|
|
Post by antigua on Jan 3, 2018 2:52:24 GMT -5
One popular mod DIY guitar modders like to do is create "hybrid" humbuckers, where you take the coils from two different P.A.F. style humbuckers and place them together in a single pickup for a "half-this", "half-that" pickup. What is known from a circuit analysis standpoint is that you won't get two strong, different resonant peaks from the two coils, but rather you calculate for the sum of inductance from the two coils, and you get a single inductance and a single prominent resonant peak. So this sort of mod is not quite as exciting as people might thing, because the two coils sort of wash together rather than retain distinct transfer functions. Hot coil vs. cool coilBut I got to wondering, if one coil has a higher inductance than another, will that coil also produce more voltage? If coils were merely inductors, such a question wouldn't apply, but since pickup coils are also a voltage source, this question arises. So I did a test where wired the screw coil of a Jazz and the screw coil of a JB together in series (neither slug coil is involved), and excited the coils in the same manner, and compared the output voltage, and it turns out that the JB, with the higher inductance produced quite a bit more voltage than the Jazz coil. The inductance of the Jazz by itself is 3.9 henries, the JB 8.0 henries. The inductance of the Jazz screw coil alone is 1.7H, and the inductance of the JB screw coil is 3.5H, about double that of the Jazz screw coil. Inductances in series add together, so 1.7H +3.5H = 5.2 henries, and I get that measured value when I test the two coils in series as well. Here are the measurement bode plots, unloaded, and loaded with a standard dummy load: no load:   with test load:  It can be seen that 3.4dB boost for the JB, the coil which has about double the inductance when measured by itself. Note that the loaded resonant peak is the same for both coils, despite the difference in amplitude, which proves that the inductances of the coil combine to create a single dominant resonance. Here is a plot showing just the JB's screw coil under load, with and without the Jazz' screw coil in series:  Note that the JB's screw coil in isolation has a resonant peak that is is about 550Hz greater, showing that even though the resonance of the JB's coil dominant compared to the Jazz, the two coils' L and C values still combine to produce a lower resonant peak. The take away from this is that when you create a hybrid coil, and wire it in series, the coil with more turns of wire dominate in terms of signal output, and so the orientation of the hybrid pickup will have a real effect on the harmonic voicing of the pickup. The second take away is that the coils combine to create a new transfer function, they don't mix together in a way that renders dual peaks, or retains their original transfer functions. Slug coil vs. screw coilA side question is, how much different is the slug coil from the screw coil, all other things being approximately equal? The JB's screw and slug coils measure 8.22k for the screw and 8.06k for the slug coil, and the inductance were nearly equal, 3.54H for the screw and 3.52H for the slug coil. So I wired the JB in standard series mode, and measured the output with the exciter over each coil, and the result was that the slug coil produced slightly more voltage: no load:  with test load:  It can be seen that the slug coil produces slightly more voltage than the screw coil, even though the screw coil appeared to be ever so slightly hotter. This makes sense, since there is clearly more steel mass near the guitar string / exciter coil with a slug than with a screw, and so there is a greater amplification of flux, or in technical jargon, there is a lower magnetic reluctance between the slug and the magnetomotive force, than there is with the screw. Strangely, the plot with the added parallel load showed a greater difference, 0.6dB versus 0.3dB. Usually added load tends to diminish differences, but in either case, the fact that the difference is less that 1dB means you won't likely hear a difference, so these coils can be considered "matched" for all intents and purposes. The take away here is that when creating hybrid coils, you don't have to worry too much about whether you're using the screw or the slug coil, because all other things being equal, they will perform about the same. Pic:  |
|
|
|
Post by wgen on Jan 3, 2018 5:49:44 GMT -5
Thank you, truly great work as always.
I'd like to ask you a couple of questions about something I don't quite understand, about the 2 screw coils in series;
1) apart from the voltage results (ie, the loaded JB screw coil having 3.4 db of boost in respect to the other coil), do the resonant peak for both coils, which I see is the same around 2.5-3 Khz in the plot with standard load for both coils, is that peak due to which inductance?
I mean; you measured 5.2 henries for both screw coils together in series, is it correct? Is this the inductance "of reference" for that 2.5-3 Khz peak?
2) this lead me to another question; I remember some of your posts in some forum in the past, which I think already answered to this, but now I don't remember where I read those, I can't find those useful informations again.
So, please excuse me if I ask if you could repeat what happens if you had two coils in parallel, instead of in series. I'm talking specifically about two very different coils like in this thread, with different inductances, as in the example of the 2 screw coils of a JB and a Jazz pickup.
Which would be the inductance "of reference", and therefore the resonant peak, of the two coils together in parallel?
Also, would the JB screw coil still have more voltage output than the screw coil of the Jazz, as in when in series?
