|
|
Post by antigua on Oct 8, 2016 2:31:27 GMT -5
www.seymourduncan.com/pickup/duncan-custom This pickup was introduced in 1979, so it's among the protohistorical aftermarket pickups. Seymour Duncan calls this a "hot PAF", though it's wound with 43 AWG wire , which has become a characteristic of "hot" humbuckers that tend not to be compared to PAFs these days. It also features a ceramic magnet, which is considered very un-PAF these days.
Seymour Duncan SH-5 Custom
Series resistance: 14.45k ohms
screw (red & green): 7.14k
slug (white & black): 7.31k
Parallel: 3.61k
Series inductance: 7.734H
screw (red & green): 3.387H
slug (white & black): 3.432H
Parallel: 1.933H
Gauss: 380G slug, 320G screw , ceramic
Unloaded: V: 5.2dB f: 5.19kHz (black)
screw (red & green): V: 4.8dB f: 8.29kHz (red)
slug (white & black): V: 6.0dB f: 7.59kHz (green)
parallel: V: 5.0dB f: 8.02kHz (gray)
Loaded (200k & 470pF): V: 1.0dB f: 2.03kHz (black)
screw (red & green): V: 1.4dB f: 3.43kHz (red)
slug (white & black): V: 2.3dB f: 3.24kHz (green)
parallel: V: 2.5dB f: 4.31kHz (gray)Bode plots:
Unloaded:  Loaded w/ 200k & 470pF:  These plots show that the screw coil sees higher eddy losses than the slug coil, as usually seems to be the case, though the 1dB difference between them is not audibly consequential. I can also be seen that parallel mode has a resonant peak that is almost 1kHz higher than either coil in split mode, and that split mode is well over 1kHz higher than series mode. The split and parallel modes put their peaks in the vicinity of a Stratocaster pickup, though with a much lower Q. IMO, this goes to show that splitting and parallel wiring tends to be more effective with these high output pickups. The ceramic magnet should give the split and parallel modes a little more kick, too. 
|
|
|
|
Post by wgen on Jun 28, 2017 4:40:16 GMT -5
www.seymourduncan.com/pickup/duncan-customThis pickup was introduced in 1979, so it's among the protohistorical aftermarket pickups. Seymour Duncan calls this a "hot PAF", though it's wound with 43 AWG wire , which has become a characteristic of "hot" humbuckers that tend not to be compared to PAFs these days. It also features a ceramic magnet, which is considered very un-PAF these days.
Seymour Duncan SH-5 Custom
Series resistance: 14.45k ohms
screw (red & green): 7.14k
slug (white & black): 7.31k
Parallel: 3.61k
Series inductance: 7.734H
screw (red & green): 3.387H
slug (white & black): 3.432H
Parallel: 1.933H
Gauss: 380G slug, 320G screw , ceramic
Unloaded: V: 5.2dB f: 5.19kHz (black)
screw (red & green): V: 4.8dB f: 8.29kHz (red)
slug (white & black): V: 6.0dB f: 7.59kHz (green)
parallel: V: 5.0dB f: 8.02kHz (gray)
Loaded (200k & 470pF): V: 1.0dB f: 2.03kHz (black)
screw (red & green): V: 1.4dB f: 3.43kHz (red)
slug (white & black): V: 2.3dB f: 3.24kHz (green)
parallel: V: 2.5dB f: 4.31kHz (gray)Bode plots:
Unloaded: Loaded w/ 200k & 470pF: These plots show that the screw coil sees higher eddy losses than the slug coil, as usually seems to be the case, though the 1dB difference between them is not audibly consequential. I can also be seen that parallel mode has a resonant peak that is almost 1kHz higher than either coil in split mode, and that split mode is well over 1kHz higher than series mode. The split and parallel modes put their peaks in the vicinity of a Stratocaster pickup, though with a much lower Q. IMO, this goes to show that splitting and parallel wiring tends to be more effective with these high output pickups. The ceramic magnet should give the split and parallel modes a little more kick, too. I was reading this analysis again, and I was thinking about the parallel wiring of a high output humbucker pickup like this One.. I have a doubt regarding the resonant peak of the parallel connection..I mean: why does it go up of just a couple of dBs in respect to the series connection? Leave alone the shift in Hz....the pickup when in parallel shows less than 2H inductance and less than 4 K of DC resistance. Shouldn't the peak be much higher with that low resistance? I'm specifically talking about the loaded graph What I'm omitting here...? Thank you very much in advance! |
|
|
|
Post by antigua on Jun 28, 2017 13:29:07 GMT -5
I was reading this analysis again, and I was thinking about the parallel wiring of a high output humbucker pickup like this One.. I have a doubt regarding the resonant peak of the parallel connection..I mean: why does it go up of just a couple of dBs in respect to the series connection? Leave alone the shift in Hz....the pickup when in parallel shows less than 2H inductance and less than 4 K of DC resistance. Shouldn't the peak be much higher with that low resistance? I'm specifically talking about the loaded graph What I'm omitting here...? Thank you very much in advance! The Q factor doesn't have much to do with series resistance, it's almost entirely determined by eddy current losses. Something else that I think might be happening, which I can test later, is in parallel mode I think the coils mutually act as eddy current causing agents, because you have continuity at both ends of both coils, so current can pass through both coils as a closed circuit. When current can flow, it's gives the magnetic field passing through the conductive loops a means by which it can converted into a current and a heat loss, hence eddy currents. I can test this later by seeing what happens when the pickups are wired in parallel, and the coils are side by side, or made to be farther apart from one another. I have tested this in the context of splitting, but not parallel wiring. |
|
|
|
Post by wgen on Jun 28, 2017 17:49:59 GMT -5
