Inner stray capacitance of passive mag transducers. A few thoughts.

[bbsailor]

Years after the facts, I feel free to share the ideas at work but I can't be more precise about the product involved, its 5spice model and the corrections applied to it, since it was not my design.
I've simply said to myself that people knowing this kind of stuff could/would extrapolate on the principles involved and maybe make them work for their own benefit. :-)

freefrog,

Go to the web site till.com to see other research on pickup resonance. The web site also describes how to make an FET buffer/amplifier in the guitar end of a guitar cable to minimize the effect of coax capacitance on the pickup resonant frequency and add gain to minimize noise pickup.

I built this buffer cable and it really has an audible effect on the pickup resonance peak frequency.

Joseph J. Rogowski

[freefrog]

Thx for sharing, bbsailor.

[freefrog]

Another contribution of the Sunday morning while the house is sleeping. Same editorial limitations than in my previous posts but I hope my selective trimmed screenshots to be clear enough to allow to "get the picture" that I wanna share.

It's about regular humbuckers.

I’ve found them variable when it comes to secondary resonances.

Because of their structure and wiring, the coils of some models produce practically identical curves from 20 to 20 000 hz. That’s the case with Gretsch Filter’Tron’s but also with Gibson ‘’Patent Sticker’’or T-Tops IME.

In other models, the coils excited separately produced more dissimilar curves, changing the whole pickup in a “double-tuned circuit” once in series. Hence a comb filtered response beyond the main resonant peak, adding itself to the comb filtering of vibrating strings by harmonic nodes modelized by Don Tillman. Hence sometimes undesirable musical in the audio range effects, due to dulled harmonics or to insisting parasitic resonances : two examples of this have been evoked in the post 2.

The other examples that I’ll evoke below are not among these problematic transducers. their coils have a relatively high “crossover frequency”, not susceptible to really affect the audio range of a played guitar. IOW, these pickups resonate in the way illustrated by the red lines in the Sspice sims of my post 2. But their crossover frequency is not high enough to remain unseen and that’s why I’ll evoke these examples.

The transducers themselves are run of the mill Gibson style humbuckers, but wound with thinner wire than usually (hence a ‘’P.A.F. like” inductance for a DCR of +/- 11.4k in series).

The first one has been fitted with a Mogami 4 conductors cable, among the least capacitive possible.

In the first screenshot below, the black and pink lines translate the resonant peaks of its coils alone (split and unloaded): one resonates at a way higher frequency than the other, just because of the parasitic capacitance increased by to the hook up cable and mentioned in the frame. The lower pitched dark red and green line translate what these dissymmetric responses become when the pickup is wired in series, leading it to comb filter the highest harmonics.



Below is now a comparative screenshot in which the second (and identical) humbucker is wired with two kinds of hook up cables. The black and pink lines now show the response of each coil when this pickup is wired in series to a braided shielded wire. As expected, the higher capacitance of such a coaxial cable drags down the resonance of both coils in series but it has no real consequences on their “crossover frequency”, because braided shielded wire doesn’t increase the capacitance of one coil more than the other.
Things differ with a Mogami cable, involved in the dark red and green lines: although it’s less capacitive than the coax cable, as shown by the higher pitched resonances, it has not the same influence on resonances when its inner bare wire is grounded (left picture) and when it’s not (right picture).



It’s just a question of capacitance added between conductors and increasing the capacitive load of one coil only. But as simulated in my post 2 and illustrated above by experimental screenshots, it changes the response of the whole pickup in the highest harmonic region.
This change might become a problem with a transducer having already two coils with “naturally” dissimilar responses, exhibiting a lower pitched “crossover frequency” like in the two examples of my post 2 (or like in the case of the stacked humbucker already commented here). Unless it would be considered as a bonus, since it can promote selectively some of the highest harmonics and contribute to a more “open” perceived sound through a clean amp / cab with extended high range. After all, double-tuned circuits are seen as desirable in other contexts.

FWIW. More later if possible.

[Antigua]

These differences you're highlighting seem to be upwards of 10kHz, but AFAIK that's pretty much beyond the operational frequency of most guitar speakers, and even with respect to the guitar strings, I think the harmonics in that region are short lived transients.

[freefrog]

The other examples that I’ll evoke below are not among these problematic transducers. their coils have a relatively high “crossover frequency”, not susceptible to really affect the audio range of a played guitar. IOW, these pickups resonate in the way illustrated by the red lines in the Sspice sims of my post 2. But their crossover frequency is not high enough to remain unseen and that’s why I’ll evoke these examples..

