frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 13:33:25 GMT -5
Yeah you're still thinking of comb filtering as it relates to the string window off the Tillman article. Yes, that's a thing too but I'm talking about as it relates to the coil's proximity to the strings.
|
|
|
Post by antigua on Mar 20, 2017 13:48:13 GMT -5
If you want to get into the HiFi voodoo jargon that you probably won't like, one phenomenon is referred to as group delay. Highs/mids/lows out of time alignment with one another, in a smear. Proprietary testing methods notwithstanding, it's easier for me to show the effects manifested than it is to show it on screen. This is where on other forums you've heard me talk about all the notes in a complex chord to be able to survive heavy gain; that it is not as much frequency dependent as it is time alignment. But really it's quite simple to see a part of it on a macro level by recording both coils from an equally wound stack, and playing around with the amplitude, looking for time alignment anomalies...And it's not necessary for you to concede to any time-based causality in a wire-wound coil, traversing closer to, and further from the string, even though I'm saying there is something there. You can just attribute it to out of phase flux disturbance of you like. The result is the same. You can even start cutting it up into different frequency bands with band pass filters and looking at sections against one another. This is unproven assertion, very unlikely to be true as, if taken at face value, this would impact transformer design. Every time you ran audio through a transformer, you'd have frequency dependent phase variance - that doesn't happen. This is exactly the kind of thing you could demonstrate for the world to see, as you've already partly described how you'd go about conducting the test. Why is the onus on me or Ken Willmott to test everything? When are you going to get your hands dirty for the greater good?
|
|
|
Post by antigua on Mar 20, 2017 13:49:36 GMT -5
Yeah you're still thinking of comb filtering as it relates to the string window off the Tillman article. Yes, that's a thing too but I'm talking about as it relates to the coil's proximity to the strings. Magnetic fields propagate at the speed of light. The reason you have comb filtering with microphones is because sound does not move at the speed of light.
|
|
|
Post by stratotarts on Mar 20, 2017 14:15:46 GMT -5
Furthermore, to have comb filtering you need signal inversion - all the induced voltages in a coil are in-phase.
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 14:21:31 GMT -5
I told you you wouldn't like it.
|
|
|
Post by stratotarts on Mar 20, 2017 14:28:12 GMT -5
I told you you wouldn't like it. There is nothing to like or dislike. You haven't really presented anything that has any meaning. I wasn't going to say it, but since you persist - it is very obvious that your knowledge of physics and electronics is rather limited. That in itself is nothing, but you're pretending otherwise.
|
|
|
Post by stratotarts on Mar 20, 2017 15:02:13 GMT -5
If you want to get into the HiFi voodoo jargon that you probably won't like, one phenomenon is referred to as group delay. No, group delay is a common engineering term. Wiki link
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 15:08:42 GMT -5
What am I pretending exactly? Do I sign PhD after my name? All I try to do is tell you guys if/when you're missing something, and the ONLY time I do is when errant conclusions are reached and then propagated as fact around the water cooler so to speak.
|
|
|
Post by ms on Mar 20, 2017 15:23:21 GMT -5
Furthermore, to have comb filtering you need signal inversion - all the induced voltages in a coil are in-phase. Not quite. If you make the coil wide enough, some lines which have gone down through the coil return through at least a few windings, and on the return trip they point in the opposite direction, contributing a small amount of signal of opposite polarity.. This causes the output to increase a bit less rapidly* than it would if those flux lines stayed outside of all windings when you wind on more wire. However, nothing cancels, there is no comb filter. * by which i really mean that it increases less per added turn than if this did not happen. (Thanks, Antigua) Frank: Now, as for group delay., the electrical circuit of the pickup, the resonance, etc., causes a frequency dependent group delay. So you can measure this, but the result has nothing to do with magnetic return paths, just with the electrical filtering action of the circuit. Frank writes: "I was asked how I seem to have some deeper information about what's happening within the coil and I replied that part of it is from Fluence R&D, partly through stack R&D..." I think most people realize by now that you do not have any deeper information. Piling up expensive test equipment and calling what'd happens R&D is a gross exaggeration. If you do not understand even the basics of the physics of pickups and the electronics of circuits, you are just providing advertising copy, not any actual knowledge.
