Why increase the pickup Q: design and measurements

[ms]

Consider leaving the pickup coil parameters as much the same as possible, but raising the Q of the coil. One way to do this while keeping the inductance the same is for the the series resistance to go down. Also at the same time, you can decrease eddy current losses. If you did one or both, then the pickup would need additional external loss in order to reestablish the Q to give the proper sound.

How would you do this? Here are two possibilities, that could be used separately, or together:
Lower the resistance of the volume control.
Run the tone control in a lower number position, resulting in the normal level of highs, since the tone control is a Q control when it is on the higher numbers.

The first would have the effect of decreasing the loss of highs resulting from turning down the volume control the since the R in series with the shunt cable capacitance would decrease for the same pot rotational position.

The second would allow the tone control to be raised when the volume control is reduced to put back some of the missing high frequencies.

Note on tone control: the tone control capacitor can be replaced with a resistor (to stop it from shorting out the signal completely) in order to make it a pure Q reducer over the entire range. IMO, this is a better tone control.

Note on 1. and 2.: You could reduce the volume R part way, and have a combination of both effects. This is more practical since pots are not available in many different resistances.

We can reduce the resistance of the coil by taking off turns. Then we can increase the inductance without adding turns back by improving the flux return path. The result is that the Q goes up. We want to make a single coil sized pickup, and so we do not have a lot of space. Also, we want the pickup to be magnetic hum canceling, without getting “humbucker sound” and keep about the same output level associated with a single coil. The only way I can see to do this involves using six small coils. A coil wound on a cylindrical core can be made tight, thus keeping the resistance down. (I have worked with six coil pickups on and off for the last 10 years, and I did make one in the sixties when I was in high school.) For a bridge pickup, the six coils with alternating electrical and magnetic polarity can be located in a straight line. This would cause nulls during string bends with a neck pickup.

So the the idea is to provide a high permeability path from the end of the coil away from the string back to the end near the string. Of course we cannot close the path because then we would be very insensitive to the vibrating string. Ferrite is used because available material can have both high permeability and resistivity in the audio range.

The coils are shown in the attachment.   The core is a 73 material cylinder .437” long and .2” in diameter and has a relative permeability of 2500. This is more than necessary. It has a small hole along its axis, and therefore it is called a “bead”. The core is bought from Amidon (as are the other ferrite pieces) and is designated as FB-73-1801. This core, a fiber washer, and a ferrite “brick” make up the bobbin as shown here: . An sc pickup cover is also shown in the picture. Ferrite plates are glued to the sides to guide the flux. This is a neck pickup under construction, using two sets of three coils, where the sets have opposite electrical and magnetic polarity. The two sets are offset to prevent cancelation when bending between the two. The bricks guide the flux from the end of the core to the plates.

The fiber washer has a hole large enough to allow the insertion of a neo magnet 1/8” in diameter and 1/32” thick, which then sits on the core.. This small magnet is sufficient because it is backed by a high permeability core.

The electrical properties of the complete pickup were measured using the methods described here: guitarnuts2.proboards.com/thread/7834/impedance-measurements-computer-interface-unit. The results of the measurements will be discussed in the following posts.

[ms]

The pickup that was measured is a bridge pickup shown here: .  The start-start, finish-finish.... connections that give excellent hum cancelation can be seen.  The impedance of the pickup was measured and is shown here:.  Capacitance was added across the pickup to bring the resonance down into the normal range.  The inductance (about 2H) is normal for a single coil pickup.  The resistance (about 2.4K) is very low.  The resonant peak is high and narrow.  The pickup is wound with #42 wire.  There is probably enough room so tat the wire size could be increased a bit, lowering the resistance a bit more.

[ms]

Now let's look at the frequency response with a resistive load of 200K.  This value is chosen because it is the standard that Antigua has established.  It is the parallel combination of a 250K volume pot and the 1M input impedance off a typical amplifier.  


 Note: This is actually not the total load on the pickup. The tone pot should should be included , too, since the tone capacitor is effectively a short at the higher frequencies where the load matters (kind of like a coupling capacitor between stages in an amplifier).  

The response with 200K load is shown here: The height of the peak is almost 10 db, higher than normal, and high enough so that the load on the pickup could be lowered.