Let's recall that I've shared selective examples of humbuckers with a lower "crossover frequency" in the post 2 and relatively detailed explanations about my experience with such things in post 8.

When harmonics quirks beyond resonant frequency are close to 20khz, it's not a problem and can even become a bonus when it comes to felt "openess". When such things happen around 11khz or 12khz (or @ lower frequencies), it becomes potentially annoying, either by lack either by local excess of upper harmonics... and pickups with this specificity are not that rare.
I share below another selective chart about the stacked HB that I've capacitively corrected years ago : crude Rz electrically induced vs response of the pickup played direct to the board (in blue) vs spectrum of a JC120 receiving a swept sine wave and heard through a SM57 (in white).
The harmonic peak shown by a vertical arrow was insisting enough to sound like tinnitus through any bright amp/cab and is the actual reason why the transducer has been corrected.

Of course, such things can't necessarily be heard through devices like a Joyo American Sound, for instance. Through a JC120 or Fender Twin played loud, it's another story.

[freefrog]

A few more "slices" of experiments. Same trimming than before for the same contextual reasons but I still hope that it will allow to “get the picture” to share.

Here are the curves of some old Gibson humbuckers with coaxial cables: a T-Top, a Shaw (mildly loaded, capacitively speaking). Linear scale, vertical steps of 1dB. No integrator used while the coils were excited separately by a low impedance air coil positioned horizontally above them, as previously.

One can see differences in the responses of slugs vs screws coils but no quirks due to parasitic capacitance in the audio range with these transducers:





Below are now some curves obtained in the same way from other HB’s, each wired in series with a 4 conductors cable, giving lower “crossover frequencies” between coils than in my post 18.

All come from famous American brands. All have been tested with resistive & capacitive loads as they would have in real life (these resistive and capacitive loads being susceptible to vary from one pickup to the other but it has no importance here IMHO: I've explained somewhere above that in my humble experience, "secondary resonances" were not really affected by overall RC loads)

Here is a set of Gibson style HBs. Black & pink= bridge pickup. Red & green = neck pickup.



Here are the resonances of a P90 sized high gain humbucker, from the same brand.




P90 sized humbuckers are interesting in this case IMHO: typical ones have cosmetically and structurally symmetrical coils with the same magnetic poles, avoiding any difference inherent to slugs and screws coils. It leaves aside a variable.

So the two other examples that I'll provide below also involve P90 sized HB’s.

In each of these two pickups, each coil will have the same magnetic poles but also the same turns count and measured inductance than the other.

But one will appear as much more altered by the side effects of stray capacitance in its 4 conductors cable.

This model has already been mentioned and its response illustrated in my post 2. The resonance of its coils in series / humbucker mode will be reproduced again in black and pink in the screenshot below.



The other pickup has also a 4 conductors cable but in series humbucker mode, its coils resonate according to the green and red curves: this model seems way less affected by parasitic capacitance of its hook up cable. Why? Sum up of my own answer to come if time permits.

Life is calling me: I wish you all a nice day / week / summer. :-)

[freefrog]

So, what is the difference between the last two P90 sized humbuckers, both with 4 conductors cables?

Splitting them and checking the resonance of each coil gives an answer.

Here are the frequencies of coils with the model which seems the most affected by stray capacitance. Green & red lines = coils in series. Black & pink = coils separately, when the PU is split (and with a rest of double resonance due to HOW it was split; I'll try to share later a few things showing how connecting conductors differently can bring noticeable differences). Vertical arrows show where its coils resonate in this case.



In this first example, the grey H shaped segment translates a relatively noticeable gap between resonant frequencies of isolated coils and this gap is paradoxically imputable to their inherent symmetry: both having similar DCR, inductance and stray capacitance, the asymmetrical stray capacitance of the 4 conductors cable just adds itself to the transducer, changing it in a double-tuned circuit.

Let’s check now the second pickup, way less affected by stray capacitance when wired in series.



Coils excited separately (split) resonate at closer frequencies, as shown by the narrower H shaped grey segment.

In this second transducer, less affected by its own cable when wired in series, both coils have the same inductance but one is inherently more capacitive than the other. This most capacitive coil is connected to the least capacitive part of the 4 conductors cable and the most capacitive part of this cable is reciprocally connected to the least capacitive coil. Resonances of separate coils are therefore closer to each other in the audio spectrum. So the “double tuning” effect becomes less present when coils are in series.
This softening effect seems also favored by the higher resistance of the most capacitive coil.

