumitagarwal]

Thanks! I was on the assumption that two typical humbucker coils in series is about 15% more inductance than the sum, but you're saying that 11% is probably a more typical number?

[sumitagarwal]

Another observation based on the limited data I'm looking at and limited understanding:
I'm finding that models that achieve a higher inductance comparative to their DCR (H/Ohms) also correlate with slightly higher resonant peaks.

Would it be reasonable to say that, other things being equal, those designs that achieve a given inductance with lower resistance will generally have higher resonant peaks?

[ms]

Feb 12, 2024 10:19:30 GMT -5 sumitagarwal said:

Another observation based on the limited data I'm looking at and limited understanding:
I'm finding that models that achieve a higher inductance comparative to their DCR (H/Ohms) also correlate with slightly higher resonant peaks.

Would it be reasonable to say that, other things being equal, those designs that achieve a given inductance with lower resistance will generally have higher resonant peaks?

Exactly.  The concept of Q was invented to cover this. The Q of a coil is 2*pi*f*L/R.  It is a bit  complicated since in  addition to loss from winding R, there is loss from eddy currents.  Often the loss from winding R is dominant, and so your measurements are right on.

[sumitagarwal]

Cool! Feels good to be learning stuff 

Now I feel like I better understand the focus that PRS has taken around increasing inductance with moderate resistance. That helps get that higher-output higher-clarity humbucking tone they're aiming for while simultaneously creating an inherently better-splitting pickup.

Would it be reasonable to say that this approach also helps mitigate the (split-coil) difference between steel poles and magnetic poles?

[ms]

Feb 12, 2024 10:35:28 GMT -5 sumitagarwal said:

Cool! Feels good to be learning stuff 

Now I feel like I better understand the focus that PRS has taken around increasing inductance with moderate resistance. That helps get that higher-output higher-clarity humbucking tone they're aiming for while simultaneously creating an inherently better-splitting pickup.

Would it be reasonable to say that this approach also helps mitigate the (split-coil) difference between steel poles and magnetic poles?

Yes, although it is not easy to achieve everything.  For example, it is easier to increase loss (lower Q) than decrease it (raise Q), but decreasing loss might be what you need to do.

It occurs to me that there is another possible approach to coil splitting humbuckers that would start with higher Q (lower loss).  Can you see what I am thinking of?

[sumitagarwal]

Would that be to start off with a relatively high-Q humbucker, split it to single (which would raise the Q further?), then add capacitors to bring Q down to the desired level?

[antigua]

Feb 12, 2024 10:35:28 GMT -5 sumitagarwal said:

Cool! Feels good to be learning stuff 

Now I feel like I better understand the focus that PRS has taken around increasing inductance with moderate resistance. That helps get that higher-output higher-clarity humbucking tone they're aiming for while simultaneously creating an inherently better-splitting pickup.

Would it be reasonable to say that this approach also helps mitigate the (split-coil) difference between steel poles and magnetic poles?

One thing that's important to remember about the higher Q factor is that it's not a desirable sound, usually. 

You can see here that a no load pot gives the highest Q factor. Guitar makers use 250k or 500k pots intentionally to reduce the Q factor. The two in parallel brings that actual resistance down to 125k or 250k respectively.  I don't know what PRS has been saying about inductance and resistance, but to date they've been making a lot of dubious claims about physics and guitar, and that sounds to me like another one.




Magnetic pole pieces give you a higher Q factor, because the AlNiCo magnet has higher electrical resistance than steel, so you get less eddy currents, less reduction in both high frequency and less reduction in Q factor. Fender uses 250k pots with their single coils to reduce the shrill sound of the high Q factor. 

It's true that when you split a humbucker, one reason is sounds anemic is because it has a lower Q factor than what you'd hear with 
magnetic pole pieces, but also the strength of the magnetic field is only about 25% as strong at the pole tops of the split humbucker. This is because a direct AlNiCo magnet is stronger than an AlNiCo magnet mounted to the underside of a humbucker, having it's magnetism redirected by steel pole pieces. Even a Strat pickup with a ceramic magnet and steel pole pieces (as found on cheap import guitars) is about half the magnetic strength of a vintage AlNiCo 5 Strat pickup. The stronger magnetism also results in more treble response at the pickup, which is why Strats and Telecasters sound so bright, even with 250k tone and volume pots.