|
|
|
Post by ms on Mar 20, 2017 15:31:22 GMT -5
What am I pretending exactly? Do I sign PhD after my name? All I try to do is tell you guys if/when you're missing something, and the ONLY time I do is when errant conclusions are reached and then propagated as fact around the water cooler so to speak. What? The topic of this discussion was (and still should be) the sampling window of the string. You started by commenting about string bending as if that were the topic, rather than part of a measuring technique. How can you tell people what they are missing when you cannot even understand what they are trying to do? You do not have the understanding you need to play the role you are claiming. Not even close.
|
|
|
Post by antigua on Mar 20, 2017 15:31:32 GMT -5
Furthermore, to have comb filtering you need signal inversion - all the induced voltages in a coil are in-phase. Not quite. If you make the coil wide enough, some lines which have gone down through the coil return through at least a few windings, and on the return trip they point in the opposite direction, contributing a small amount of signal of opposite polarity.. This causes the output to increase a bit less rapidly than it would if those flux lines stayed outside of all windings when you wind on more wire. However, nothing cancels, there is no comb filter. You say "less rapidly", so you do think there is some sort of time consequence involved? My understanding, based on Faraday's law, says the consequence is very simple, flux change induces voltage. If you input flux, and simultaneously input opposite flux, you simply reduce the overall flux. You just reduce the induced voltage, plain and simple. The return path is of a much lower flux density than the primary path, so this effect is very small, no matter what. You could call it a small cancellation. As far as I can tell, there are no time based consequences, no delays. Furthermore, to have comb filtering you need signal inversion - all the induced voltages in a coil are in-phase. It doesn't have to be perfectly inverted, perfect 180 degree inversion just ensures maximum comb filtering. Even a slight out of phase state causes comb filtering, the key ingredient is really that the two overlapping signals be the same (or very nearly the same) signal.
|
|
|
Post by ms on Mar 20, 2017 15:34:37 GMT -5
Not quite. If you make the coil wide enough, some lines which have gone down through the coil return through at least a few windings, and on the return trip they point in the opposite direction, contributing a small amount of signal of opposite polarity.. This causes the output to increase a bit less rapidly than it would if those flux lines stayed outside of all windings when you wind on more wire. However, nothing cancels, there is no comb filter. You say "less rapidly", so you do think there is some sort of time consequence involved? My understanding, based on Faraday's law, says the consequence is very simple, flux change induces voltage. If you input flux, and simultaneously input opposite flux, you simply reduce the overall flux. You just reduce the induced voltage, plain and simple. The return path is of a much lower flux density than the primary path, so this effect is very small, no matter what. You could call it a small cancellation. As far as I can tell, there are no time based consequences, no delays. No sorry, I do not mean time, I mean that the signal strength increases by a small amount per turn wound on than if returning fluxed not intersect some of the windings. Thanks for catching that.
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 15:35:27 GMT -5
How is anything that I have said in disagreement with the two things you just posted? Group delay and return flux are both responsible for anomalies. Where did I say something outside of that?
|
|
|
Post by antigua on Mar 20, 2017 15:40:01 GMT -5
You say "less rapidly", so you do think there is some sort of time consequence involved? My understanding, based on Faraday's law, says the consequence is very simple, flux change induces voltage. If you input flux, and simultaneously input opposite flux, you simply reduce the overall flux. You just reduce the induced voltage, plain and simple. The return path is of a much lower flux density than the primary path, so this effect is very small, no matter what. You could call it a small cancellation. As far as I can tell, there are no time based consequences, no delays. No sorry, I do not mean time, I mean that the signal strength increases by a small amount per turn wound on than if returning fluxed not intersect some of the windings. Thanks for catching that. I know what you mean, I've thought in those terms to. Since putting wire on a bobbin takes time, I think of more winds = more time.
|
|
|
Post by ms on Mar 20, 2017 15:44:39 GMT -5
How is anything that I have said in disagreement with the two things you just posted? Group delay and return flux are both responsible for anomalies. Where did I say something outside of that? I do not think return flux is responsible for anomalies. Its effects are apparent from the law of magnetic induction. I guess I misunderstood what you were trying to do by brining up groups delay. That reply was directed to me? I am not sure since you do not quote, or otherwise indicate who you are addressing.
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 15:45:28 GMT -5
Technically 180° wouldn't be referred to as comb filtering because it's full range, right?
