Duncan Designed Mustang Pickups, Analysis & Review

[Antigua]

Driving the pickup with a coil does not accurately show the effects of the magnetic circuit. As to how far off from reality that is, I do not know, I suppose loading a pickup with poles which have yet to be magnetized would give some clues, but would likely raise more questions then answers. Could be useful though.

A guitar string is not a significant factor in the magnetic circuit, because it is very small and very far away from the pickup. It's presence probably doesn't alter the inductance of the pickup more a fraction of a millihenry. That's something that could be determined with an LCR meter.

A tone wheel is how the Hammond B3 and similar organs generated sound, pretty much a gear spun over the pickup, the teeth of the gear interact with the field similarly as a string would. It is not a perfect analog of a string, but closer then a coil and it allows the frequency to be swept since pitch is a function of rpm.

Interesting, not a bad idea, but as you say, still somewhat unlike a guitar string, and it would a cumbersome test rig to create. I also have issues with electric motors producing noise. I'm not sure how they deal with that in a Hammond.

I have been mulling over how to accurately judge the effects of the magnetic circuit for awhile now and the tone wheel is my best solution yet. I have also considered winding a coil right onto a offcut of a string, has some advantages especially in labor, but still not ideal.

I think it's best that the test coil have an air core, the keep the reluctance path as high as possible between the test equipment and the thing being tested.

[oid]

A guitar string is not a significant factor in the magnetic circuit, because it is very small and very far away from the pickup. It's presence probably doesn't alter the inductance of the pickup more a fraction of a millihenry. That's something that could be determined with an LCR meter.

Sure, it is not a significant factor while at rest, pluck the string and we get output. It is at least as important as any other part of the circuit since if you remove it, we no longer have a working system.

Interesting, not a bad idea, but as you say, still somewhat unlike a guitar string, and it would a cumbersome test rig to create. I also have issues with electric motors producing noise. I'm not sure how they deal with that in a Hammond.

Unlike a guitar string, but more like a string then your air coil. Creating it is mostly an issue of the time required which is probably on par with the time it took you to create/calibrate your rig, or just an issue of cost and buying a tone wheel from an organ. The motor is not really an issue, there are plenty of quite motors out there, if there were not we would not have had tape recorders or record lathes and the world of music would be a very different place. Even if you were to use a noisy motor, that noise is easy enough to remove from the equation.. But as I said, if I do go down this path I plan on using a hand cranked grinder, which has no motor and I do not need to design/build a control circuit to get a frequency sweep.

I think it's best that the test coil have an air core, the keep the reluctance path as high as possible between the test equipment and the thing being tested.

But that is not accurate to reality, driving the pickup that way does not show what a pickup actually sees, you do not even need the magnetic circuit to drive the pickup this way, you just created an air coupled transformer were one coil has an magnetic core and one does not. A pickup and a string is a generator (transducer), not a transformer, your test is electrical to electrical and not mechanical to electrical. As to how accurate this is, I have found nothing concrete to address this and such studies as yours are fundamentally flawed without this information. After doing the tests it may indeed turn out that driving the pickup with a coil is close enough, or even accurate, as of now it is just an assumption.

In the end I think a combination of your test, testing without the magnetic circuit, the tone wheel and the string wound coil would provide a great deal of information. None of them are ideal, but the comparative results would be interesting, all four tests tell us different things about the whole and also tell us (hopefully) how reliable any of the other tests are.

[Antigua]

A guitar string is not a significant factor in the magnetic circuit, because it is very small and very far away from the pickup. It's presence probably doesn't alter the inductance of the pickup more a fraction of a millihenry. That's something that could be determined with an LCR meter.

Sure, it is not a significant factor while at rest, pluck the string and we get output. It is at least as important as any other part of the circuit since if you remove it, we no longer have a working system.