You can get a pickup with steel pole pieces to have the higher Q factor by simply using 1meg or no load pots, but a lot of people who try this will tell you it sounds bad. The steel pole pieces also drive up the inductance of the pickup a lot, due to high permeability, so they almost always have a lower resonant peak at the same time. For example, a Strat pickup with AlNiCo pole pieces will be 2 henries, with steel pole pieces that same pickup will come out to 3 henries. And if a high Q factor is bad, a high Q factor at a lower audible frequency is even worse. If a person wanted to really fake an AlNiCo pickup with steel, they would have to assure that the inductance is especially low, put a very strong magnet underneath, and use high resistance tone and volume pots... it's easier to just use a real AlNiCo poled pickup than go through that kind of trouble. 

All together it's a balancing act, but the main point is that there is no reason to strive for a high Q factor if you're going to turn around and reduce it with 500k tone and volume pots. 

[ms]

Feb 12, 2024 11:32:16 GMT -5 sumitagarwal said:

Would that be to start off with a relatively high-Q humbucker, split it to single (which would raise the Q further?), then add capacitors to bring Q down to the desired level?

Capacitance would bring down the resonant frequency and that is certainly an important tool.  But, yes, starting with a higher Q humbucker is it, and the way I see to do that is to make it from two Fender type sc coils (Alnico cores rather than slugs or screws), not a new idea, but maybe not fully explored for this purpose.  Use 250K pots and you nail the split coil single coil sound.  Put the two coils in series with the 250K pot and you are moving in the right direction because you are starting with higher Q coils that might need to be lowered (as the loss induced by the 250K would do) when used as a humbucker.  I would make the two coils in one pickup the same for good hum cancelation, and  lower the resonant frequency as necessary by switching in parallel capacitors (and resistors if necessary).  The switching here could get a bit complicated.  Something to think about.

[sumitagarwal]

Feb 12, 2024 11:47:36 GMT -5 ms said:

Feb 12, 2024 11:32:16 GMT -5 sumitagarwal said:

Would that be to start off with a relatively high-Q humbucker, split it to single (which would raise the Q further?), then add capacitors to bring Q down to the desired level?

Capacitance would bring down the resonant frequency and that is certainly an important tool.  But, yes, starting with a higher Q humbucker is it, and the way I see to do that is to make it from two Fender type sc coils (Alnico cores rather than slugs or screws), not a new idea, but maybe not fully explored for this purpose.  Use 250K pots and you nail the split coil single coil sound.  Put the two coils in series with the 250K pot and you are moving in the right direction because you are starting with higher Q coils that might need to be lowered (as the loss induced by the 250K would do) when used as a humbucker.  I would make the two coils in one pickup the same for good hum cancelation, and  lower the resonant frequency as necessary by switching in parallel capacitors (and resistors if necessary).  The switching here could get a bit complicated.  Something to think about.

That's basically the design of the Seymour Duncan Stag Mag, right? I haven't looked into it much, maybe I should.

antigua regardless of PRS's marketing (much of which does seem like bunk) their results are interesting in the TCI and TCI "S" pickups.

Going by the data in the Guitar Pickup Database and DCR measurements by Phillip McKnight, it seems the basic design is an asymmetric wind with the slug coils about twice as strong as the screw coil, A5 magnets, relatively high inductance for the DCR (5.85H, 8.84 Ohms), and a 2.8Khz resonant peak that seems intentionally pegged right between a PAF and Fender single coils (like with a lot of PRS approaches trying to find the center between Gibson and Fender).

Interestingly they apparently don't use a ground wire, which may be to keep less pickup hardware out of the circuit? And from photos of the USA version TCI I think the baseplates may even be plastic. For the non-US TCI "S" it looks like earlier versions used a nickel-silver baseplate and later revised to brass.

[ms]

Feb 12, 2024 12:01:41 GMT -5 sumitagarwal said:

That's basically the design of the Seymour Duncan Stag Mag, right? I haven't looked into it much, maybe I should.

antigua regardless of PRS's marketing (much of which does seem like bunk) their results are interesting in the TCI and TCI "S" pickups.

Going by the data in the Guitar Pickup Database and DCR measurements by Phillip McKnight, it seems the basic design is an asymmetric wind with the slug coils about twice as strong as the screw coil, A5 magnets, relatively high inductance for the DCR (5.85H, 8.84 Ohms), and a 2.8Khz resonant peak that seems intentionally pegged right between a PAF and Fender single coils (like with a lot of PRS approaches trying to find the center between Gibson and Fender).