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 15:48:27 GMT -5
How is anything that I have said in disagreement with the two things you just posted? Group delay and return flux are both responsible for anomalies. Where did I say something outside of that? I do not think return flux is responsible for anomalies. Its effects are apparent from the law of magnetic induction. I guess I misunderstood what you were trying to do by brining up groups delay. That reply was directed to me? I am not sure since you do not quote, or otherwise indicate who you are addressing. Yeah I guess I'll use quote more. It's cumbersome on mobile and I didn't realize so many people were jumping in when I replied. It's a real party line right now.
|
|
|
Post by antigua on Mar 20, 2017 15:53:36 GMT -5
AFAIK, in order to get group delay, you need some sort of "grouping" agent. For example, you have an analogue multiband EQ, some frequency bands pass through some capacitors, some bands pass through others, and then all arrive at the output with various phases as a result of RLC filtering, hence "groups" of "delay". You can also have group delay with speakers, again due to the physical distinction of moving energy through air. With a speaker, low frequencies are more movement-restricted than high frequencies.
How would/could this concept manifest in a guitar pickup?
I apologize for straying from the topic of magnetic aperture, but I think there is value in novel suppositions about pickups that have never before been evaluated in a physics-focused space.
|
|
|
Post by ms on Mar 20, 2017 16:04:21 GMT -5
AFAIK, in order to get group delay, you need some sort of "grouping" agent. For example, you have an analogue multiband EQ, some frequency bands pass through some capacitors, some bands pass through others, and then all arrive at the output with various phases as a result of RLC filtering, hence "groups" of "delay". You can also have group delay with speakers, again due to the physicality of moving air with a speaker, low frequencies are more movement-restricted than high frequencies. How would/could this happen in a guitar pickup? I apologize for straying from the topic of magnetic aperture, but I think there is value in novel suppositions about pickups that have never before been evaluated in a physics-focused space. Antigua, group delay is a physics and engineering term applying to things like wave probation and signal propagation through filters. Sratotarts gave a link earlier.
|
|
|
Post by ms on Mar 20, 2017 16:05:21 GMT -5
I do not think return flux is responsible for anomalies. Its effects are apparent from the law of magnetic induction. I guess I misunderstood what you were trying to do by brining up groups delay. That reply was directed to me? I am not sure since you do not quote, or otherwise indicate who you are addressing. Yeah I guess I'll use quote more. It's cumbersome on mobile and I didn't realize so many people were jumping in when I replied. It's a real party line right now. Yes, mobiles are difficult for this. I do not think I could do it at all.
|
|
|
Post by antigua on Mar 20, 2017 16:07:49 GMT -5
AFAIK, in order to get group delay, you need some sort of "grouping" agent. For example, you have an analogue multiband EQ, some frequency bands pass through some capacitors, some bands pass through others, and then all arrive at the output with various phases as a result of RLC filtering, hence "groups" of "delay". You can also have group delay with speakers, again due to the physicality of moving air with a speaker, low frequencies are more movement-restricted than high frequencies. How would/could this happen in a guitar pickup? I apologize for straying from the topic of magnetic aperture, but I think there is value in novel suppositions about pickups that have never before been evaluated in a physics-focused space. Antigua, group delay is a physics and engineering term applying to things like wave probation and signal propagation through filters. Sratotarts gave a link earlier. From this link: They say "Many components of an audio reproduction chain...", so I went looking for causes, and didn't see anything that could be correlated with a guitar pickup, nothing concerning magnetic coupling. There doesn't appear to be anything here.
|
|
|
Post by stratotarts on Mar 20, 2017 16:08:01 GMT -5
Would not signals induced by the alternating field in the return path, be in some fixed phase relationship to the signals induced by the alternating field in the main path? If that is the case, only the overall amplitude of the signal will be affected by any portions of the coil that intersect the return path. Since, the string is always in a fixed position relative to any part of the coil winding. If there is no phase shift, there is no group delay.
|
|
|
Post by antigua on Mar 20, 2017 16:13:01 GMT -5
Would not signals induced by the alternating field in the return path, be in some fixed phase relationship to the signals induced by the alternating field in the main path? If that is the case, only the overall amplitude of the signal will be affected by any portions of the coil that intersect the return path. Yes. Magnetic fields propagate at light speed. There is no mechanism for delay. If a pickup's winds where a bunch of microphones, then you'd get "smearing" by virtue of sound waves arriving at the first wind before it arrives at the last wind, but magnetic waves are not sound. I believe this mental picture is probably what gave rise to the notion.