[antigua]

Maybe a super high Q pickup could give rise to an interesting tone, especially if the resonant peak was lowered with capacitance, making for a wah-wah like high-Q tone. For the most part, guitarists intentionally discard Q with 250k/500k pots in lieu of 1meg pots.

I think an interesting experiment (or product prototype) to try would be to vary the inductance of the pickup by raising or lowering the ferrite, towards or away from the coil. If the inductance only changed by 300mH, guitarists might not perceive much difference, but if the inductance could be varied by 1 henry or more, it could possibly result in a pickup that has both "vintage" and "hot" modes.

It's probably a lot to ask for, but I found in a previous experiment that the steepness slope beyond the resonant peak also
played a role in perceived brightness, since it dictates the rate of roll off by frequency, and I wonder of there is any way that the steepness of that slope could be changeable they mechanical or passive electrical means. 

[Charlie Honkmeister]

I don't think that increasing the maximum possible Q should be a sonic design goal in itself,  but it can be a yardstick for measuring some other parameters (internal resistance of the coil, lower eddy losses, efficiency of the magnetic circuit)  that are worth pursuing in a good pickup design. 

From a pure transducer engineering standpoint, designing a transducer that has a big resonant peak in the very frequency range you're trying to transduce, is just nuts.  Nobody would do it that way these days;  we're just faced with the consequences of what "they" did back then, for their own reasons, and are used to what "that" sounds like. 

If you are chasing a subjectively better tonality but one which is still familiar as an "electric guitar sound,"   you aren't going to need resonant peaks more than 6-8.5 dB,  because when loaded by volume and tone controls, through a cable,  into the amp input impedance, that's the practical Q you're probably going to get with a passive setup and a conventional design.   And that's what we are all used to hearing.  

But maximizing the possible Q on the bench might be indicative that some other things in the pickup design are good and won't get in the way of you going in the direction of "vintage overall sound, but better."   You'd  have in general more flexibility to tailor the sound outside the pickup,  as Mike (ms) pointed out.   

But if you aren't trying to "emulate but improve on", the tonalities we know and love from the instrument, and/or if you want the absolute most tonal flexibility possible,  IMHO it doesn't make sense to have a resonant peak at all;  you should go for "hi-fi flat"  and then do any signal processing you want downstream.  

[ms]

May 1, 2017 3:39:15 GMT -5 antigua said:

Maybe a super high Q pickup could give rise to an interesting tone, especially if the resonant peak was lowered with capacitance, making for a wah-wah like high-Q tone. For the most part, guitarists intentionally discard Q with 250k/500k pots in lieu of 1meg pots.

I think an interesting experiment (or product prototype) to try would be to vary the inductance of the pickup by raising or lowering the ferrite, towards or away from the coil. If the inductance only changed by 300mH, guitarists might not perceive much difference, but if the inductance could be varied by 1 henry or more, it could possibly result in a pickup that has both "vintage" and "hot" modes.

It's probably a lot to ask for, but I found in a previous experiment that the steepness slope beyond the resonant peak also
played a role in perceived brightness, since it dictates the rate of roll off by frequency, and I wonder of there is any way that the steepness of that slope could be changeable they mechanical or passive electrical means. 

Well maybe you could, but it is hard enough to make these pickups without making the ferrite movable!

[ms]

May 1, 2017 9:27:35 GMT -5 Charlie Honkmeister said:

I don't think that increasing the maximum possible Q should be a sonic design goal in itself,  but it can be a yardstick for measuring some other parameters (internal resistance of the coil, lower eddy losses, efficiency of the magnetic circuit)  that are worth pursuing in a good pickup design. 

From a pure transducer engineering standpoint, designing a transducer that has a big resonant peak in the very frequency range you're trying to transduce, is just nuts.  Nobody would do it that way these days;  we're just faced with the consequences of what "they" did back then, for their own reasons, and are used to what "that" sounds like. 

If you are chasing a subjectively better tonality but one which is still familiar as an "electric guitar sound,"   you aren't going to need resonant peaks more than 6-8.5 dB,  because when loaded by volume and tone controls, through a cable,  into the amp input impedance, that's the practical Q you're probably going to get with a passive setup and a conventional design.   And that's what we are all used to hearing.  

But maximizing the possible Q on the bench might be indicative that some other things in the pickup design are good and won't get in the way of you going in the direction of "vintage overall sound, but better."   You'd  have in general more flexibility to tailor the sound outside the pickup,  as Mike (ms) pointed out.   