Thank you very much in advance!
|
|
|
|
Post by antigua on Jan 3, 2018 8:57:05 GMT -5
Thank you, truly great work as always. I'd like to ask you a couple of questions about something I don't quite understand, about the 2 screw coils in series; 1) apart from the voltage results (ie, the loaded JB screw coil having 3.4 db of boost in respect to the other coil), do the resonant peak for both coils, which I see is the same around 2.5-3 Khz in the plot with standard load for both coils, is that peak due to which inductance? I mean; you measured 5.2 henries for both screw coils together in series, is it correct? Is this the inductance "of reference" for that 2.5-3 Khz peak? 2) this lead me to another question; I remember some of your posts in some forum in the past, which I think already answered to this, but now I don't remember where I read those, I can't find those useful informations again. So, please excuse me if I ask if you could repeat what happens if you had two coils in parallel, instead of in series. I'm talking specifically about two very different coils like in this thread, with different inductances, as in the example of the 2 screw coils of a JB and a Jazz pickup. Which would be the inductance "of reference", and therefore the resonant peak, of the two coils together in parallel? Also, would the JB screw coil still have more voltage output than the screw coil of the Jazz, as in when in series? Thank you very much in advance! When it comes to the voltage generated, it's only the inductance of the coil receiving the flux change that is operative, because in that capacity, each screw coil is acting like the second coil of an individual transformer, where the exciter coil or the guitar string are acting as the primary coil. This is the case even if the pickup is wired in parallel, so the JB's screw coil would still generate more voltage, even if it's wired in parallel with the Jazz screw coil. But when it comes to the LC filtering that creates the resonant cut off, then the inductance and capacitance of both coils are in play at that point, because whereas the voltage generated by each coil depends on the relationship between that coil and the magnetomotive energy source, the LC filtering is a result of the entire circuit combined, and that includes the guitar cable as well. The inductances of the two screw coils in series add together; 1.7H + 3.5H = 5.2H, so the 5.2 value is the one that factors into the dominant resonant peak. It's such a dominant peak that any individual coil resonance that might exist is nowhere to be seen in the plot. Note that the capacitance that factors into the dominant resonant peak also combines in series for a single calculated value, but unlike inductors, the series capacitance actually becomes smaller. When then coils are wired in prallel, then the reverse is the case; the inductance drops and the capacitance goes way up, but despite swapping the relative magnitudes of L and C, the resonant peak will usually still be somewhat higher in parallel mode. |
|
|
|
Post by ms on Jan 3, 2018 12:34:31 GMT -5
When it comes to the voltage generated, it's only the inductance of the coil receiving the flux change that is operative, because in that capacity, each screw coil is acting like the second coil of an individual transformer, where the exciter coil or the guitar string are acting as the primary coil. This is the case even if the pickup is wired in parallel, so the JB's screw coil would still generate more voltage, even if it's wired in parallel with the Jazz screw coil. But when it comes to the LC filtering that creates the resonant cut off, then the inductance and capacitance of both coils are in play at that point, because whereas the voltage generated by each coil depends on the relationship between that coil and the magnetomotive energy source, the LC filtering is a result of the entire circuit combined, and that includes the guitar cable as well. The inductances of the two screw coils in series add together; 1.7H + 3.5H = 5.2H, so the 5.2 value is the one that factors into the dominant resonant peak. It's such a dominant peak that any individual coil resonance that might exist is nowhere to be seen in the plot. Note that the capacitance that factors into the dominant resonant peak also combines in series for a single calculated value, but unlike inductors, the series capacitance actually becomes smaller. When then coils are wired in prallel, then the reverse is the case; the inductance drops and the capacitance goes way up, but despite swapping the relative magnitudes of L and C, the resonant peak will usually still be somewhat higher in parallel mode. It is a little more complicated that that. If the coils are put in parallel, and we want the voltage across the two coils: Consider the impedance of each coil as one element in a voltage divider. First, what happens with the voltage generated in the JB? In this case the JB acts as the series leg of the voltage divider, while the jazz is the shunt leg. So what you get across the pickups is: Vjb(Zjazz)/(Zjazz + Zjb). The voltage from the jazz is processed in a similar way: Vjazz(Zjb)/(Vjb + Zjazz). The total voltage is the sum of the two. (The Zs are frequency dependent; so this is not so simple.) Look at a simple example: Suppose that the two coils have identical impedances and generate identical voltages. The the two legs of the voltage dividers are the same and you get half the voltage from each coil. They add, and the result is equal the voltage from one coil, a result that you already knew from inspection. |
|
Deleted
Deleted Member
Posts: 0
Likes:
|
Post by Deleted on Jan 3, 2018 15:17:07 GMT -5
Vjazz(Zjb)/(Vjb + Zjazz). maybe Vjazz(Zjb)/( VjbZjb + Zjazz) a typo? |
|
|
|
Post by reTrEaD on Jan 3, 2018 15:38:31 GMT -5
No typo. Simple voltage division equation.