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Mar 20, 2017 16:26:34 GMT -5
AFAIK, in order to get group delay, you need some sort of "grouping" agent. For example, you have an analogue multiband EQ, some frequency bands pass through some capacitors, some bands pass through others, and then all arrive at the output with various phases as a result of RLC filtering, hence "groups" of "delay". You can also have group delay with speakers, again due to the physical distinction of moving energy through air. With a speaker, low frequencies are more movement-restricted than high frequencies. How would/could this concept manifest in a guitar pickup? I apologize for straying from the topic of magnetic aperture, but I think there is value in novel suppositions about pickups that have never before been evaluated in a physics-focused space. It is true that group delay does not relate to the string window discussion. We got there because I said neglecting neighboring poles impacts the string window test you were conducting, and suggested an electric piano style tine perpendicular to the would-be string axis. Perhaps where only the tip is magnetic. From there we get to phase anomalies and eventually ms is talking about how if a coil is wide enough, the return path is a variable. What I'm saying is that with one lone pole in the center of the pickup, you do have coil "width" in the return path because your return path is cyllindrical. Neighboring poles would be adding focus, and pushing the return path out and around. Now, unlike ms I don't feel the coil has to be unconventionally wide to be affected. There is return flux practically on the skin, admittedly minuscule but as you've seen, I don't generally dismiss minutiae. So I'm happy to agree to disagree about it's meaningfulness in a fully loaded coil.
|
|
|
Post by ms on Mar 20, 2017 16:46:15 GMT -5
Antigua, group delay is a physics and engineering term applying to things like wave probation and signal propagation through filters. Sratotarts gave a link earlier. From this link: They say "Many components of an audio reproduction chain...", so I went looking for causes, and didn't see anything that could be correlated with a guitar pickup, nothing concerning magnetic coupling. There doesn't appear to be anything here. A filter causes group delay. The electric circuit of a pickup is a resonant low pass filter. It has a group delay that varies with frequency.
|
|
|
Post by antigua on Mar 20, 2017 16:51:44 GMT -5
From this link: They say "Many components of an audio reproduction chain...", so I went looking for causes, and didn't see anything that could be correlated with a guitar pickup, nothing concerning magnetic coupling. There doesn't appear to be anything here. A filter causes group delay. The electric circuit of a pickup is a resonant low pass filter. It has a group delay that varies with frequency. That's a good point. The Velleman support phase plotting, but not as easily as the the CGR-101. Are you referring to the phase change that occurs as you approach and cross over LC resonance? If so, I think that's all there is to see here, as this is an LC function than can't be tailored apart from the resonance itself. One would always follow with the other. Never mind the fact that the phase curve is probably inaudible, otherwise.
|
|
|
Post by antigua on Mar 20, 2017 16:57:18 GMT -5
AFAIK, in order to get group delay, you need some sort of "grouping" agent. For example, you have an analogue multiband EQ, some frequency bands pass through some capacitors, some bands pass through others, and then all arrive at the output with various phases as a result of RLC filtering, hence "groups" of "delay". You can also have group delay with speakers, again due to the physical distinction of moving energy through air. With a speaker, low frequencies are more movement-restricted than high frequencies. How would/could this concept manifest in a guitar pickup? I apologize for straying from the topic of magnetic aperture, but I think there is value in novel suppositions about pickups that have never before been evaluated in a physics-focused space. It is true that group delay does not relate to the string window discussion. We got there because I said neglecting neighboring poles impacts the string window test you were conducting, and suggested an electric piano style tine perpendicular to the would-be string axis. Perhaps where only the tip is magnetic. From there we get to phase anomalies and eventually ms is talking about how if a coil is wide enough, the return path is a variable. What I'm saying is that with one lone pole in the center of the pickup, you do have coil "width" in the return path because your return path is cyllindrical. Neighboring poles would be adding focus, and pushing the return path out and around. Now, unlike ms I don't feel the coil has to be unconventionally wide to be affected. There is return flux practically on the skin, admittedly minuscule but as you've seen, I don't generally dismiss minutiae. So I'm happy to agree to disagree about it's meaningfulness in a fully loaded coil. Without being certain what you mean, I'm fairly certain that this point is addressed in the above posts: the flux change is instantaneous. No arrangement of pole piece or coil would "slow down" the rate at which magnetic flux traverses the coil in any given location. As for the "Tillman" sense of comb filtering, by adding slugs to the left or the right of the slug of interest, you're not widening or narrowing the aperture width, so you're not changing anything other than overall output amplitude (and inductance in the coil).