Interestingly they apparently don't use a ground wire, which may be to keep less pickup hardware out of the circuit? And from photos of the USA version TCI I think the baseplates may even be plastic. For the non-US TCI "S" it looks like earlier versions used a nickel-silver baseplate and later revised to brass.

Well I had not heard of that. A couple of comments:

Why traditional stagger?  Do they really want your plain G string to be too loud?

They seem to imply that you keep your 500K pots since they recommend a Pearly Gates in the neck position.  They also claim it is high output.  I think they are running a delicate balancing act by using smaller than #42 wire to increase the damping to compensate for  the 500K pots in the sc mode, as well as allowing more turns to keep the sc output level high.  I suppose this limits the range of possible hb characteristics, but it could be what you want, right?  What I am thinking of would have a different range of possibilities.

[sumitagarwal]

Feb 12, 2024 12:27:17 GMT -5 ms said:

Feb 12, 2024 12:01:41 GMT -5 sumitagarwal said:

That's basically the design of the Seymour Duncan Stag Mag, right? I haven't looked into it much, maybe I should.

antigua regardless of PRS's marketing (much of which does seem like bunk) their results are interesting in the TCI and TCI "S" pickups.

Going by the data in the Guitar Pickup Database and DCR measurements by Phillip McKnight, it seems the basic design is an asymmetric wind with the slug coils about twice as strong as the screw coil, A5 magnets, relatively high inductance for the DCR (5.85H, 8.84 Ohms), and a 2.8Khz resonant peak that seems intentionally pegged right between a PAF and Fender single coils (like with a lot of PRS approaches trying to find the center between Gibson and Fender).

Interestingly they apparently don't use a ground wire, which may be to keep less pickup hardware out of the circuit? And from photos of the USA version TCI I think the baseplates may even be plastic. For the non-US TCI "S" it looks like earlier versions used a nickel-silver baseplate and later revised to brass.

Well I had not heard of that. A couple of comments:

Why traditional stagger?  Do they really want your plain G string to be too loud?

They seem to imply that you keep your 500K pots since they recommend a Pearly Gates in the neck position.  They also claim it is high output.  I think they are running a delicate balancing act by using smaller than #42 wire to increase the damping to compensate for  the 500K pots in the sc mode, as well as allowing more turns to keep the sc output level high.  I suppose this limits the range of possible hb characteristics, but it could be what you want, right?  What I am thinking of would have a different range of possibilities.

Yea, that stagger is definitely a head-scratcher.

I've also seen a number of anecdotes that it's a better neck pickup than bridge pickup.

The design and idea is interesting. I feel like the best use of this something like this is combined with a relatively hot conventional bridge humbucker for good versions of the following four tones:

*Strat neck solo (Stag Mag neck outer coil)
*Strat notch position (Stag Mag neck inner coil + bridge humbucker inner coil)
*Tele middle position (Stag Mag neck outer coil + bridge humbucker outer coil)
*Hot bridge humbucker solo

[ms]

Feb 12, 2024 13:09:46 GMT -5 sumitagarwal said:

Yea, that stagger is definitely a head-scratcher.

I've also seen a number of anecdotes that it's a better neck pickup than bridge pickup.

The design and idea is interesting. I feel like the best use of this something like this is combined with a relatively hot conventional bridge humbucker for good versions of the following four tones:

*Strat neck solo (Stag Mag neck outer coil)
*Strat notch position (Stag Mag neck inner coil + bridge humbucker inner coil)
*Tele middle position (Stag Mag neck outer coil + bridge humbucker outer coil)
*Hot bridge humbucker solo

In my old fashioned view, I would want to emphasize the bridge pickup. IMO, the tele bridge is the classic electric guitar sound, while the PAF bridge, not such a great tone considered alone, is a good contrast to the tele sound.

[sumitagarwal]

Feb 12, 2024 13:44:29 GMT -5 ms said:

Feb 12, 2024 13:09:46 GMT -5 sumitagarwal said:

Yea, that stagger is definitely a head-scratcher.

I've also seen a number of anecdotes that it's a better neck pickup than bridge pickup.

