Electric guitar pickup behavior studies

[genmce]

I found this searching for the influence of inductance measurement of pickups and placement.
I still have questions. I will post a new thread.

I don't know if this should go in reference section or not, please move if mod deems appropriate.

Here is the link

[JohnH]

Thanks for that link. Ive been watching out for that kind of information for a few years and that is one of the best pieces of work that i have seen, with both theory and testing. I have been putting that into a spreadsheet GuitarFreak, and its in our reference section. It could help you compare tone effects due to inductance, that you asked on the SP thread.

A short answer on that is that inductance is the most important electrical property affecting guitar tone, and the higher it is the thicker the tone will be.

[Deleted]

too bad the pdf is a scanned version of the paper and not searchable.

[genmce]

Aug 5, 2015 17:30:30 GMT -5 JohnH said:

Thanks for that link. Ive been watching out for that kind of information for a few years and that is one of the best pieces of work that i have seen, with both theory and testing. I have been putting that into a spreadsheet GuitarFreak, and its in our reference section. It could help you compare tone effects due to inductance, that you asked on the SP thread.

A short answer on that is that inductance is the most important electrical property affecting guitar tone, and the higher it is the thicker the tone will be.

Does thicker (higher inductance) result in more or less quack? Do I want mismatched inductance for more quack?

Yes - I should look at GuitarFreak again. I don't know how to interpret or what will show me increased quack in gf? Is it the hump mentioned here, if not, what should quack look like in gf?

I am after the lay rule of thumb when choosing which pup to put where. My understanding is the highest dc resistance should be in the bridge or is the dc resistance not helpful or is that more about balancing output? Then should the thickest (highest induction) pickup be in bridge?
Is getting pickups with balanced inductance important?

[newey]

My understanding is the highest dc resistance should be in the bridge or is the dc resistance not helpful or is that more about balancing output?

Yes. Yes. And, yes.

The search for the elusive "quack", or a search to maximize a property known as "quack", is not something that can done simply by comparing electrical characteristics, for a couple of reasons. First, the "quack" itself can only be subjectively defined (trying to find agreement between guitarists as to what "quack" is will quickly lead one down the rabbit hole . . . )

Second, whatever "quack" is, it is dependent upon the interaction of two single-coil pickups wired together in parallel, and dependent upon the harmonic nodes along the guitar string where those pickups are placed. In a Strat-type design, the pickup placement is largely a given (although I have often wondered if some of the imported Strat "clones" may vary the placement a bit, probably only by a few mm one way or another. Also, if the scale length isn't 25.5", all bets are off).

It is perhaps helpful to bear in mind the history of the Strat here. "Quack" wasn't a feature on the original Strats, as they had a 3-way selector switch, and only one pickup was in operation at a given time. Folks found that they could balance the lever in between positions 1 and 2, and in between 2 and 3, and thereby have either N + M or M + B. That led to people cutting notches in the lever switch, making a 5-way switch out of the 3-way.

At that point, there was already discussion of "Quack" at positions 2 and 4 of the now-5-way switch. At this time, middle pickups were not yet RWRP, and the idea of a "matched set" of pickups with varying resistances/inductances still lay in the future as well. All three pickups on vintage Strats were identical. So, "quack" does not depend upon having a RWRP middle pickup, and does not depend upon having a higher DC resistance or inductance in one or the other pickup.

The "notch" mod became so common that eventually, in the early 1970s, Fender made a factory 5-way switch standard, and began specifying a RWRP middle pickup. It was around that same time that Fender began providing a bridge pickup with a bit higher resistance (and inductance) to better match output levels. Later still, "matched sets" of pickups, where all 3 were different, became available in the aftermarket.

Now, one could argue that differences between the pickups, whether RWRP or differing inductances, might increase/decrease the perceived "quack"- but now we're headed down that rabbit hole again . . .

Many guitarists are in search of that "vintage" Strat tone, and don't want a RWRP middle, for example (they want the vintage noise along with the vintage tone, apparently). Those folks will tell you that "quack" lies down that road. So, go figure.