But if you aren't trying to "emulate but improve on", the tonalities we know and love from the instrument, and/or if you want the absolute most tonal flexibility possible,  IMHO it doesn't make sense to have a resonant peak at all;  you should go for "hi-fi flat"  and then do any signal processing you want downstream.  

It is a goal because it allows you to use lower value vol pots, reducing the amount the tone changes with the volume control position.   Or it allows you to run the tone control lower in normal operation so that you have some boost available for when you need it.  Or some combination of the two.

[Charlie Honkmeister]

May 1, 2017 9:32:39 GMT -5 ms said:

May 1, 2017 9:27:35 GMT -5 Charlie Honkmeister said:

I don't think that increasing the maximum possible Q should be a sonic design goal in itself,  but it can be a yardstick for measuring some other parameters (internal resistance of the coil, lower eddy losses, efficiency of the magnetic circuit)  that are worth pursuing in a good pickup design. 

From a pure transducer engineering standpoint, designing a transducer that has a big resonant peak in the very frequency range you're trying to transduce, is just nuts.  Nobody would do it that way these days;  we're just faced with the consequences of what "they" did back then, for their own reasons, and are used to what "that" sounds like. 

If you are chasing a subjectively better tonality but one which is still familiar as an "electric guitar sound,"   you aren't going to need resonant peaks more than 6-8.5 dB,  because when loaded by volume and tone controls, through a cable,  into the amp input impedance, that's the practical Q you're probably going to get with a passive setup and a conventional design.   And that's what we are all used to hearing.  

But maximizing the possible Q on the bench might be indicative that some other things in the pickup design are good and won't get in the way of you going in the direction of "vintage overall sound, but better."   You'd  have in general more flexibility to tailor the sound outside the pickup,  as Mike (ms) pointed out.   

But if you aren't trying to "emulate but improve on", the tonalities we know and love from the instrument, and/or if you want the absolute most tonal flexibility possible,  IMHO it doesn't make sense to have a resonant peak at all;  you should go for "hi-fi flat"  and then do any signal processing you want downstream.  

It is a goal because it allows you to use lower value vol pots, reducing the amount the tone changes with the volume control position.   Or it allows you to run the tone control lower in normal operation so that you have some boost available for when you need it.  Or some combination of the two.

   OK,  higher possible Q  allows you to move toward lower source impedance for the pickup, for about the same output level,  so you can do those things.   I'm on board with that.   If you get a high efficiency magnetic circuit, you can lower turns and/or use lower AWG wire.  The lower the source impedance, the more flexibility you have on usability of the volume and tone controls.  Also,  the cable capacitance will affect the pickup resonant frequency less.   All good things. 

[ms]

May 1, 2017 10:46:27 GMT -5 Charlie Honkmeister said:

May 1, 2017 9:32:39 GMT -5 ms said:

It is a goal because it allows you to use lower value vol pots, reducing the amount the tone changes with the volume control position.   Or it allows you to run the tone control lower in normal operation so that you have some boost available for when you need it.  Or some combination of the two.

   OK,  higher possible Q  allows you to move toward lower source impedance for the pickup, for about the same output level,  so you can do those things.   I'm on board with that.   If you get a high efficiency magnetic circuit, you can lower turns and/or use lower AWG wire.  The lower the source impedance, the more flexibility you have on usability of the volume and tone controls.  Also,  the cable capacitance will affect the pickup resonant frequency less.   All good things. 

Right. Lowering the series R affects the resonant frequency a bit.  Adding resistance in parallel does not.

[stratotarts]

Where are the magnets?

[ms]

May 6, 2017 8:18:36 GMT -5 stratotarts said:

Where are the magnets?

The magnets are tiny neo "buttons", 1/8" dia., 1/32" thick; they go on the pole piece closest to the strings and fit against the pole piece inside the hole in the fiber washer.  You do not need much of a magnet with a high permeability pole piece.  Installing the magnets lowers the inductance by a bit under 1%, so the permanent field has moved the material a small amount along its hysteresis curve, but no where near saturation.  There is a thin plastic cover over the pole pieces and magnets (that is, covering the top of the pickup) because neo corrodes quickly even when plated when exposed a guitar playing environment.

Using two of these magnets causes the beginnings of stratitis.  A somewhat larger magnet would probably be OK.