The design and idea is interesting. I feel like the best use of this something like this is combined with a relatively hot conventional bridge humbucker for good versions of the following four tones:

*Strat neck solo (Stag Mag neck outer coil)
*Strat notch position (Stag Mag neck inner coil + bridge humbucker inner coil)
*Tele middle position (Stag Mag neck outer coil + bridge humbucker outer coil)
*Hot bridge humbucker solo

In my old fashioned view, I would want to emphasize the bridge pickup. IMO, the tele bridge is the classic electric guitar sound, while the PAF bridge, not such a great tone considered alone, is a good contrast to the tele sound.

For classic electric guitar, absolutely. But I think the Nashville Tele already has those bases pretty well covered.

For a more versatile classic-to-modern pickup and electronics layout purpose-built from the ground up I think that bridge humbucker is critical. Of course even better would be to rig up a bridge humbucker that splits to a good Tele-esque tone, so then we have these 5:

*Strat neck solo (Stag Mag neck outer coil)
*Strat notch position (Stag Mag neck inner coil + bridge humbucker inner coil)
*Tele middle position (Stag Mag neck outer coil + bridge humbucker outer coil)
*Tele bridge solo (Hot bridge humbucker inner coil)
*Hot bridge humbucker solo

[sumitagarwal]

Looks like the Duncan Stag Mag is not the only humbucker design with magnetic poles. The Rio Grande Tallboy and Muy Grande are a very similar approach, and the Fralin Twangmaster is also along the same lines but different.

[ms]

Feb 12, 2024 16:06:27 GMT -5 sumitagarwal said:

Looks like the Duncan Stag Mag is not the only humbucker design with magnetic poles. The Rio Grande Tallboy and Muy Grande are a very similar approach, and the Fralin Twangmaster is also along the same lines but different.

and the rio grande halfbreed.  Those are expensive pickups.

The Fralin twangmaster is something else.  Not sure what, but it looks like it gets hum canceling with two coils, each containing three of the pole pieces.  The coils are separated (in the direction along the strings) to prevent field cancelation I guess, but at the expense of continuity in location.   I think.

[sumitagarwal]

Yea, the Twangmaster looks very similar to the Railhammer Humcutter, although the Humcutter is "reversed" compared to normal Fender bridge pickup orientation.

Twangmaster seems really cool, especially for a bridge position, but unfortunately I do need a bridge humbucker sound.

[sumitagarwal]

Feb 11, 2024 16:41:16 GMT -5 sumitagarwal said:

Thanks! I was on the assumption that two typical humbucker coils in series is about 15% more inductance than the sum, but you're saying that 11% is probably a more typical number?

antigua following up with some additional thoughts here:

The 11% additional inductance from the coupling, does that apply to parallel wiring as well? I.e. for coils with an inductance of 2H we'd get:
((2*2)/(2+2)) * 1.11 = 1.11H?

Also, does this account for the difference in split tones when leaving the other coil out of circuit versus shunting to ground?

Other coil out of circuit, remaining coil: 2H
Shunted coil to ground, remaining coil: 2*1.11 = 2.22H?

[antigua]

Feb 13, 2024 8:42:31 GMT -5 sumitagarwal said:

Feb 11, 2024 16:41:16 GMT -5 sumitagarwal said:

Thanks! I was on the assumption that two typical humbucker coils in series is about 15% more inductance than the sum, but you're saying that 11% is probably a more typical number?

antigua following up with some additional thoughts here:

The 11% additional inductance from the coupling, does that apply to parallel wiring as well? I.e. for coils with an inductance of 2H we'd get:
((2*2)/(2+2)) * 1.11 = 1.11H?

Also, does this account for the difference in split tones when leaving the other coil out of circuit versus shunting to ground?

Other coil out of circuit, remaining coil: 2H
Shunted coil to ground, remaining coil: 2*1.11 = 2.22H?

Yeah the parallel inductance is greater than  (coilA * coilB ) / ( coilA + coilB )

I checked with a couple humbucker that I measured parallel inductance for, a most of the time I don't measure that. Both come out higher than what the math suggests. For the SD 59, the parallel inductance is 13% greater than what the formula suggests for the coils by themselves. 

guitarnuts2.proboards.com/thread/7763/seymour-duncan-custom-analysis-review

guitarnuts2.proboards.com/thread/7995/seymour-duncan-duckbucker-analysis-review

When I say coils by themselves, they were in-situ when I measured their inductance, but if I had unscrewed them from the base plate and have them physically far apart, their inductance would drop by some small amount, I'm not sure by how much. 

[sumitagarwal]

Thanks antigua, this is very interesting since we expect the single coils to be getting some induction boost from each other when used singly already, and whatever this boost is is increased further when they're operating in parallel.