I was reading this analysis again, and I was thinking about the parallel wiring of a high output humbucker pickup like this One.. I have a doubt regarding the resonant peak of the parallel connection..I mean: why does it go up of just a couple of dBs in respect to the series connection? Leave alone the shift in Hz....the pickup when in parallel shows less than 2H inductance and less than 4 K of DC resistance. Shouldn't the peak be much higher with that low resistance? I'm specifically talking about the loaded graph What I'm omitting here...? Thank you very much in advance! The Q factor doesn't have much to do with series resistance, it's almost entirely determined by eddy current losses. Something else that I think might be happening, which I can test later, is in parallel mode I think the coils mutually act as eddy current causing agents, because you have continuity at both ends of both coils, so current can pass through both coils as a closed circuit. When current can flow, it's gives the magnetic field passing through the conductive loops a means by which it can converted into a current and a heat loss, hence eddy currents. I can test this later by seeing what happens when the pickups are wired in parallel, and the coils are side by side, or made to be farther apart from one another. I have tested this in the context of splitting, but not parallel wiring. Thank you, I suspected that it could be something related to Eddy current losses. I downloaded a simple pickup simulator program some time ago, I think it was the Lite version of the One from the user JohnH here of the Guitarnuts forum. When I insert your exact values of this Custom humbucker it always give me something around 3 dB higher Q than your testing, especially when in parallel. I didn't consider that Eddy current losses could damp the response some. What would be the reasons of these losses, considering the uncovered design? Maybe the steel slugs and screws, in respect to alnico poles? Thank you again! |
|
|
|
Post by antigua on Jun 28, 2017 20:07:26 GMT -5
The Q factor doesn't have much to do with series resistance, it's almost entirely determined by eddy current losses. Something else that I think might be happening, which I can test later, is in parallel mode I think the coils mutually act as eddy current causing agents, because you have continuity at both ends of both coils, so current can pass through both coils as a closed circuit. When current can flow, it's gives the magnetic field passing through the conductive loops a means by which it can converted into a current and a heat loss, hence eddy currents. I can test this later by seeing what happens when the pickups are wired in parallel, and the coils are side by side, or made to be farther apart from one another. I have tested this in the context of splitting, but not parallel wiring. Thank you, I suspected that it could be something related to Eddy current losses. I downloaded a simple pickup simulator program some time ago, I think it was the Lite version of the One from the user JohnH here of the Guitarnuts forum. When I insert your exact values of this Custom humbucker it always give me something around 3 dB higher Q than your testing, especially when in parallel. I didn't consider that Eddy current losses could damp the response some. What would be the reasons of these losses, considering the uncovered design? Maybe the steel slugs and screws, in respect to alnico poles? Thank you again! An adjacent coil can actually cause eddy current losses also, because, when wired in parallel, it meets the physical criteria necessary to turn a changing magnetic field into an electrical current. The coil is connected at both ends, so it's a closed circuit, and it can pass a current around itself. If you split a humbucker, this also happens, because the shorted out coil has continuity. Some believe that to split a humbucker and get a true single coil tone you want to also disconnect the ground side of that second coil, but there are other things that prevent a split humbucker from sounding like an ideal single coil. In this respect, you can also say the two coils are behaving like a transformer, and it's causing a loss in LC resonance efficiency in the primary coil. I'll test this out to make sure this is what is accounting for the low Q, though. |
|
|
|
Post by wgen on Jun 29, 2017 2:00:19 GMT -5
Thank you. I read your posts about capacitance in coil tap modes in the past, but these informations about eddy currents with parallel and split humbuckers are very interesting.
Apart from those related to coil geometry, do you think that, given two pickups both with the same inductance and the same resistance, but One with alnico poles and the other with steel poles (like the example of two coils wired in parallel in respect to a real single coil perhaps)...would this second one have more losses than the first when loaded?
Now I guess the question isn't well put because in the Real world I don't know how that could happen (ie, with parallel coils you would have those losses you talked about in your previous post so this aspect would eventually be masked from that, or, if you have steel above magnet bar this pickup would have either different inductance or different resistance if you underwound it for the same inductance, and so on..)
So let's just say: in a loaded guitar circuit, would steel poles lead to some dB more of losses in respect to alnicos?
I think to remember that in your Filtertrons analysis, all things being on par, the bigger/longer steel screws lead to more losses than shorter ones...so I was now thinking if the same happens changing the pole pieces materials, but I'd be interested in loaded situation because if when loaded it's just a minor difference of a fraction of a dB I guess that wouldn't be audible.
Thank you!