This single pickup guitar has two 250K vol. controls and two 500K tone controls, limited to 10K minimum with series resistors, no caps.  The switch selects either vol/tone combination.


So, for example, you can have clean and dirty presets. where the clean preset uses a low volume setting and  high tone setting (with these very bright pickups) to make up for the loss of  highs due to the low volume setting.  The dirty preset uses full volume, but much reduced tone to get the right Q for good distortion.

[ms]

Here is a single coil sized High Q pickup that cancels hum and can be used in the neck position without D-G cancelation when bending. (Of course the Q can lowered easily to make a normal pickup keeping the other qualities.) The pickup also has satisfactory output level, almost 200 mv peak-to-peak from a single string
This post describes how the pickup is made. The magnetic parts used in constructing the six small bobbins are shown in the first attachment.

They are a ferrite pole piece, actually a “bead”,0.2”x.0437” (73 material, but a somewhat lower permeability material would work as well), a small ferrite “brick”, (0.567”x.3”x.135”), and a rectangular neodymium magnet, 3/16”x3/8”x1/32”. The ferrite is from Amidon, and the magnet from K&J. These are assembled into a bobbin as shown in the second attachment.

When the pickup is assembled, the magnet points across the strings to help spread the field in that direction, while confining it in the direction along the string. The brick points along the string, increasing the isolation between coils. Three coils are made with one magnetic polarity, three with the other.

The third attachment

shows a coil after winding on 3200 turns of #41 wire. An additional pole piece has been used to help confine the magnetic flux. (This raises inductance and increases sensitivity.) This is not the ideally shaped ferrite piece for this purpose, but it is all that I have that is the right length.

One more photo is in the next post.

[ms]

The final photo shows one coil mounted in the cover.

The coil is located as close as possible to one side. The next two are mounted in the same orientation, and these three all have the same magnetic polarity. The final three are rotated 180 degrees and mounted as close as possible to the other side.

The cols are wired s, f-s, f-s, f-f, s-f, s-f, s.

The separation between the two sets along the strings is just under .25”. This exceeds pole diameter of .2”, and so when the D or G string is bent to the middle, the oppositely oriented magnetization regions on the string from the two pole pieces are separated enough to avoid significant cancelation.

In the next post I will discuss some of the “artifacts” often associated with split type pickups and show that they are not a problem with this pickup.

[antigua]

Thanks for sharing this concept. High Q and hum cancelling are obvious selling points for this design, although with a high Q you have to think about the actual user controls that would be required to tame it to a guitarist's liking. Usually guitarists think you turn all the controls to 10 for the best tone, but with a high Q, the best tone might be at "8", and they're not going to know that, and might not discover this on their own. 

I like to idea of a pickup that has dedicated controls for Q and peak frequency, I think it can be done with PCB sized components beneath a passive pickups. There was one such product, the "Dialtone":



but this product apparently failed. The inventor claimed he was no longer interesting in marketing the idea. It might well have been a marketing failure more than a product failure, though. The biggest problem IMO is that it didn't look vintage, but also there was Strat / Tele version, and I'd gamble that Strat / Tele  players are more fussy with tone knobs than users of humbuckers. I think this high Q concept couples with discreet controls could hold a lot of appeal to Fender players. 

[ms]

Jun 27, 2017 23:21:43 GMT -5 antigua said:

Thanks for sharing this concept. High Q and hum cancelling are obvious selling points for this design, although with a high Q you have to think about the actual user controls that would be required to tame it to a guitarist's liking. Usually guitarists think you turn all the controls to 10 for the best tone, but with a high Q, the best tone might be at "8", and they're not going to know that, and might not discover this on their own. 

I like to idea of a pickup that has dedicated controls for Q and peak frequency, I think it can be done with PCB sized components beneath a passive pickups. There was one such product, the "Dialtone":



but this product apparently failed. The inventor claimed he was no longer interesting in marketing the idea. It might well have been a marketing failure more than a product failure, though. The biggest problem IMO is that it didn't look vintage, but also there was Strat / Tele version, and I'd gamble that Strat / Tele  players are more fussy with tone knobs than users of humbuckers. I think this high Q concept couples with discreet controls could hold a lot of appeal to Fender players. 

Well, that product probably looks pretty bad to most guitarists.  It seems likely that there is some sacrifice in performance.  If it is hum bucking, both coils cannot receive signal, and so it has lower output.  Balancing out hum with two coils that would have to be so different is more difficult.