Interesting, not a bad idea, but as you say, still somewhat unlike a guitar string, and it would a cumbersome test rig to create. I also have issues with electric motors producing noise. I'm not sure how they deal with that in a Hammond.

Unlike a guitar string, but more like a string then your air coil. Creating it is mostly an issue of the time required which is probably on par with the time it took you to create/calibrate your rig, or just an issue of cost and buying a tone wheel from an organ. The motor is not really an issue, there are plenty of quite motors out there, if there were not we would not have had tape recorders or record lathes and the world of music would be a very different place. Even if you were to use a noisy motor, that noise is easy enough to remove from the equation.. But as I said, if I do go down this path I plan on using a hand cranked grinder, which has no motor and I do not need to design/build a control circuit to get a frequency sweep.

I think it's best that the test coil have an air core, the keep the reluctance path as high as possible between the test equipment and the thing being tested.

But that is not accurate to reality, driving the pickup that way does not show what a pickup actually sees, you do not even need the magnetic circuit to drive the pickup this way, you just created an air coupled transformer were one coil has an magnetic core and one does not. A pickup and a string is a generator (transducer), not a transformer, your test is electrical to electrical and not mechanical to electrical. As to how accurate this is, I have found nothing concrete to address this and such studies as yours are fundamentally flawed without this information. After doing the tests it may indeed turn out that driving the pickup with a coil is close enough, or even accurate, as of now it is just an assumption.

In the end I think a combination of your test, testing without the magnetic circuit, the tone wheel and the string wound coil would provide a great deal of information. None of them are ideal, but the comparative results would be interesting, all four tests tell us different things about the whole and also tell us (hopefully) how reliable any of the other tests are.

A guitar pickup has RLC characteristics. It's functionally equivalent to a low pass filter with a resonance. The question at hand is: is a moving guitar string required in order to identify the RLC values, or rather the low pass function, of a guitar pickup? The answer is no. In fact, the resonant peak can be found by simply frequency sweeping the coil directly, through it's two leads, and measuring the voltage across the pickup. The benefit of using an external magnetic field is that it will also reveal eddy current losses that manifest as a result of voltage having been induced by way of external flux change, regardless of where that flux change might originate from.

[oid]

The same is true for a dynamic mic, if you were to measure its response with your method you would get the nice clean response of a lowpass filter with a relatively high q formed by the coil and the load. Measure the same mic in an anechoic chamber with a good flat sound source and you got a nuanced frequency response full of peaks and valleys. You are ignoring the input and focusing solely on the output. The lowpass formed by the coil and load is applied to what the pickup sees and would only be accurate if the input is perfectly flat beyond the range of that filter.

The question of whether or not you can identify The lowpass function of the pickup has never been the issue, any decent text book will tell you the answer to that, my question has always been in regards to the input and the magnetic circuit, I have not even mentioned the filter except in response to you, this function of magnetic pickups is well understood and documented.

[Antigua]

The same is true for a dynamic mic, if you were to measure its response with your method you would get the nice clean response of a lowpass filter with a relatively high q formed by the coil and the load. Measure the same mic in an anechoic chamber with a good flat sound source and you got a nuanced frequency response full of peaks and valleys. You are ignoring the input and focusing solely on the output. The lowpass formed by the coil and load is applied to what the pickup sees and would only be accurate if the input is perfectly flat beyond the range of that filter.

The question of whether or not you can identify The lowpass function of the pickup has never been the issue, any decent text book will tell you the answer to that, my question has always been in regards to the input and the magnetic circuit, I have not even mentioned the filter except in response to you, this function of magnetic pickups is well understood and documented.

There must be a misunderstanding then as to the intention of the plot. It's sole purpose is to expose the resonant peak frequency, the Q factor, eddy current losses, as well as allow for the intrinsic capacitance to be calculated from the peak f.