|
|
|
|
Post by antigua on Jun 29, 2017 19:24:39 GMT -5
Here is a parallel plot with the second coil in the proper position, and the second coil removed from the base plate and moved to the side:  I normalized the amplitude so that the plots would overlap. Once the second coil was moved aside, it received far less signal from the driver coil, causing an overall drop in amplitude. It can be seen that the resonance increased by about 1dB, which is fairly trivial, but enough to show an "eddy current", or inductive coupling relationship between the two coils that retards the LC resonance slightly. Therefore it seems that most of the eddy current losses are steel core related, and not owing to the neighboring coil. The peak frequency did go up by 800Hz, I wasn't expecting that. That owes to a drop in mutual capacitance and inductance. Using an LCR meter set to 120Hz, series resistance mode, I measured 1.27 henries with the pickup configured normally, and 1.02 henries with the coils apart, so it looks like the 250mH change in inductance it fairly significant. I'd guess its the other coil's core being nearby that increases the inductance.  |
|
|
|
Post by wgen on Jun 30, 2017 7:30:05 GMT -5
Here is a parallel plot with the second coil in the proper position, and the second coil removed from the base plate and moved to the side: I normalized the amplitude so that the plots would overlap. Once the second coil was moved aside, it received far less signal from the driver coil, causing an overall drop in amplitude. It can be seen that the resonance increased by about 1dB, which is fairly trivial, but enough to show an "eddy current", or inductive coupling relationship between the two coils that retards the LC resonance slightly. Therefore it seems that most of the eddy current losses are steel core related, and not owing to the neighboring coil. The peak frequency did go up by 800Hz, I wasn't expecting that. That owes to a drop in mutual capacitance and inductance. Using an LCR meter set to 120Hz, series resistance mode, I measured 1.27 henries with the pickup configured normally, and 1.02 henries with the coils apart, so it looks like the 250mH change in inductance it fairly significant. I'd guess its the other coil's core being nearby that increases the inductance. Thank you for the test! That's interesting, and it shows also that part of the difference from a demo in respect to a proper test may be because of these eddy currents of the close coils in parallel. In fact, when I insert the same inductance and the same resistance in the demo software I get higer Q than your test. But I understand that the demo doesn't consider these eddy currents, it uses the values as if they are from an alnico single coil with those specs, I guess, instead of this "real world" parallel coils with slugs and screws. Another question I thought about this Duncan Custom humbucker is around the wire gauge. I read it's 43 AWG wire, for a resistance of 14+ K. I was wondering how it does compare to a PAF with 42 AWG and around 8 K resistance. I mean: leave alone the Q of the peak, that if I understood correctly is lower because of the high resistance with this Custom humbucker. But..why the high inductance? Is it correct if I assume that not only the Custom pickup has 43 AWG, but it is also overwound if compared to a PAF with 42 AWG? I hope this question is clear enough. How much DC resistance should a 43 AWG wire humbucker have to achieve the same inductance as a PAF with 42 AWG? Thank you very much again! |
|
|
|
Post by antigua on Jun 30, 2017 18:21:24 GMT -5
Apart from those related to coil geometry, do you think that, given two pickups both with the same inductance and the same resistance, but One with alnico poles and the other with steel poles (like the example of two coils wired in parallel in respect to a real single coil perhaps)...would this second one have more losses than the first when loaded? Now I guess the question isn't well put because in the Real world I don't know how that could happen (ie, with parallel coils you would have those losses you talked about in your previous post so this aspect would eventually be masked from that, or, if you have steel above magnet bar this pickup would have either different inductance or different resistance if you underwound it for the same inductance, and so on..) So let's just say: in a loaded guitar circuit, would steel poles lead to some dB more of losses in respect to alnicos? I think to remember that in your Filtertrons analysis, all things being on par, the bigger/longer steel screws lead to more losses than shorter ones...so I was now thinking if the same happens changing the pole pieces materials, but I'd be interested in loaded situation because if when loaded it's just a minor difference of a fraction of a dB I guess that wouldn't be audible. Thank you! The steel cores definitely lead to more resonance losses than AlNiCo, because the steel core has higher magnetic permeability and higher conductivity than AlNiCo. More steel means more resonance loss, so if you can reduce the size of the steel, you reduce resonance losses, and you reduce inductance at the same time. As a separate matter, higher permeable steel also means higher inductance, where in this case the conductivity is irrelevant. And then as yet another separate matter, the higher permeable steel pole pieces also amplify the magnetism of the moving guitar strings more than AlNiCo will, resulting in higher voltage output, and this is also regardless of the steel's conductivity. Since a single parameter change can have multiple effects, it helps to separate them all out and consider them individually. Swapping the AlNiCo for steel changes several properties of the system all at once. Speaking of multiple consequences for a given parameter, there are two different contexts of eddy current losses at work: 1) the eddy currents caused by the moving guitar string, and 2) eddy currents caused by the magnetic field that is created when current passes though the coil(s). The reason this is important is that if you have a chunk of metal in the center of the pickup (pole pieces), or on the top of the pickup (a cover), it will cause eddy losses due to both the moving guitar string as well as the AC current in the coil, but if you have a chunk on the bottom (base plates) you only get AC/coil eddy currents due to the fact that guitar string is farther away, which according to these tests, causes far less resonance loss. |
|
|
|
Post by antigua on Jun 30, 2017 18:29:16 GMT -5