As for the controls:A guitar tone control is a Q control, except nearer to zero where the big C lowers the resonance, and as I have said before, throwing out the C and making it a better Q control is the thing to do.  So I do not see why you need a Q control on the pickup.  If you just have a resonant frequency control, I think that could be constructed entirely in the mounting frame, using tiny chip capacitors without affecting the construction of the pickup itself.  In that case, the frame could be the product, where you would likely use a pickup of your choice more towards the under-wound side so that you have more room to lower the frequency with C.

[ms]

Jun 27, 2017 23:21:43 GMT -5 antigua said:

Thanks for sharing this concept. High Q and hum cancelling are obvious selling points for this design, although with a high Q you have to think about the actual user controls that would be required to tame it to a guitarist's liking. Usually guitarists think you turn all the controls to 10 for the best tone, but with a high Q, the best tone might be at "8", and they're not going to know that, and might not discover this on their own. 

It is easy to throw away the High Q if that is not of interest:

1.  Use smaller wire.  The result is an even smaller coil or, use more turns and get more output and a lower resonance.

2. Add series resistance. Kind of like 1., but without the advantage of a smaller coil.

3.  Add parallel resistance.  This just pre-loads the pickup independent of the effect of the pots.

4. Use lower resistance pots.  Easy to do on a guitar with a full pick guard.  Not so easy with a guitar requiring pots with a long threaded bushing.  You can get them in any value, but other than the standard values, you usually need to order many.

[ms]

Let's list the artifacts a pickup with this split design can have:

1.  The wound and plain strings are at slightly different locations along the strings, and so the sounds do not quite match.

2.  Pickup from the G string by the D coil (and vice versa) has opposite polarity (or out of phase, as some would say).

3. When bending, for example to the middle, the D or G string is picked up by both coils equally, but the two coils are in slightly different locations along the string, leading to slightly different harmonic contents.

There might be more.

A discussion of these artifacts requires the use of the effects of pickup location and string sampling aperture.  The standard reference on these topics is Tillman: www.till.com/articles/PickupResponse/.  This is good on the location, but the apertures he assumes are  way too big, and thus very much underestimate the levels of audible higher string harmonics.  Obviously, a pickup cannot have an aperture larger than the effective width of string magnetization, that is, the width with string magnetization such that a significant component of of B field is produced through the coil and pointing along its axis (Law of magnetic induction). The next post will review this topic and show the effects.

[antigua]

Even though the tone control can serve as a Q control, I'm thinking of this as a "trim" control, so that your primary tone control can be parked at "10", and then a separate control can finely tune the Q factor separately.

Figuring out where to cram these controls is probably a bigger challenge than anything else. An alternative would be to have a tone control with a detent, so that it has a "sweet" center spot that is easy to find. The Fender passive TBX control pretty much does this exact thing. The relative lack of popularity of these things goes to show the extent to which guitarists are married to arbitrary conventions, which is partly why I'm into studying existing designs and no so much interested in pickup design. Good ideas fall on deaf ears. It bothers me that people accept what they do, and will wax poetic about for pages on end, without really caring to understand how it works at a fundamental level.

[Charlie Honkmeister]

Jun 28, 2017 13:34:52 GMT -5 antigua said:

Even though the tone control can serve as a Q control, I'm thinking of this as a "trim" control, so that your primary tone control can be parked at "10", and then a separate control can finely tune the Q factor separately.

Figuring out where to cram these controls is probably a bigger challenge than anything else. An alternative would be to have a tone control with a detent, so that it has a "sweet" center spot that is easy to find. The Fender passive TBX control pretty much does this exact thing. The relative lack of popularity of these things goes to show the extent to which guitarists are married to arbitrary conventions, which is partly why I'm into studying existing designs and no so much interested in pickup design. Good ideas fall on deaf ears. It bothers me that people accept what they do, and will wax poetic about for pages on end, without really caring to understand how it works at a fundamental level.

   Wow,  Antigua,  that's about three interesting and "don't get me started" topics rolled into one.   Couple of comments:

   1.  Really moderate resonant peaks with 5-8 dB of peak before rolloff, is typical of vintage passive guitar pickups and electronics, through a cable, into about a 1 Meg amp input resistance.   I've been experimenting with Q and found that keeping Q relatively constant in that range, with series and shunt resistive loading,  and just moving the resonant peak around with varying shunt capacitance, works pretty well for humbucker style tonality all the way into some Gretsch/Rick/Strat/Tele bridge-y type sounds.  