|
|
|
Post by ms on Mar 20, 2017 17:34:40 GMT -5
It is true that group delay does not relate to the string window discussion. We got there because I said neglecting neighboring poles impacts the string window test you were conducting, and suggested an electric piano style tine perpendicular to the would-be string axis. Perhaps where only the tip is magnetic. From there we get to phase anomalies and eventually ms is talking about how if a coil is wide enough, the return path is a variable. What I'm saying is that with one lone pole in the center of the pickup, you do have coil "width" in the return path because your return path is cyllindrical. Neighboring poles would be adding focus, and pushing the return path out and around. Now, unlike ms I don't feel the coil has to be unconventionally wide to be affected. There is return flux practically on the skin, admittedly minuscule but as you've seen, I don't generally dismiss minutiae. So I'm happy to agree to disagree about it's meaningfulness in a fully loaded coil. Without being certain what you mean, I'm fairly certain that this point is addressed in the above posts: the flux change is instantaneous. No arrangement of pole piece or coil would "slow down" the rate at which magnetic flux traverses the coil in any given location. As for the "Tillman" sense of comb filtering, by adding slugs to the left or the right of the slug of interest, you're not widening or narrowing the aperture width, so you're not changing anything other than overall output amplitude (and inductance in the coil). This is not quite true. Ferromagnetic material can slow down propagation below the speed of light in a vacuum. But it does not slow it down enough to make a significant effect in the audio range. This is roughly analogous to the speed of propagation of rf in a coax cable. In that case the dielectric material (the plastic between the inner and outer conductors) causes the the propagation to slow below the speed of light in a vacuum. Speeds could be .82 tines the speed of light or perhaps .95 depending upon the construction of the coax line.
|
|
|
Post by antigua on Mar 20, 2017 17:37:57 GMT -5
Without being certain what you mean, I'm fairly certain that this point is addressed in the above posts: the flux change is instantaneous. No arrangement of pole piece or coil would "slow down" the rate at which magnetic flux traverses the coil in any given location. As for the "Tillman" sense of comb filtering, by adding slugs to the left or the right of the slug of interest, you're not widening or narrowing the aperture width, so you're not changing anything other than overall output amplitude (and inductance in the coil). This is not quite true. Ferromagnetic material can slow down propagation below the speed of light in a vacuum. But it does not slow it down enough to make a significant effect in the audio range. This is roughly analogous to the speed of propagation of rf in a coax cable. In that case the dielectric material (the plastic between the inner and outer conductors) causes the the propagation to slow below the speed of light in a vacuum. Speeds could be .82 tines the speed of light or perhaps .95 depending upon the construction of the coax line. That's a good point, at least with the magnetic permeability it takes times for magnetic domains to orient, and I suppose that presents a high frequency hysteresis loss, but yeah, not applicable at or below 20kHz.
|
|
|
Post by ms on Mar 21, 2017 10:17:22 GMT -5
In the first post, results for the variation of string magnetization along the string were shown. The results were obtained with an air core pickup coil, and thus they represent a spatial average over the sensitive range of the coil. The result were:
"Three measurements are made by carefully plucking the string, one centered over the guitar pickup, and then slid along the strings by 3/32 and 3/16 inches. The measurements are converted to db referred to the centered measurement, and the measurements are 0 db, -3.1 db, and -18.4 db. "
This post presents some results for the other half of the pickup problem: excite the the coil from different locations along where the string would be. This is done with a very small coil with a diameter about half of a pickup pole piece with three turns. The coil is driven with about 1 ampere at one KHz and the output the guitar is measured using an fft analyzer for each position. (The signal level excited by the small coil is not very big, and so it is a good idea to use a small frequency range and integrate for a few seconds.)
The results for the same displacements from the center are 0 db, -1.8 db, and -4.7 db. Moving to the edge of the pickup gives -15.1 db. This is outside the coil.
This result is quite a bit wider than the result for string magnetization. This is not surprising; although we expect the strongest concentration of flux through the pole piece, short pole pieces confine the flux imperfectly, and so the amplification of the incident field is not that high. This means that the pickup of flux coming into the coil but missing the pole is significant.
So the overall result for both measurements added is 0 db, -4.9 db, and -27.1 db for centered, 2/32", and 3/16" from the center. The conclusion that the pole piece sets the sampling distance for this pickup is reinforced a bit although these are approximate measurements, and might vary some when better measurements are made.
The next thing to measure is a humbucker.
|
|