Thank you for the test! That's interesting, and it shows also that part of the difference from a demo in respect to a proper test may be because of these eddy currents of the close coils in parallel. In fact, when I insert the same inductance and the same resistance in the demo software I get higer Q than your test. But I understand that the demo doesn't consider these eddy currents, it uses the values as if they are from an alnico single coil with those specs, I guess, instead of this "real world" parallel coils with slugs and screws. Another question I thought about this Duncan Custom humbucker is around the wire gauge. I read it's 43 AWG wire, for a resistance of 14+ K. I was wondering how it does compare to a PAF with 42 AWG and around 8 K resistance. I mean: leave alone the Q of the peak, that if I understood correctly is lower because of the high resistance with this Custom humbucker. But..why the high inductance? Is it correct if I assume that not only the Custom pickup has 43 AWG, but it is also overwound if compared to a PAF with 42 AWG? I hope this question is clear enough. How much DC resistance should a 43 AWG wire humbucker have to achieve the same inductance as a PAF with 42 AWG? Thank you very much again! JohnH has a separate parameter you have to input to model the eddy currents. It's still sort of a mystery as to how you model eddy currents, but he has a system worked out that requires curve matching in order to derive the mystery value. A 43 AWG pickup is almost always overwound relative to a 42 AWG pickup. They use 43 AWG instead of 42 AWG in order to fit more wire on the bobbin, not for any other reason. It's both widely believe and demonstrated in testing that the wire gauge is not much of a determinant by itself guitarnuts2.proboards.com/thread/7928/electrical-differences-wire-gauge-practical . The 7 henrty inductance of the Custom and the loaded peak of 2kHz makes the Custom essentially a PAF shaped P-90. Most "PAF" derived pickups on the market using 42AWG have close to 5H inductance with a loaded peak around 2.5kHz. One thing I do find with finer wire, though, is that due to the reduced surface area of smaller wire, you can achieve a lower capacitance with 43 or 44 AWG without having to take as much care to "scatter wind" the coil. |
|
|
|
Post by wgen on Jul 1, 2017 8:43:14 GMT -5
Apart from those related to coil geometry, do you think that, given two pickups both with the same inductance and the same resistance, but One with alnico poles and the other with steel poles (like the example of two coils wired in parallel in respect to a real single coil perhaps)...would this second one have more losses than the first when loaded? Now I guess the question isn't well put because in the Real world I don't know how that could happen (ie, with parallel coils you would have those losses you talked about in your previous post so this aspect would eventually be masked from that, or, if you have steel above magnet bar this pickup would have either different inductance or different resistance if you underwound it for the same inductance, and so on..) So let's just say: in a loaded guitar circuit, would steel poles lead to some dB more of losses in respect to alnicos? I think to remember that in your Filtertrons analysis, all things being on par, the bigger/longer steel screws lead to more losses than shorter ones...so I was now thinking if the same happens changing the pole pieces materials, but I'd be interested in loaded situation because if when loaded it's just a minor difference of a fraction of a dB I guess that wouldn't be audible. Thank you! The steel cores definitely lead to more resonance losses than AlNiCo, because the steel core has higher magnetic permeability and higher conductivity than AlNiCo. More steel means more resonance loss, so if you can reduce the size of the steel, you reduce resonance losses, and you reduce inductance at the same time. As a separate matter, higher permeable steel also means higher inductance, where in this case the conductivity is irrelevant. And then as yet another separate matter, the higher permeable steel pole pieces also amplify the magnetism of the moving guitar strings more than AlNiCo will, resulting in higher voltage output, and this is also regardless of the steel's conductivity. Since a single parameter change can have multiple effects, it helps to separate them all out and consider them individually. Swapping the AlNiCo for steel changes several properties of the system all at once. Speaking of multiple consequences for a given parameter, there are two different contexts of eddy current losses at work: 1) the eddy currents caused by the moving guitar string, and 2) eddy currents caused by the magnetic field that is created when current passes though the coil(s). The reason this is important is that if you have a chunk of metal in the center of the pickup (pole pieces), or on the top of the pickup (a cover), it will cause eddy losses due to both the moving guitar string as well as the AC current in the coil, but if you have a chunk on the bottom (base plates) you only get AC/coil eddy currents due to the fact that guitar string is farther away, which according to these tests, causes far less resonance loss. Thank you very much for both the answers. Now I would be courious about a test with the same exact humbucker pickup and different pole pieces..hex, slugs, screws, alnicos.. even blade!...and all the possible combinations. That would be too complex though, too many variables I guess. However, now I guess that if they put hex poles instead of slugs in Screamin Demon pickups, just to name One, it may be for some reasons. But it always has the screws coil in series, so who knows? How much difference it makes |
|
|
|
Post by antigua on Jul 1, 2017 13:42:48 GMT -5
Thank you very much for both the answers. Now I would be courious about a test with the same exact humbucker pickup and different pole pieces..hex, slugs, screws, alnicos.. even blade!...and all the possible combinations. That would be too complex though, too many variables I guess. However, now I guess that if they put hex poles instead of slugs in Screamin Demon pickups, just to name One, it may be for some reasons. But it always has the screws coil in series, so who knows? How much difference it makes I did an analysis of an SD '59 with various parts removed www.strat-talk.com/threads/seymour-duncan-59-neck-metals-analysis.394003/ , and the overall finding was what you'd expect, that with ever piece of metal you take away, the inductance drops and the Q factor rises, and that removing the core metal (the screws and slugs) has a bigger impact than removing the external metals, such as the base plate or AlNiCo magnet. When it comes to replacing hex, and filister screws, the big differences are #1, how massive is the metal in the core? For example if you look at the PAF Pro and Super Distortion guitarnuts2.proboards.com/thread/7737/dimarzio-super-distortion-analysis-review , they both have filister screws, but the PAF Pro's threading diameter is matched with a vintage PAF, where as the Super D. screws are as wide around as the head of the screw:  The larger cores produce more inductance, because you're taking away air (or plastic as it were) and replacing it with atoms that can assume a magnetic polarity, and aid in the formation of the magnetic field. The #2 difference is whether there is screw material sticking out the bottom, analyzed here guitarnuts2.proboards.com/thread/7877/relation-fillister-screw-length-density . In summary, the longer the screw is out the bottom, the weaker the magnet is out the top, because the lower half of the screw has opposite magnetic polarity, relative to the top, due to the fact that the bar magnet intersects the screws in their middle. So in the image below, if the tops of these screws are south facing, then so are the bottoms of the screws, which is weird when you think about it.  