But as to the further information you're talking about, a guitar pickup is not like a microphone. A microphone's response curve is dictated by physical considerations, such as what material the diaphragm is made of, or how thick it is, etc. A magnetic guitar pickup doesn't relate to the strings in a physical way, only in a magnetic way. There is a physical aspect in that the magnetic pull on the strings effects the way they vibrate, but that's beyond the purview of a pickup analysis, and can be inferred from the gauss measurement.

[oid]

There must be a misunderstanding then as to the intention of the plot. It's sole purpose is to expose the resonant peak frequency, the Q factor, eddy current losses, as well as allow for the intrinsic capacitance to be calculated from the peak f.

But as to the further information you're talking about, a guitar pickup is not like a microphone. A microphone's response curve is dictated by physical considerations, such as what material the diaphragm is made of, or how thick it is, etc. A magnetic guitar pickup doesn't relate to the strings in a physical way, only in a magnetic way. There is a physical aspect in that the magnetic pull on the strings effects the way they vibrate, but that's beyond the purview of a pickup analysis, and can be inferred from the gauss measurement.

I know exactly what the plot is about, as I have said since the beginning, my interests are in the interaction of the string with the magnetic field, the magnetic circuit and their effects on the actual response of the pickup. In this sense my analogy is as correct as an analogy can be and I think you are intelligent enough to see the parallel.

How can something as fundamental to the operation of a guitar pickup as the interaction between sting and magnetic circuit be out of the purview of pickup analysis? A simple measurement of gauss while the string is at rest tells little, the output of a hall effect sensor plotted on a scope would give some insight, but we could not interpret that data properly without a test such as I have suggested, same as your data, it lacks the needed context to be used to its full potential. We can infer all we want from your data, but why infer when you can test?

[Antigua]

There must be a misunderstanding then as to the intention of the plot. It's sole purpose is to expose the resonant peak frequency, the Q factor, eddy current losses, as well as allow for the intrinsic capacitance to be calculated from the peak f.

But as to the further information you're talking about, a guitar pickup is not like a microphone. A microphone's response curve is dictated by physical considerations, such as what material the diaphragm is made of, or how thick it is, etc. A magnetic guitar pickup doesn't relate to the strings in a physical way, only in a magnetic way. There is a physical aspect in that the magnetic pull on the strings effects the way they vibrate, but that's beyond the purview of a pickup analysis, and can be inferred from the gauss measurement.

I know exactly what the plot is about, as I have said since the beginning, my interests are in the interaction of the string with the magnetic field, the magnetic circuit and their effects on the actual response of the pickup. In this sense my analogy is as correct as an analogy can be and I think you are intelligent enough to see the parallel.

How can something as fundamental to the operation of a guitar pickup as the interaction between sting and magnetic circuit be out of the purview of pickup analysis? A simple measurement of gauss while the string is at rest tells little, the output of a hall effect sensor plotted on a scope would give some insight, but we could not interpret that data properly without a test such as I have suggested, same as your data, it lacks the needed context to be used to its full potential. We can infer all we want from your data, but why infer when you can test?

It would help me greatly if you could explain how you anticipate that the magnetic circuit should impact the tone, because that can mean various things. If you go into more detail, I can think about a suitable test or plot to relate the magnetic circuit and a tonal outcome.

[oid]

It would help me greatly if you could explain how you anticipate that the magnetic circuit should impact the tone, because that can mean various things. If you go into more detail, I can think about a suitable test or plot to relate the magnetic circuit and a tonal outcome.

In the post that started all this I believe I stated my interests, and numorous times after that, to find out what effect the magnetic circuit has on the overall sound. It does not matter what I expect, my expectations have little to do with what is going on between the pickup and the strings.

[Antigua]

It would help me greatly if you could explain how you anticipate that the magnetic circuit should impact the tone, because that can mean various things. If you go into more detail, I can think about a suitable test or plot to relate the magnetic circuit and a tonal outcome.