The problem with higher Q is that it also means a faster rolloff after resonance.  And, with most amps and guitar speakers,  the speaker starts rolling off about 5 KHz so unless you have a relatively high resonant frequency (think Strat and Tele),  high Q works against getting at least some of those highs out, for "clarity" in cleaner to light crunch styles.  Once you get into moderate to heavy distortion,  you should limit bandwidth, but I don't think a high Q low-mid peak helps you there except for maybe extreme metal leads with lots of FX.    
   
If you're staying with trying to emulate "traditional"  vintage tonality, staying with low to moderate Q values is the best course.  Having a Q control which controls resistive loading is possible but I have a hard time translating that into something that a guitarist can make sense of and use in a live performance.   That may have been a factor in the demise of the Dialtone product.  Just because it's infinitely tweakable doesn't guarantee that it's musically usable or desirable for the player.   

I think that a single-knob variable resonant frequency tone control with more-or-less constant Q is a lot more useful and makes sense as a performance control that guitarists already think they understand,  therefore don't object to.  Of course,  I've spent the last couple of years developing exactly such a control,  so please take my comments with that in mind.   

2.  That being said, having a way to "trim" Q  could be useful in a scenario where you wanted to emulate several pickups, or wanted to compensate for different series/parallel combinations of pickups needing different loading, depending on the series/parallel combination, to sound the best.    But that would be done better with switch poles, rather than having a Q "trim" control as a separate pot on the instrument IMHO.   



-Charlie

 

[Charlie Honkmeister]

I just wanted to make sure that in my last post I didn't sound contradictory to some of my other posts on this thread. 

Having a pickup with a high intrinsic or maximum Q is a good thing because it means that from an engineering standpoint it is more efficient  (there are maybe fewer eddy losses and lower wiring resistance,  and/or a better magnetic circuit) and also it's more flexible in terms of being able to voice it externally over a wider range with resistive and capacitive loading. 

[antigua]

Jun 28, 2017 15:59:46 GMT -5 Charlie Honkmeister said:

Jun 28, 2017 13:34:52 GMT -5 antigua said:

Even though the tone control can serve as a Q control, I'm thinking of this as a "trim" control, so that your primary tone control can be parked at "10", and then a separate control can finely tune the Q factor separately.

Figuring out where to cram these controls is probably a bigger challenge than anything else. An alternative would be to have a tone control with a detent, so that it has a "sweet" center spot that is easy to find. The Fender passive TBX control pretty much does this exact thing. The relative lack of popularity of these things goes to show the extent to which guitarists are married to arbitrary conventions, which is partly why I'm into studying existing designs and no so much interested in pickup design. Good ideas fall on deaf ears. It bothers me that people accept what they do, and will wax poetic about for pages on end, without really caring to understand how it works at a fundamental level.

   Wow,  Antigua,  that's about three interesting and "don't get me started" topics rolled into one.   Couple of comments:

   1.  Really moderate resonant peaks with 5-8 dB of peak before rolloff, is typical of vintage passive guitar pickups and electronics, through a cable, into about a 1 Meg amp input resistance.   I've been experimenting with Q and found that keeping Q relatively constant in that range, with series and shunt resistive loading,  and just moving the resonant peak around with varying shunt capacitance, works pretty well for humbucker style tonality all the way into some Gretsch/Rick/Strat/Tele bridge-y type sounds.  

The problem with higher Q is that it also means a faster rolloff after resonance.  And, with most amps and guitar speakers,  the speaker starts rolling off about 5 KHz so unless you have a relatively high resonant frequency (think Strat and Tele),  high Q works against getting at least some of those highs out, for "clarity" in cleaner to light crunch styles.  Once you get into moderate to heavy distortion,  you should limit bandwidth, but I don't think a high Q low-mid peak helps you there except for maybe extreme metal leads with lots of FX.    
   
If you're staying with trying to emulate "traditional"  vintage tonality, staying with low to moderate Q values is the best course.  Having a Q control which controls resistive loading is possible but I have a hard time translating that into something that a guitarist can make sense of and use in a live performance.   That may have been a factor in the demise of the Dialtone product.  Just because it's infinitely tweakable doesn't guarantee that it's musically usable or desirable for the player.   