This isn't so much the case with slugs, because rather than meeting the magnet at the middle, the slugs meet the magnet at their ends. This makes the slugs magnetically stronger at the top, because you don't have that magnetic "tug of war" going on - People talk about "pulling" magnetic fields here and there, but that's just a mental abstraction, the physical reality is that magnetic fields sum together (it's made up of lots of little atoms facing in particular directions), and when you have a long leg sticking out the bottom of a pickup, you have a "2 + -1 = 1" situation, where -1 is the leg of the screw poking out the bottom. Snipping off the bottoms of the screws is like erasing that -1 from the equation, so that you're left with just "2", and the pickup therefore becomes magnetically stronger at the strings. I didn't measure inductance with the screw ends clipped off, but the inductance would be a little higher with the long legs, because the extra steel will aid the formation of a magnetic field from the coil's point of view. It's yet another instance where different magnetic fields behave differently to the same pieces of metal, all because of their differing geometric relationship with that metal. The steel never reaches magnetic saturation with these relatively weak magnets, and so it can support magnetic fields pointing in various directions all at the same time, and they all just add together as a magnetic sum. |
|
|
|
Post by wgen on Jul 2, 2017 16:53:02 GMT -5
Thank you very much for both the answers. Now I would be courious about a test with the same exact humbucker pickup and different pole pieces..hex, slugs, screws, alnicos.. even blade!...and all the possible combinations. That would be too complex though, too many variables I guess. However, now I guess that if they put hex poles instead of slugs in Screamin Demon pickups, just to name One, it may be for some reasons. But it always has the screws coil in series, so who knows? How much difference it makes I did an analysis of an SD '59 with various parts removed www.strat-talk.com/threads/seymour-duncan-59-neck-metals-analysis.394003/ , and the overall finding was what you'd expect, that with ever piece of metal you take away, the inductance drops and the Q factor rises, and that removing the core metal (the screws and slugs) has a bigger impact than removing the external metals, such as the base plate or AlNiCo magnet. When it comes to replacing hex, and filister screws, the big differences are #1, how massive is the metal in the core? For example if you look at the PAF Pro and Super Distortion guitarnuts2.proboards.com/thread/7737/dimarzio-super-distortion-analysis-review , they both have filister screws, but the PAF Pro's threading diameter is matched with a vintage PAF, where as the Super D. screws are as wide around as the head of the screw: The larger cores produce more inductance, because you're taking away air (or plastic as it were) and replacing it with atoms that can assume a magnetic polarity, and aid in the formation of the magnetic field. The #2 difference is whether there is screw material sticking out the bottom, analyzed here guitarnuts2.proboards.com/thread/7877/relation-fillister-screw-length-density . In summary, the longer the screw is out the bottom, the weaker the magnet is out the top, because the lower half of the screw has opposite magnetic polarity, relative to the top, due to the fact that the bar magnet intersects the screws in their middle. So in the image below, if the tops of these screws are south facing, then so are the bottoms of the screws, which is weird when you think about it. This isn't so much the case with slugs, because rather than meeting the magnet at the middle, the slugs meet the magnet at their ends. This makes the slugs magnetically stronger at the top, because you don't have that magnetic "tug of war" going on - People talk about "pulling" magnetic fields here and there, but that's just a mental abstraction, the physical reality is that magnetic fields sum together (it's made up of lots of little atoms facing in particular directions), and when you have a long leg sticking out the bottom of a pickup, you have a "2 + -1 = 1" situation, where -1 is the leg of the screw poking out the bottom. Snipping off the bottoms of the screws is like erasing that -1 from the equation, so that you're left with just "2", and the pickup therefore becomes magnetically stronger at the strings. I didn't measure inductance with the screw ends clipped off, but the inductance would be a little higher with the long legs, because the extra steel will aid the formation of a magnetic field from the coil's point of view. It's yet another instance where different magnetic fields behave differently to the same pieces of metal, all because of their differing geometric relationship with that metal. The steel never reaches magnetic saturation with these relatively weak magnets, and so it can support magnetic fields pointing in various directions all at the same time, and they all just add together as a magnetic sum. Thank you very much for the link to the threads, very very good read. I find interesting how different pole pieces damp the peak. And from your test it seems that hex pieces damp just like screws.. So I guess that a double hex coils pickup instead of slugs and screws would lead to more damping, because slugs are a little more trasparent from your test, is it correct? I'd be very courious about a blade pole..How should something like that be considered? Shouldn't it make for some serius damping...? As a side note, I was wondering about the comparison of the Duncan Custom and the Super Distortion. Both are very close DC resistance (14.45 and 14.61), from the graph of the loaded series wiring, let's say, they have almost the same response, but....the Super Distorsion has a much lower inductance, hasn't it? Around 6.5 H, while the Custom has 7.7 So, why the same tonal response from the graph, if not a slightly more damping from the Super Distorsion? Shouldn't the Super D have a higher resonant peak? So, couldn't this be because of the double hex coils instead of the classic screws and slugs of the Custom? Thank you again! |
|
|
|
Post by antigua on Jul 2, 2017 17:31:29 GMT -5
Thank you very much for the link to the threads, very very good read. I find interesting how different pole pieces damp the peak. And from your test it seems that hex pieces damp just like screws.. So I guess that a double hex coils pickup instead of slugs and screws would lead to more damping, because slugs are a little more trasparent from your test, is it correct? I'd be very courious about a blade pole..How should something like that be considered? Shouldn't it make for some serius damping...? If by transparent you mean less eddy current losses, the slugs were less lossy than the screws, interestingly. I think the reason is because, for one thing, there is a "keeper" bar under there also, and the screws are longer, extending out the bottom. They might even be different grades of steel, between the screws and slugs. Do you have an example of a "blade pole" pickup? Do you mean something like this?  