When in single-coil mode would the inductance be affected at all based on whether the unused coil was open or closed? I saw your write-up here about the damping (and, based on PRS's design changes soon after you wrote that, seems they may have too!): guitarnuts2.proboards.com/thread/7769/damping-caused-unused-splitting-humbucker

My understanding isn't strong enough to know if there is any relationship here between the damping effects and any changes in induction.

[antigua]

Feb 13, 2024 13:32:55 GMT -5 sumitagarwal said:

Thanks antigua, this is very interesting since we expect the single coils to be getting some induction boost from each other when used singly already, and whatever this boost is is increased further when they're operating in parallel.

When in single-coil mode would the inductance be affected at all based on whether the unused coil was open or closed? I saw your write-up here about the damping (and, based on PRS's design changes soon after you wrote that, seems they may have too!): guitarnuts2.proboards.com/thread/7769/damping-caused-unused-splitting-humbucker

My understanding isn't strong enough to know if there is any relationship here between the damping effects and any changes in induction.

I don't really know. The graphs show a higher frequency when the second coils is closed, and that suggests a lower inductance, but that seems counterintuitive. I can do a test later and find out. 

If you don't already own a DE-5000 I'd recommend picking on up because you seem to have a want for good volume of inductance information, and the DE-5000 makes its fairly effortless. 

[sumitagarwal]

Feb 13, 2024 17:28:12 GMT -5 antigua said:

Feb 13, 2024 13:32:55 GMT -5 sumitagarwal said:

Thanks antigua, this is very interesting since we expect the single coils to be getting some induction boost from each other when used singly already, and whatever this boost is is increased further when they're operating in parallel.

When in single-coil mode would the inductance be affected at all based on whether the unused coil was open or closed? I saw your write-up here about the damping (and, based on PRS's design changes soon after you wrote that, seems they may have too!): guitarnuts2.proboards.com/thread/7769/damping-caused-unused-splitting-humbucker

My understanding isn't strong enough to know if there is any relationship here between the damping effects and any changes in induction.

I don't really know. The graphs show a higher frequency when the second coils is closed, and that suggests a lower inductance, but that seems counterintuitive. I can do a test later and find out. 

If you don't already own a DE-5000 I'd recommend picking on up because you seem to have a want for good volume of inductance information, and the DE-5000 makes its fairly effortless. 

Hrmm... I've been promising myself that I'll be reducing my number of hobbies, not increasing them. Although that's seeming unlikely..

I worry that at that point I'll end up buying a winder and tooling too.

[Yogi B]

Feb 13, 2024 17:28:12 GMT -5 antigua said:

I don't really know. The graphs show a higher frequency when the second coils is closed, and that suggests a lower inductance, but that seems counterintuitive.

With both coils connected we can direct the coupling into either positive or negative feedback. This is easiest to reason about in the series case: the current through one coil induces a voltage in the other coil, but the current that is developed must pass through both coils — adding to (or subtracting from, depending on the phase relationship) the original current.

However, when the other coil is shorted, the current from the mutual inductance flows through only the shorted coil — there's no direct electrical feedback path. Of course, this current is still directed so as to oppose the magnetic field of the non-shorted coil, thus lowering the effective inductance.

The magnitude of this reduction is proportional to the coupling coefficient (k) squared, since the resultant change in output current is a result of being coupled twice (once over to the shorted coil & then back to the non-shorted coil). Additionally, since it is voltage that is induced yet current that determines magnetic field strength, the reduction in inductance is a function of the impedance of the shunted coil. Therefore, the reduction is frequency dependent and complex (the overall change in impedance is not limited to purely a change in reactance).

Finally, just to be clear, since |k| ≤ 1, then k² ≤ |k|. For example: assuming k is in the range 10—15%, then k² is in the range 1—2.25%. As such, the change in inductance from (un-)shorting the unused coil should be much smaller than the change (in individual inductances) between having one or two coils selected. That being said, looking purely at inductance is ignoring that the strings still induce a voltage into the deselected coil and, if that's allowed to develop a current (i.e. if the coil is shunted), that signal would be coupled across to the selected coil (at an amount proportional to just k, unsquared).

Edit: adjusted inequalities relating to k, since it's possible define negative values as indicating reversed phase relationships — this makes it clearer that the reduction in inductance due to shunting is unaffected by phase (since it's proportional to k²).