[JohnH]

Yes since we don't really know what quack is, and can't usually agree on what it sounds like, we can't easily say what it looks like or how to measure it. But I think it has a lot to do with high frequency interactions between harmonics from two pickups. I've been trying to capture this, see the last post about guitarfreak, which is not online yet.

[genmce]

newey - sounds like I'm trapped in the rabbit hole!

JohnH - Ok so when do you plan to post it?
I have downloaded newest gf version 5.01 and am trying to understand - it has been a while since I messed with it gf - it has changed significantly. When is 5.1 due out?

[JohnH]

Aug 7, 2015 12:48:45 GMT -5 genmce said:

newey - sounds like I'm trapped in the rabbit hole!

JohnH - Ok so when do you plan to post it?
I have downloaded newest gf version 5.01 and am trying to understand - it has been a while since I messed with it gf - it has changed significantly. When is 5.1 due out?

There's no deadline for GuitarFreak, it's just an occasional project for me though Im happy to share it. All the versions so far have just focussed on the electrical interactions, which (provided I have done them correctly), are on a fairly solid basis of theory. You can use this, or other more generally modelling programs such as 5Spice, to make comparisons between electrical circuits, and it can lead to some useful decisions about component values that are helpful.

But they miss out on many very important aspects of string vibration and pickup location and characteristics, which we know practically from experience. The reference that you posted, and a couple of others such as Tillman go into these. But I have never seen anywhere, examples of putting the string vibration and pickup theory into the same analysis as the electrical theory. That's what I'm playing with now, but I am cautious about claiming too much about it, other than that it is interesting and there is much to learn. Even finding a way to produce a graph that is helpful without being over-simplified, is not obvious, when real guitar signals are a confusing mess of varying frequencies. I intend to post the new version soon.

However, I think the results that I'm getting are doing at least somewhat of a better job of assessing combinations of pickups, than are the pure electrical analyses. At least the following are covered and are represented in the results, where without using the new vibration and pickup module, they are not: These are:

The effect of pickups nearer the bridge having a lower bass response
The sensing width of single and humbucker pickups, rolling off high frequencies (in addition to electrical factors)
Out of phase effects and cancelling of harmonics when two pickups are selected

One big problem is getting the right data to go into such an analysis. Its based on bits and pieces of what is available, and some speculative guesswork.

With regard to quack etc, what we know is that whatever it is, you get it from B+M or N+M combinations of single coils. It doesn't happen with humbuckers, and although B+N is another great single coil combination tone, we don't normally hear it described as 'quack'

So, from my post on GF, here is a best-guess plot showing the six basic ways to use a B and an M single coil:

If you had a 6-way rotary, you could have these on a knob.

The thick red and dark blue lines represent the pure single tones, and you can see how there is less bass in the bridge pickup. The  two out of phase combos are very evident by the strong bass cancellation, and the series settings show the higher output, but with earlier high roll-off (which is mainly due to the electrical aspects)

The usual 'quack' position is labelled 'B+M para' in pale blue. What is evident here is the strong high peak, rising above the others in the high treble range, and also, the general midrange is slightly less than the singles. This is due to, for each individual note where one pickup tries to form a peak, the other is slightly cancelling it. The plots are an envelope of all strings and all notes at once. Individual notes fall within and below this envelope.

So from that, you can tentatively see how keeping that high peak available may be a key aspect to preserving and enhancing 'quack'. Once the pickup position and type is decided, that is mostly an electrical effect. It will be promoted by less loading in terms of higher pots values and shorter cables, and single-coil pickups that are not too 'hot'.  If you want to explore that analytically, just electrical modelling will be OK. But Id be happy to send you a new 'Beta' version of GF if you want to try what I am still experimenting with. But if your priority is to wire up a great guitar rather than get into a lot of theory, then Id suggest not to base too much on these graphs and calculations.
   

[JohnH]

There is a rich vein of random info to explore on this subject! Here is a bit more investigating into the 'Nature of Quack'

The graphs above are based on an envelope response of all strings, and taking the resulting max values within a range of -2/+3 semitones to give a smooth shape. But if that max function is switched off, you get this, envelope for all strings:



The B+M tone is revealed to have many more variations in the upper register than either single pickup.