Judging from the mass alone, I'd think the damping would be far lower. In general, a wider core will cause higher eddy currents, because more of the magnetic field will pass through it an induce a swirling current, but this blade is so thin that I think the losses would be really low. As a side note, I was wondering about the comparison of the Duncan Custom and the Super Distortion. Both are very close DC resistance (14.45 and 14.61), from the graph of the loaded series wiring, let's say, they have almost the same response, but....the Super Distorsion has a much lower inductance, hasn't it? Around 6.5 H, while the Custom has 7.7 So, why the same tonal response from the graph, if not a slightly more damping from the Super Distorsion? Shouldn't the Super D have a higher resonant peak? So, couldn't this be because of the double hex coils instead of the classic screws and slugs of the Custom? Thank you again! I do see a higher peak f for the Super D, 6.29kHz unloaded and and 2.38kHz loaded, compared to 5.19kHz and 2.03kHz for the SH-5 Custom respectively, which is consistent with the Super D having lower inductance than the SH-5 Custom. We don't know exactly what wire gauge is used, so if the Super D has finer wire it would read the same resistance for a lower wind count. The grades of steel used for the Super D's hex screws versus the SH-5 Customs's screws and slugs might also be a factor. I'd have to do more testing to determine what accounts for the difference. I'm interested in sourcing screws and slugs that are made of specific steel alloys in order to determine the possible range of variation based on alloy. |
|
|
|
Post by wgen on Jul 3, 2017 6:07:44 GMT -5
Thank you very much for the link to the threads, very very good read. I find interesting how different pole pieces damp the peak. And from your test it seems that hex pieces damp just like screws.. So I guess that a double hex coils pickup instead of slugs and screws would lead to more damping, because slugs are a little more trasparent from your test, is it correct? I'd be very courious about a blade pole..How should something like that be considered? Shouldn't it make for some serius damping...? If by transparent you mean less eddy current losses, the slugs were less lossy than the screws, interestingly. I think the reason is because, for one thing, there is a "keeper" bar under there also, and the screws are longer, extending out the bottom. They might even be different grades of steel, between the screws and slugs. Do you have an example of a "blade pole" pickup? Do you mean something like this? Judging from the mass alone, I'd think the damping would be far lower. In general, a wider core will cause higher eddy currents, because more of the magnetic field will pass through it an induce a swirling current, but this blade is so thin that I think the losses would be really low. As a side note, I was wondering about the comparison of the Duncan Custom and the Super Distortion. Both are very close DC resistance (14.45 and 14.61), from the graph of the loaded series wiring, let's say, they have almost the same response, but....the Super Distorsion has a much lower inductance, hasn't it? Around 6.5 H, while the Custom has 7.7 So, why the same tonal response from the graph, if not a slightly more damping from the Super Distorsion? Shouldn't the Super D have a higher resonant peak? So, couldn't this be because of the double hex coils instead of the classic screws and slugs of the Custom? Thank you again! I do see a higher peak f for the Super D, 6.29kHz unloaded and and 2.38kHz loaded, compared to 5.19kHz and 2.03kHz for the SH-5 Custom respectively, which is consistent with the Super D having lower inductance than the SH-5 Custom. We don't know exactly what wire gauge is used, so if the Super D has finer wire it would read the same resistance for a lower wind count. The grades of steel used for the Super D's hex screws versus the SH-5 Customs's screws and slugs might also be a factor. I'd have to do more testing to determine what accounts for the difference. I'm interested in sourcing screws and slugs that are made of specific steel alloys in order to determine the possible range of variation based on alloy. Thank you again. As a blade humbucker I was thinking about more like this one (Dimarzio X2N), which has quite thick blades. Thank you anyway for all the explanations! Attachments:
|
|
|
|
Post by antigua on Jul 3, 2017 14:17:38 GMT -5
Thank you again. As a blade humbucker I was thinking about more like this one (Dimarzio X2N), which has quite thick blades. Thank you anyway for all the explanations! It appears that the X2N's blades are ceramic magnets, so it would have rather low eddy current losses in that case. I bet the Q factor is pretty high, but I don't have one to test. In this test www.strat-talk.com/threads/seymour-duncan-59-neck-metals-analysis.394003/ I never tried, "with baseplate, no screws or slugs" which would most closely match this scenario, but based on other base plate tests, I know that it causes minimal losses by itself, so I imagine the Q factor is pretty high, if you have ceramic cores. |
|
frankfalbo
Meter Reader 1st Class
Posts: 74 
Likes: 1
|
Post by frankfalbo on Jul 3, 2017 16:06:26 GMT -5
It appears that the X2N's blades are ceramic magnets, so they would have rather eddy current losses... The blades aren't ceramic. It's a ceramic base magnet with blades. |
|
|
|
Post by antigua on Jul 4, 2017 2:24:57 GMT -5
It appears that the X2N's blades are ceramic magnets, so they would have rather eddy current losses... The blades aren't ceramic. It's a ceramic base magnet with blades. Do you know if the blades are as thick through at the base as they are at the bobbin tops? If the blades are large and long chunks of steel, then both inductance and eddy current damping should be very substantial. I can't find any pics on the web that reveal whats going on between the opaque base plate and the bobbins. |
|
|
|
Post by wgen on Jul 4, 2017 11:07:11 GMT -5
The blades aren't ceramic. It's a ceramic base magnet with blades. Do you know if the blades are as thick through at the base as they are at the bobbin tops? If the blades are large and long chunks of steel, then both inductance and eddy current damping should be very substantial. I can't find any pics on the web that reveal whats going on between the opaque base plate and the bobbins. While I'd still be very curious if frankfalbo knows something more about the X2N blades... I found a spec about the Duncan JB. If that's correct it should have 44 AWG and about 16 k DC resistance (series wiring). I was wondering how it would compare to the Duncan Custom you tested. I know the JB has alnico 5 magnet and not ceramic, but I'd be interested if 44 vs 43 AWG and a little difference about resistance would lead to much different results. Have you eventually ever take some notes about the JB? Thank you very much anyway! |
|
|
|
Post by antigua on Jul 4, 2017 14:08:53 GMT -5