In the post that started all this I believe I stated my interests, and numorous times after that, to find out what effect the magnetic circuit has on the overall sound. It does not matter what I expect, my expectations have little to do with what is going on between the pickup and the strings.

It's my understanding that the magnetic circuit has fairly little influence on the produced tone, and more to do with the overall output volume. Per Faraday's law, voltage induced by flux change through a loop per a given period of time, and so the magnetic circuit will play a role in determining the amount of flux change that is able to occur, and therefore the overall output voltage. What makes one pickup produce a difference tone from another owes almost entirely to different peak resonant frequencies.

In order for the tone to change, there would have to be a frequency dependence, or a change over time. A change in frequency is seen in the form of eddy current resistance, which increases with the rise of frequency, and that can be seen in the lower Q factor, relative to air core. As for change over time, the magnetic pull upon the guitar string causes the string to vibrate asymmetrically, interfering with the moving string in ways which lead to a difference in harmonic content over time, which constitutes a difference in tone. The reason a pickup with a strong magnet sounds different from a pickup with a weaker magnet is the same reason a pickup sounds different when it's nearer or further from the strings. To a large extent, the intensity of that effect can be inferred by observing the magnetic strength of the pole pieces. The precise effect it has depends on where the pickup is positioned and how far it is from the strings, and a few other qualities of the guitar strings, factors which are not intrinsic to the pickup itself.

[oid]

It's my understanding that the magnetic circuit has fairly little influence on the produced tone, and more to do with the overall output volume. Per Faraday's law, voltage induced by flux change through a loop per a given period of time, and so the magnetic circuit will play a role in determining the amount of flux change that is able to occur, and therefore the overall output voltage. What makes one pickup produce a difference tone from another owes almost entirely to different peak resonant frequencies.

In order for the tone to change, there would have to be a frequency dependence, or a change over time. A change in frequency is seen in the form of eddy current resistance, which increases with the rise of frequency, and that can be seen in the lower Q factor, relative to air core. As for change over time, the magnetic pull upon the guitar string causes the string to vibrate asymmetrically, interfering with the moving string in ways which lead to a difference in harmonic content over time, which constitutes a difference in tone. The reason a pickup with a strong magnet sounds different from a pickup with a weaker magnet is the same reason a pickup sounds different when it's nearer or further from the strings. To a large extent, the intensity of that effect can be inferred by observing the magnetic strength of the pole pieces. The precise effect it has depends on where the pickup is positioned and how far it is from the strings, and a few other qualities of the guitar strings, factors which are not intrinsic to the pickup itself.

It does not matter if the strings are not intrinsic to the pickup itself, they are fundamental to the operation, remove the strings and what is left?

Simple example, what happens when the strings own amplitude carries the string past the edge of the pole? We can bend the string into this region and hear the change, but what happens when this is happening 110 times a second? 220 times a second? 330 times a second? etc.

The Jaguars claw largely removes this effect and acts more like a rail, how does this effect the sound?

Even if such effects are not the dominant source of a pickups sound they are likely important to the sound and quantifying them would give great insight to pickups and their design and increase the usefulness of the data you already gathered.

[Antigua]

It does not matter if the strings are not intrinsic to the pickup itself, they are fundamental to the operation, remove the strings and what is left?

Simple example, what happens when the strings own amplitude carries the string past the edge of the pole? We can bend the string into this region and hear the change, but what happens when this is happening 110 times a second? 220 times a second? 330 times a second? etc.

You get very quiet even order harmonics. This happens with any pole piece, so it's not particular to this pickup so much as it's particular to the pole pieces themselves, which are used for various Fender pickups. More on this here http://www.physics.princeton.edu/~mcdon ... guitar.pdf

The Jaguars claw largely removes this effect and acts more like a rail, how does this effect the sound?

The claw does not act like a rail, it's not magnetized strongly enough to further magnetize the strings above it. Removing the claw shield completely didn't have a negative effect on the output level. Even if it were a magnet itself, and not just a piece of steel, it would still not cancel out the symmetrical component of the pole pieces.