Now consider just one string (3rd string G in this case):



Phase cancellations are causing about an octave in which there is almost no output from this string. In this case it is roughly between about 1200hZ and 2400hZ. Each of the other strings will be doing something similar, scaled to their own frequencies.

I don't know what but that all has to mean something!

[genmce]

Wow - there is, that comb filter effect (I believe someone mentioned that in the quack thread)! Very clear to see.
I am very excited to see this. I am very interested to see this. I did not mean to pressure you ton complete this, I was measly curious.
To me - your model shows the "quack" very clearly. So how to use this as a model to predict the amount and location of those cancelations based on
circuit components, pick up values, etc.

I know you have said the quack is more pronounced given no load.
To my ears - as I fade tone down from 10 quack is still there, albeit less intense.
Since the effect seems to lie in that range 1khz - 2.4khz and I find tone open or no load a bit
too harsh lately... there must be a way to get more control of this, in a conscious way.
Could passive components be added to intensify/modulate the effect, ala dial a duck but more as a way to move the comb arround the spectrum.
I don't mean anything drastic - inductor/caps on a push pull or some sort of variable inductor.
Can you add inductance to a pickup to make it perform differently/possibly better, for this purpose, then of course to be able to switch it out of circuit
So did your curves come from standard single coil models?
What happens to the model when the b = 7.1k 2.6h and m = 6.4k 2.4h just for an example (since that is what is in my guitar now)?

[JohnH]

Interesting, here is the best I can figure out for the 3rd string on your guitar, assuming the pickups are placed as a Strat:

Its quite similar to the previous one (which was based on not much different to yours)

An assumption that I had to make along with the data that you provided, was that the relative output of the bridge pickup would be higher, approximately in proportion to the ratio of inductance and/or resistance. More turns = more L and R and more output, I guessed +0.8db.

The comb frequency effect is not related to electrical effects but to the physical placement of pickups and the tuning frequency of the open string. So I don't see a way of shifting it with an electrical control, but you could maybe control the intensity of cancellations to some extent.

[genmce]

Seems like a subtle changes, that I really had to hunt for.

A thought occurs to me related to hss, as I am about to start stratsp hss, if this quack is related more to pickup position, would spliting the humbucker to use the coil closer to the middle pickup, increase the comb effect?

[JohnH]

Thanks for these questions. They help me test out my system to see what kind of issues it can help with.

I have an HSS strat, with 6.2k Texas specials and a 8.4k humbucker. Here is a group of plots relating to that, which take account of the two coil positions of the split humbucker:



The thick red, dark blue and green lines are the combinations with M, and B coil1 is nearest to the bridge.

There's some significant difference to see and hear there. Of the two split combo settings, only one will be humcancelling. On my guitar, I can get either full B+M or split B+M, using the bridge coil furthest from the bridge. There is a small but significant difference between those two.

[genmce]

Thanks, that looks great!
It seems to suggest proximity of the coils does influence the comb effect. Using that closer coil2 with m on position 2 would be best for my quacky goal.

So - see my StratSP hss post. I see polarity changing but I do not see where humbucker coil is being selected to be different from switch position 2 (bm) and 3 (nb).

[JohnH]

Im pleased that these plots give some helpful insight.

Last night I started investigating another key tonal factor which has so far been missing, which is the relative balance of harmonics as you play across the strings, and how this affects overall frequency response.

The online version of GF 5.01 and earlier assumed a flat input response spectrum and only captures electrical effects.

The beta version that Ive been using recently combines that with the effects of pickup width and location, but still assumes that the strings are vibrating such that there is equal string velocity for each harmonic. (=linear reduction of amplitude as frequency increases).  String velocity determines voltage output from pickups.

The ref. that you posted implies that higher harmonics fall away quicker than that and can also take account of picking position. I think the current beta GF wiold imply picking extremely close to the bridge.
 
The physics and maths gets very involved and im beyond the range of what I can properly understand. So I ran some physical tests recording a guitar via a buffer into a mixer set with flat EQ. There does seem to be a further  rolloff of higher frequencies visible in spectrun plots that needs to be represented somehow before analysis can approach reality.

Work in progress...