Do you know if the blades are as thick through at the base as they are at the bobbin tops? If the blades are large and long chunks of steel, then both inductance and eddy current damping should be very substantial. I can't find any pics on the web that reveal whats going on between the opaque base plate and the bobbins. While I'd still be very curious if frankfalbo knows something more about the X2N blades... I found a spec about the Duncan JB. If that's correct it should have 44 AWG and about 16 k DC resistance (series wiring). I was wondering how it would compare to the Duncan Custom you tested. I know the JB has alnico 5 magnet and not ceramic, but I'd be interested if 44 vs 43 AWG and a little difference about resistance would lead to much different results. Have you eventually ever take some notes about the JB? Thank you very much anyway! I don't have a JB on hand, but everything I've seen, including the origination story of the JB, says that it's merely a wildly overwound PAF, with prototypes having been made from the stock pickups of a Gibson Flying V You have to remember the context, that the earliest aftermarket pickup models came out when there was no alternative, and not much experience with such things, so "wildly overwound PAF" or "wildly overwound Strat pickup" was not as boring of a concept back then as it seems to be now. So, having seen similar pickups plenty of times, I'm not inclined to spend money on such a thing, even for the sake of testing. It would feel to me like spending $50 for a resistor of a particular value. As far as 43 vs 44 AWG, all the evidence suggests it makes no difference. That seemed to be the subjective consensus on MEF, and my own test suggested this as well guitarnuts2.proboards.com/thread/7928/electrical-differences-wire-gauge-practical |
|
|
|
Post by wgen on Jul 4, 2017 16:01:36 GMT -5
While I'd still be very curious if frankfalbo knows something more about the X2N blades... I found a spec about the Duncan JB. If that's correct it should have 44 AWG and about 16 k DC resistance (series wiring). I was wondering how it would compare to the Duncan Custom you tested. I know the JB has alnico 5 magnet and not ceramic, but I'd be interested if 44 vs 43 AWG and a little difference about resistance would lead to much different results. Have you eventually ever take some notes about the JB? Thank you very much anyway! I don't have a JB on hand, but everything I've seen, including the origination story of the JB, says that it's merely a wildly overwound PAF, with prototypes having been made from the stock pickups of a Gibson Flying V You have to remember the context, that the earliest aftermarket pickup models came out when there was no alternative, and not much experience with such things, so "wildly overwound PAF" or "wildly overwound Strat pickup" was not as boring of a concept back then as it seems to be now. So, having seen similar pickups plenty of times, I'm not inclined to spend money on such a thing, even for the sake of testing. It would feel to me like spending $50 for a resistor of a particular value. As far as 43 vs 44 AWG, all the evidence suggests it makes no difference. That seemed to be the subjective consensus on MEF, and my own test suggested this as well guitarnuts2.proboards.com/thread/7928/electrical-differences-wire-gauge-practicalYeah, I think I can understand perfectly what you mean... I'll add that, just my opinion of course, these types of pickup are just that, boring and One trick pony, because by definition they miss some detail to achieve not too shrill distorted tones... Fact is, I was simply intrigued just by that thread about different AWG wire. It appeared that you could achieve basically the same response by winding pickups but with thinner wire. So I was just wondering if 44 awg of the JB would make for the same overall inductance of the Custom 43 AWG pickup, just like in your single coils comparison thread.. With some approximation, I guess so..it would be quite similar to the situation of the 3 singles with different AWG wire, and different DC resistance too, but similar inductance |
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Jul 5, 2017 23:12:49 GMT -5
While I'd still be very curious if frankfalbo knows something more about the X2N blades... Not much more can be said. You can get one used and peel the tape off, but I imagine you're all smart enough to think logically about it, and realize that if the blades are about the same thickness as the diameter of their slugs, or a 10-32 set screw, then...Wouldn't it make sense that they go all the way down and make contact with the magnet? They wouldn't add a machining process just to use a keeper bar or something, and it was long before their Airbucker designs.... I don't have a JB on hand, but everything I've seen, including the origination story of the JB, says that it's merely a wildly overwound PAF, with prototypes having been made from the stock pickups of a Gibson Flying V...having seen similar pickups plenty of times, I'm not inclined to spend money on such a thing, even for the sake of testing. It would feel to me like spending $50 for a resistor of a particular value. As far as 43 vs 44 AWG, all the evidence suggests it makes no difference. That seemed to be the subjective consensus on MEF, and my own test suggested this as well guitarnuts2.proboards.com/thread/7928/electrical-differences-wire-gauge-practicalYou honestly believe that wire gauge "makes no difference"? Makes no difference to what exactly? I mean, I'm honestly curious as to what pickups you'd play and say that they sound "the same" or that the difference "doesn't matter". Like, is the Alnico II Pro and the Pearly Gates the "same pickup" to you? Or the 59 and the Custom 5 is maybe a better example. Do you think they sound the same? |
|
|
|
Post by stratotarts on Jul 6, 2017 19:47:42 GMT -5
You honestly believe that wire gauge "makes no difference"? Makes no difference to what exactly? I mean, I'm honestly curious as to what pickups you'd play and say that they sound "the same" or that the difference "doesn't matter". Like, is the Alnico II Pro and the Pearly Gates the "same pickup" to you? Or the 59 and the Custom 5 is maybe a better example. Do you think they sound the same? I think the statement was made in the manner of "all other things being equal". Are the pickup pairs you mention identical in all respects including turns count, except for wire guage? |
|
|
|
Post by antigua on Jul 6, 2017 20:50:25 GMT -5
I think two pickups can seem to sound different simply by having different names.
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Jul 7, 2017 3:05:36 GMT -5
Let me pose the question this way: Are you two suggesting that, you could wind two otherwise identical humbuckers, one with 42 single build poly and the other with 43 single build poly, and by targeting the same resonant peak frequency you could make the two pickups sound indistinguishable to one another?
|
|
|
|
Post by ms on Jul 7, 2017 7:34:53 GMT -5
Let me pose the question this way: Are you two suggesting that, you could wind two otherwise identical humbuckers, one with 42 single build poly and the other with 43 single build poly, and by targeting the same resonant peak frequency you could make the two pickups sound indistinguishable to one another? I do not think so. What Antigua has been saying is that the inductance varies little when winding with different wire sizes, therefore implying that the resonant frequencies, with the cable capacitance the dominant effect over coil capacitance, will vary by little. That is, it happens automatically. Also what he has been saying is that it is the resistance that varies, and that is obviously true. However, the size of the effect this resistance change has on the Q of the coils, and thus the frequency response of the pickup, depends on the relative losses from the wire and from eddy currents. In pickups with steel poles, the eddy losses generally dominate. In pickups with low loss poles, variations in he coil resistance have more of an effect on the frequency response.