I think that a single-knob variable resonant frequency tone control with more-or-less constant Q is a lot more useful and makes sense as a performance control that guitarists already think they understand,  therefore don't object to.  Of course,  I've spent the last couple of years developing exactly such a control,  so please take my comments with that in mind.   

2.  That being said, having a way to "trim" Q  could be useful in a scenario where you wanted to emulate several pickups, or wanted to compensate for different series/parallel combinations of pickups needing different loading, depending on the series/parallel combination, to sound the best.    But that would be done better with switch poles, rather than having a Q "trim" control as a separate pot on the instrument IMHO.   



-Charlie

I agree with all your points. I actually do modify both the peak and the Q in live jamming, with guitars that I've wired to control both, but I had to "learn" how to use the new tools, and make then serve a musical purpose. Most people are playing standards, or original music that is derivative of standards, which is why 50's tech always seems to hit the spot. Most guitarists see all this as a solution in search of a problem.

Marketing any new idea is therefore near impossible, but that being said, I would market a Q / f variable pickup to all the guys out there who buy ten sets of pickups, who are somehow never satisfied with what they have, and boast that if this doesn't make them happy, nothing could. I'd probably target the Fender market, as the Gibson guys are stuck on the "perfect PAF", they'd ask if your Q/f pickup was wound on a Leesona winder and had a wood spacer, that's where they're at.

The prototype here is promising because it would fit the Fender foot print, although it would be a tough cross sale with Zexcoil, and Lace, and other non-stack noise-reduced pickups with an atypical appearance. 

[ms]

Of all the changes you might make to guitar controls, setting it up so that the tone control normally runs below ten is the simplest to learn to use because it just means that you use the guitar more like all your other audio devices that have tone controls.  Also I find that a properly functioning tone control, which adjusts Q over the whole range, is more useful than a resonant frequency adjustment if you have just one of the two.  This is especially true when the pickup is an sc replacement since you can set it up so that the res. fr. is at the top of the system response and then have a wide range of tones available with a single control.  But preferences do vary, of course.

[antigua]

Jun 29, 2017 7:00:43 GMT -5 ms said:

Of all the changes you might make to guitar controls, setting it up so that the tone control normally runs below ten is the simplest to learn to use because it just means that you use the guitar more like all your other audio devices that have tone controls.  Also I find that a properly functioning tone control, which adjusts Q over the whole range, is more useful than a resonant frequency adjustment if you have just one of the two.  This is especially true when the pickup is an sc replacement since you can set it up so that the res. fr. is at the top of the system response and then have a wide range of tones available with a single control.  But preferences do vary, of course.

I agree that the Q control is more useful, it's better than the volume control of modulating how present the guitar is in the band's mix. If I start singing and the chords are brighter than my voice, I reach down and just roll back the tone until the guitar is not outshining me. Same thing if the drummer wants to do a solo, etc. With some practice, controlling the Q is a constant benefit. I mostly use the frequency (cap) control to get different shades of square wave distortion, or to render a P-90 like clean warm out of Fender single coils, but that's not something I do constantly.

[Charlie Honkmeister]

Jun 29, 2017 7:00:43 GMT -5 ms said:

Of all the changes you might make to guitar controls, setting it up so that the tone control normally runs below ten is the simplest to learn to use because it just means that you use the guitar more like all your other audio devices that have tone controls.  Also I find that a properly functioning tone control, which adjusts Q over the whole range, is more useful than a resonant frequency adjustment if you have just one of the two.  This is especially true when the pickup is an sc replacement since you can set it up so that the res. fr. is at the top of the system response and then have a wide range of tones available with a single control.  But preferences do vary, of course.

To be honest, I'm not totally convinced yet, by listening and playing, that variable resonant frequency/constant Q is the absolute best way to go on Strat-type instruments,  although I'm completely convinced on 2-humbucker guitars.  So I appreciate your comments on using a conventional tone control with SC type pickups. 

One target for using a variable resonant frequency on a 3-SC Strat type guitar is to get the pickup resonance down to about 2 KHz,  to be able to hit the Hendrix or SRV tonality with a Strat and still be able to "dial back up"  and do all the other tones a Strat can do.   Jimi did it (also by touch and technique, of course) by using a 50 foot crappy cable with a lot of capacitance, according to Bill Lawrence, and SRV did it (I think) by of course, touch and technique,  but also by really large gauge strings, the tone control,  and also paralleling to multiple amps, which added a lot of capacitance.    