[genmce]

As I finished the strat sp hss - and remembered that I had not balanced pickup height...(hanging fragment)

Pup height (distance from vibrating string) will have a direct effect on pup output level, yes?
You mentioned having to estimate output level on plots above.
I have read, in several other threads, about increasing quack by lowering middle pup.
I don't know how one would quantify the amount of lowered output.
I would be interested to see if the comb is more or less pronounced (mb) related to m output level (raising/lowering m pup). I know my focus has been about quack, however, this information could apply to help make conscious decisions when wiring. In the "dial a duck" newey mentioned, the guy put volume pot on the m - what might that look like, modeled?
I may need to read that article I posted, I have only skimmed it, because most of it was over my head.

[JohnH]

Here is a plot in which the M pickup has less output in 1db steps. The B humbucker is split to just the coil nearest to M, and would have a nominal relative output of -1db is that mode, while the M steps from, +2 to -5db. It seems the less output it makes, the more the hole in the upper midrange drops away. I don't know how to properly relate output to pickup height, you could test I suppose.

[genmce]

This graph is very interesting.

I note the difference between -4 and -5db is significant.
The bottom of the cuts seem to move as well. -4db seems to have sharper cuts, though at different frequencies, than the -5db, suggesting a sweet spot...

This implies to me that a volume pot across the middle pickup (dial-a-duck)might give some control over the max cut, frequency and an easier ability to find the sweet (quack) spot, whatever that is. This may be an oversimplification as the pot will have an interaction as well.

[JohnH]

You are right about a volume pot affecting the results. So far, all my models with two pickups assume they are directly connected to each other before heading off towards master tone and volume circuits.

Id like to extend the system to be able to deal with things like twin volume controls, blending circuits and partly bypassed combinations. There's a heap of extra maths that would be involved in that. I would do it by putting in enough extra components into a single model, so that any such arrangement can be modelled by using high or low values to negate unwanted parts. Im pausing to consider what that might entail and how to present and control it so that it is still a useful and not too complicated tool.

In interested in any comments and suggestions as to what kind of arrangements might be most usefully modelled and how they might be presented.

[genmce]

Aug 12, 2015 16:36:36 GMT -5 JohnH said:

You are right about a volume pot affecting the results. So far, all my models with two pickups assume they are directly connected to each other before heading off towards master tone and volume circuits.

Id like to extend the system to be able to deal with things like twin volume controls, blending circuits and partly bypassed combinations. There's a heap of extra maths that would be involved in that. I would do it by putting in enough extra components into a single model, so that any such arrangement can be modelled by using high or low values to negate unwanted parts. Im pausing to consider what that might entail and how to present and control it so that it is still a useful and not too complicated tool.

In interested in any comments and suggestions as to what kind of arrangements might be most usefully modelled and how they might be presented.

Hey - what about a modular approach, much like you have going with tone control setups?
A different tab to select some preset ideas. Two volume controls or volume control on each pickup?
Default being a more stock LP or Strat.

With an advanced tab to allow things you mention above.

I really like those spectrum graphs based on a note.

[helmuth]

JohnH: How did you get these curves?
Did you physically measure them or are they computer calculations?

For pickup measuring technique look here:
www.planetz.com/wp-content/uploads/2014/10/Pickup_Measuring_Technique.pdf

or
www.buildyourguitar.com/resources/lemme/index.htm

regards, Helmuth

[JohnH]

Feb 17, 2016 5:56:46 GMT -5 helmuth said:

JohnH: How did you get these curves?
Did you physically measure them or are they computer calculations?

For pickup measuring technique look here:
www.planetz.com/wp-content/uploads/2014/10/Pickup_Measuring_Technique.pdf

or
www.buildyourguitar.com/resources/lemme/index.htm

regards, Helmuth

Dear Helmuth

Thankyou very much for joining our forum. My plots are calculated responses, made using a spreadsheet called GuitarFreak:

The basis of it is your work, which I reference there. So thanks again! Your writing was very enlightening and inspiring.


I've had several versions of the spreadsheet, starting with the pure electrical responses based on simple pickup representations. More recently, I started to add the effects of string vibration harmonics and pickup location - still very much under development.

regards
John