So it seems to me that wire size changes in a hum bucker lead to small enough effects that any listening tests must be the carefully done, double blind kind. |
|
|
|
Post by antigua on Jul 7, 2017 12:29:15 GMT -5
There should be at least one forum on the internet where there's some acknowledgement that psychoacoustics are a thing that exists. The premise behind "but don't you hear a difference?" relies not on the idea that one's hearing is as infallible as an electronic measuring device, but more reliable than an electronic measuring device. The concept of psychoacoustics, of course, says our hearing isn't so reliable. If a person wants to pretend their hearing is as objective as a multi-meter, they can find refuge in every single other guitar pickup forum on the internet.
I have a test I'm planning to conduct, ASAP, that involves listening to sound clips side by side, without knowing which sound clip is which, but with these warm summer months, I have to get outside a little.
|
|
frankfalbo
Meter Reader 1st Class
Posts: 74
Likes: 1
|
Post by frankfalbo on Jul 7, 2017 14:06:00 GMT -5
The premise behind "but don't you hear a difference?" relies not on the idea that one's hearing is as infallible as an electronic measuring device, but more reliable than an electronic measuring device. I have a test I'm planning to conduct, ASAP, that involves listening to sound clips side by side, without knowing which sound clip is which, but with these warm summer months, I have to get outside a little. Ok so 2 things. First, if the subject(s) DO hear (or repeatably sense) a difference between A and B while test equipment shows sameness, then in my experience the test methods just haven't brought out the deltas and need to be reconsidered. (Sorry in advance for storytime, but it's required) I was just talking to someone today about my experience with a Swedish virtuoso who, repeatedly told some engineers that A and B were not the same, while the engineers (to this day) could never produce a test that bore that out. However, the artist "drives" these items like a race car driver, faster and louder than any of us will. So I flew to his house, and observed his situation. Guitar>dimed pedal>dimed Marshall>screaming 4x12's>mics>compressor/mic pre>console>studio monitors...listening at very reasonable volume levels. The result was, that now, I too, could hear and sense what he was talking about when he switched from product A to product B. Whether he was playing or I was playing. He was not wrong. I came back to headquarters and loosely duplicated the setup. 50 watt Marshall mic'd up in the sound room with extension cables upstairs to listen in headphones while the amp was raging away. Under that situation, I was able to show others what he was talking about. It never mattered because we didn't make the product. But it's a teaching moment. What's the difference between 100 and 104mph? Nothing if you're obeying the speed limit. But if you're racing it could be the difference between whether there's a shimmy around the turns. All this to say, your blind sound tests are simply not effective. They risk proving nothing, or a falsehoods. The trouble is that if I have 2 pickups that are close to one another, I might be able to play a certain way to show you that when you do ABC, they don't sound that different. But when you do XYZ, you can see one performs better at X, the other performs better at Y, and so on. The controlled strum test, or one player playing their style of play on two different pickups, is ineffective if that style of play, on that particular rig, doesn't magnify the deltas. Beyond that, the sensation and response of the gain chain to the dynamics are often difficult to hear in a recorded track, but very easy to feel when you're behind the guitar, much like the race car driver can feel the shimmy but the viewer sees nothing while watching the race. A blind listener test is much like testing different vehicles by riding blindfolded passenger, while someone else drives. You will only sense a limited number of differences, while the driver will sense more because they're behind the wheel. I'm not here to pee in anyone's cornflakes, I'm telling you that you could make a 42/43/44 pickup set that sounds similar on a recording, but in my hands I could play in such a way to demonstrate why the differences are more meaningful than just what a recorded sound clip can convey. |
|
Deleted
Deleted Member
Posts: 0
Likes:
|
Post by Deleted on Jul 7, 2017 14:25:36 GMT -5
I guess its the first time I read this by someone : it is listening + playing, AT THE SAME TIME!
When I flipped 81/85 the other way around, I could very much hear the difference. Tell me to blind test this by giving my guitar to another person and I might fail all of the tests. The end result may be the same or similar. But the player (driver) gains confidence by knowing that he/she has the right tools which can boost his/her performance.
I think we got an oxymoron here : ppl spend pages after pages on the electric characteristics of guitar pups, just to conclude that in the end not much or all of this matters. Heck, we live in the digital era, any signal can be easily converted real time into almost anything. What I mean, if someone wants to keep on being an "analog" diehard desperado (which we gotta admit is dealing with 1920's technology), he better stick to it in every detail, and leave the opposition's basic narrative (blind tests, etc) aside. Ok English is not my native tang, I might have failed giving the exact meaning here, what I write describes best the "wood matters" vs "doesn't matter" classic holy war, but this can be easily transcribed into more modern parallels as well (electricity, digital, modelling, synths etc ...).
|
|
|
|
Post by ms on Jul 7, 2017 14:33:24 GMT -5
I was just talking to someone today about my experience with a Swedish virtuoso who, repeatedly told some engineers that A and B were not the same, while the engineers (to this day) could never produce a test that bore that out. ... All this to say, your blind sound tests are simply not effective. They risk proving nothing, or a falsehoods. The trouble is that if I have 2 pickups that are close to one another, I might be able to play a certain way to show you that when you do ABC, they don't sound that different. But when you do XYZ, you can see one performs better at X, the other performs better at Y, and so on. But if your Swedish virtuoso cannot hear the difference when he does not know whether it is A or B, then it is not a difference that matters. If he can hear it without knowing if A or B, then it does not matter whether you can measure with test equipment or not, because it is none the less real. As for the second part of the quote, you do what you describe there, you are fooling yourself and anyone else you can convince. It is like that silver wire. A six percent increase in conductivity in a pickup where the resistance of the coil is of secondary importance and not only can you hear the difference (a possibility, but unproven), but it is an improvement, something worth spending quite a bit of money to obtain. I do not think so, not without significant double blind tests. |
|