I'm almost finished building a Strat-type instrument with Zexcoil Z-Series pickups, which has the variable resonant frequency tone control on the 1st tone pot, and a neck/bridge no-load blend pot on the 2nd.  I'm eagerly anticipating testing and playing the new guitar and will be trying to discover whether conventional or "greasebucket"  tone control,  variable resonant frequency/constant Q,  or variable resonant frequency/increasing Q with frequency,  works the best.  Since the guitar has an onboard  buffer,  any of those schemes can be done fairly easily without cable capacitance rearing its ugly head.   

[ms]

Jun 29, 2017 11:39:11 GMT -5 Charlie Honkmeister said:

To be honest, I'm not totally convinced yet, by listening and playing, that variable resonant frequency/constant Q is the absolute best way to go on Strat-type instruments,  although I'm completely convinced on 2-humbucker guitars.  

Yes, there can be a problem with the  sound if the Q is too high as the frequency is lowered.  One solution that works in some cases is to switch an RC in parallel, rather than just a C.  That way you can adjust the R you use with each C.

[ms]

We need to look at all the factors that affect the frequency response of the electric guitar system in order to asses the effects of various possible “artifacts” that might affect the performance of the pickup under discussion. For example the response is strongly affected by the electrical circuit the pickup is part of, the amplifier, and especially the speaker. But there is another side to this, and this is how the pickup samples the strings. And so it is necessary to understand something about the vibrating string to see how this happens. Tillman’s work (http://www.till.com/articles/PickupResponse/) is an excellent reference for understanding the displacement of the various string harmonics, and he shows how the aperture of the pickup can act as a filter to reduce the higher harmonics, and thus frequencies. But fails to properly include the filtering effect of the pickup aperture. He assumed that the aperture is somewhat wider than the width of the coil, but is actually more like the width of a pole piece. This means that he underestimated the frequency response by about a factor of four, and we need to get this right.

Following Tillman (http://www.till.com/articles/PickupResponse/), consider this equation:

yn = An*n*sin(pi*n*x),
where:
yn: the displacement of the string from its rest position for the nth harmonic.
x: the location along the string, from 0 to 1 (unit scale length); x=0 at the bridge.
n: the harmonic number, 1, 2, 3….
An: the amplitude of the nth harmonic

The argument of sin is in radians where pi radians is 180 degrees.

Let all the An be one for simplicity.

If we let n be a set of fixed numbers (1, 2, 3, 4) and plot y with x as a variable we get the first attachment.

Also plotted as vertical lines are typical positions of a bridge and neck pickup. So look at the relative string displacements of the first four harmonics at the bridge pickup. These displacements increase with increasing harmonic number. This plot should be compared to Tillman’s equivalent plot to make sure that they agree.

On the other hand, if we set x to a fixed number (where the pickup is located) and let n vary, we can get a number describing the displacement for each harmonic. But this number is for an x value on the string, not for a pickup, which receives signal over some range of x values, the aperture. We can get the pickup response by summing over many closely spaced points with in the range, multiplying each of the point responses by an appropriate weighting function. This summing process reduces the response at the higher harmonics.

Tillman assumed a 1 inch square aperture fir a single coil pickup. The assumption here, based on field measurements (which do not have perfect accuracy) is that at the edge of the pole piece, the field is 80% of its value at the center, and that it falls of rapidly with increasing distance from the center. Thus it is fairly well represented represented by a Gaussian function, with the scale constant set to achieve the 80% value based on the pole diameter supplied to the software.

The computations for neck and bridge pickups are shown in the next two attachments.

Computations were made at the harmonics and the results plotted with filled circles. The circles are connected by lines to aid in visualizing the trends. The differences between the two plots are as expected for the two pick locations, the same as used by Tillman. The common feature is the gradual decrease in the sampled levels. This is the result of the pole width. The roll off is about 4 db at the 54th harmonic. This is about 8KHz with the D string, and is greater than the 5 KHz limit set by the guitar speaker. Thus, it is the electrical circuit that is determining the frequency limitations, not the string sampling as implied by Tillman’s results, which are about a factor of four lower.