The reason why it's more resistive and more capacitive with a same inductance has apparently to be found in the only physical difference between coils (both stuffed with the same magnetic poles, for the record): in this second pickup, one measures 5.8k and the other, 12k, because they were wound with different wire gauges.


Now, different wire gauges are not mandatory to correct the double tuning effect and one can do more against that than simply dissociating the bare ground wire from other conductors. Below are the resonances of a pickup identical to the first (IOW: problematic) P90 sized HB once its least capacitive coil simply associated to an external capacitor (precisely: a pair of caps measuring168pF in this case): green and red lines show the resonances of the stock pickup in series. Black and pink show the same thing with the added 168pF symmetrizing artificially the capacitive load of its 4 conductors cable + coils.



These various iterations don’t sound the same: once capacitively corrected, the “problematic” HB exhibits no more the paradoxical blend of closed nasal tone and harmonic harshness that it had in its stock version. The sonic difference is subtle but it’s there.

FWIW. No need to mention the brands and models involved, since I'm not in a commercial perspective. But once again, they all come from well known American winders.

EDIT- When parasitic capacitance is manipulated as explained above, the "double tuning effect" can be either softened / cancelled, either increased and/or moved along the frequential range beyond main resonance. There's many possibilities here, evoking me the action of a VariTone, but applied to high harmonics rather than to fundamental notes.

[freefrog]

Another test, with a "Hybrid" humbucker whose coils have different wire gauges and turns counts.

Screws coil = 4.3k, 2H.
Slugs coil = 8k, 3.15H.

Below is how its coils resonate in SC and HB (series) modes, through a 4 conductors cable wired in 2 different ways: with ground (and therefore bare ground wire) on the screws coil side (left) then on the slugs coil side (right). Resistive and capacitive loads mimic a single volume pot of approximatively 500k and 3m of standard guitar cable.



In this case secondary peaks and dips influence even the main resonant peaks in the right pic.

It makes harmonic differences obvious when the pickup is played: direct to the board through a 1M input, the two options cause a difference of 6dB to 9dB around 2khz, 6dB to 12dB around 4 or 4.5khz, 6 to 9dB around 5 or 5.5khz and so on... the second option (right pic) makes the spectrum of the pickup much more "comb filtered", leading it to interact in a not desirable way with the harmonic comb filtering modelized by Don Tillman.

IOW, it's almost a different pickup according to the wiring used.

That's why I share the last trimmed screenshot above: it might give some ideas to those who use hybrid humbuckers (like reversing the wiring AND the magnet to give a more even harmonic response to the pickup, while keeping it in phase with other transducers).

I don't know if I'll be able to continue this topic. FWIW, as it is.

[dcoyle]

Just came across this thread and I am very grateful you spent the time to share what you learned. Thank you.

[freefrog]

Impulse add of a Sunday morning, inspired by a recent topic elsewhere where my contribution has been useless. Maybe it will make more sense to share it here, who knows?... it's about the inner capacitance of coils.

Below is the electrically induced impulse response of P.A.F. style humbuckers with identical specs, except for their coils - either hand wound with PE insulated 42AWG, either machine wound ones (with plain enamel as well as polysol insulation).

The hand wound model involved is characterized by an obviously low inner capacitance: unloaded, it has the highest pitched resonant frequency I've been able to measure on P.A.F. replicas.

Whatever were the parts associated to its coils (different baseplates, keeper bars, slugs, screw poles, magnets, covers) and the host guitars used (2 Les Paul's, a semi-hollow, a Flying V, an Explorer, a super Strat), this model has consistently exhibited a fast and strong attack: it has virtually no perceived compression - and is not always nice to hear because of this: the post that I'm writing is NOT an advertising for hand wound coils, it's a testimonial.

My pics have been trimmed for the same questions of intellectual property than before but follow the usual scales (output vertically, time horizontally) and should therefore make sensible what happens IME with the mentioned hand wound model: even when its attack is slowed down by Foucault currents due to a cover, it's faster than machine wound pickups with identical LR specs. And it clearly affects the feel when this model of pickup is played, in our experience here... YMMV (and I don't plan to argue about that: not enough free time, thx for your understanding).

FWIW.