[oid]

You get very quiet even order harmonics

And that is one effect caused by the magnetic circuit.

The claw does not act like a rail

I said act "MORE" like a rail, the effects of the claw on the shape of the magnetic field can be easily demonstrated.

In your write ups you admit to discrepancies that you can not explain, but you push against devising any further tests and continue on as though you have everything figured out and that the minutiae are unimportant, I just do not get this.

[Antigua]

You get very quiet even order harmonics

And that is one effect caused by the magnetic circuit.

It's not characteristic of the pickup, though. The degree to which this happens have be inferred by just looking at the width of the pole piece. It's also a very quiet effect, so it's not especially worth determining.

The claw does not act like a rail

I said act "MORE" like a rail, the effects of the claw on the shape of the magnetic field can be easily demonstrated.

I read you correctly, but based on the low amount of flux involvement between the shield, the pole piece and the strings, there will be no audible similarity to a rail. The flux at the top of the pole piece is around 1000 gauss, the flux at the top of the claw will be a tiny fraction of that. It too would have to be close to 1000G in order to compete with the pole pieces for overall effect.

In your write ups you admit to discrepancies that you can not explain, but you push against devising any further tests and continue on as though you have everything figured out and that the minutiae are unimportant, I just do not get this.

Likewise, you're trying to infer that the vagarity proves your claim that the magnetic circuit is of relevance to the heard product of the pickup, which is equally unsubstantiated.

If you'd like to undertake this endeavor, be my guest. It's a non-trivial amount of work, and nobody is paying me.

[oid]

It's not characteristic of the pickup, though. The degree to which this happens have be inferred by just looking at the width of the pole piece. It's also a very quiet effect, so it's not especially worth determining.

It does not take much change in harmonic content to affect overall tonality especially in the early harmonics such as the second and fourth, this would not be show in your plots since it would be entirely within the pass band of the filter formed by the coil and its load.

I read you correctly, but based on the low amount of flux involvement between the shield, the pole piece and the strings, there will be no audible similarity to a rail. The flux at the top of the pole piece is around 1000 gauss, the flux at the top of the claw will be a tiny fraction of that.

Try the old iron fillings on paper test, it shows things your meter can not tell you. If you did read me correctly you still chose ignore what I asked, I did not suggest it would sound anything like a rail, or suggest what the impact on sound would be, I asked a question regarding the effect of the change in shape and its effect on sound and it was not rhetorical.

Likewise, you're trying to infer that the vagarity proves your claim that the magnetic circuit is of relevance to the heard product of the pickup, which is equally unsubstantiated.

I am not trying to prove anything, I have admitted to not knowing the answers, and have even said that I may be completely wrong, I am just looking for answers

If you'd like to undertake this endeavor, be my guest. It's a non-trivial amount of work, and nobody is paying me.

The only thing I ever asked of you Is one measurement and advice on devising tests, never once even implied that I am trying to get you to undertake this for me.

[Antigua]

You know my view is that the magnetic circuit is principally involved in determining the overall amplitude of the pickup, where as the coil is mostly determinant in the RLC low pass filtering, and all other harmonics are dictated by the placement of the pickup, and any comb filtering that might occur with more than one coil (not the case here). The McDonald paper I linked to talks about the even harmonics and their low amplitude. If you only consider this analysis to be an RLC analysis, that's fine with me.

I have done a lot of practical testing with actual guitar strings as well, and the issue there is that a lot of changes will result in differences that are either too small to register visually, or are smaller than some other dependent variable, for example, the variation in picking the guitar string might caused a harmonic variance that easily exceeds the harmonic variance of magnetic string pull, which is telling in its own right, but also makes it difficult to say conclusive things about the effects of magnetic string pull. For that reason, it's easier to approach some of these things as though models, or mathematical models, as McDonald did.