|
|
|
|
Post by antigua on Jan 3, 2018 16:06:59 GMT -5
It is a little more complicated that that. If the coils are put in parallel, and we want the voltage across the two coils: Consider the impedance of each coil as one element in a voltage divider. First, what happens with the voltage generated in the JB? In this case the JB acts as the series leg of the voltage divider, while the jazz is the shunt leg. So what you get across the pickups is: Vjb(Zjazz)/(Zjazz + Zjb). The voltage from the jazz is processed in a similar way: Vjazz(Zjb)/(Vjb + Zjazz). The total voltage is the sum of the two. (The Zs are frequency dependent; so this is not so simple.) Look at a simple example: Suppose that the two coils have identical impedances and generate identical voltages. The the two legs of the voltage dividers are the same and you get half the voltage from each coil. They add, and the result is equal the voltage from one coil, a result that you already knew from inspection. I think you're talking about the final output voltage calculation, which was important when settling arguments about parallel versus split output, but what's at issue in this case is which coil contributes more voltage, or a stronger signal, when the pickup is wired in either series or parallel modes. This is important because suppose you actually made a hybrid pickup out of a JB and a Jazz, and you wanted to put it in the bridge. It appears that if you put the "louder" JB coil bridge-side, you will get more bright "bridge" harmonics in the final output, but if you put the JB coil neck-side, you will get more of that larger-displacement lower harmonics that are further from the bridge, and so you'd get more low end and mids from the pickup. I'll set this test back up later and try it in parallel to get practical measurements and see what happens. My understanding is that the JB coil should once again produce a higher voltage when excited, and I'll be curious to see if that bares out, but also the degree to which is does, relative to series mode. |
|
Deleted
Deleted Member
Posts: 0
Likes:
|
Post by Deleted on Jan 3, 2018 16:09:02 GMT -5
No typo. Simple voltage division equation. WTF are you talking about? In which book you read is it considered normal to add impedance to voltage, or apples to oranges ? Look at the freaking denominator. |
|
|
|
Post by antigua on Jan 3, 2018 16:10:55 GMT -5
Looks like a copy paste error when you view them side by side.
Vjb(Zjazz)/(Zjazz + Zjb)
Vjazz(Zjb)/(Vjb + Zjazz)
|
|
Deleted
Deleted Member
Posts: 0
Likes:
|
Post by Deleted on Jan 3, 2018 16:12:57 GMT -5
Antigua, while we'are at it, some quickie, how is on-parallel volatge-wise compared to the output of the strongest / weakest coil? Is it smth in between? Cause ppl who have experimented with dummy coils in parallel said that the result was weak.
|
|
|
|
Post by antigua on Jan 3, 2018 16:29:39 GMT -5
Antigua, while we'are at it, some quickie, how is on-parallel volatge-wise compared to the output of the strongest / weakest coil? Is it smth in between? Cause ppl who have experimented with dummy coils in parallel said that the result was weak. I'll do a practical parallel test in a few hours in order to see how the same test varies when the screw coils are in parallel. It stands to reason that a parallel dummy coil will gut the output level, because you're giving the current another route around the pickup, while doing nothing to increase the voltage otherwise. Resistances or impedances in parallel means less of either or both, meaning lower voltage across them. I did a test a few months ago that demonstrated this guitarnuts2.proboards.com/post/80448/thread . If you take a humbucker and you wire it in parallel, both coils are generating voltage, and they output level is nearly the same as if you split the pickup and use either coil completely by itself. But if you keep it wired in parallel, and simply removed one of the coils from the humbucker housing, you now have a parallel dummy coil off to the side that is not also generating voltage. Since only one of the two coils is producing voltage, the voltage drops way below what it would be split, or parallel with both coils generating a voltage. In the practical test I linked to, there was a 4.7dB drop, and so I'd expect a similar substantial drop in the case of a parallel dummy coil. I wonder about the Ilitch dummy coil, I haven't researched it yet, but it gets good reviews and I'm wondering what parallel or series consequences is has on the passive circuit of a Strat. |
|
|
|
Post by ms on Jan 3, 2018 19:16:18 GMT -5
No typo. Simple voltage division equation. Sorry about the typo. |
|
|
|
Post by reTrEaD on Jan 3, 2018 20:04:55 GMT -5
Yeah, I looked back to your post and saw the first expression was correct. Missed the error in the second. Mea Culpa. |
|
Deleted
Deleted Member
Posts: 0
Likes:
|
Post by Deleted on Jan 4, 2018 2:00:40 GMT -5
Antigua, while we'are at it, some quickie, how is on-parallel volatge-wise compared to the output of the strongest / weakest coil? Is it smth in between? Cause ppl who have experimented with dummy coils in parallel said that the result was weak. I'll do a practical parallel test in a few hours in order to see how the same test varies when the screw coils are in parallel. It stands to reason that a parallel dummy coil will gut the output level, because you're giving the current another route around the pickup, while doing nothing to increase the voltage otherwise. Resistances or impedances in parallel means less of either or both, meaning lower voltage across them. I did a test a few months ago that demonstrated this guitarnuts2.proboards.com/post/80448/thread . If you take a humbucker and you wire it in parallel, both coils are generating voltage, and they output level is nearly the same as if you split the pickup and use either coil completely by itself. But if you keep it wired in parallel, and simply removed one of the coils from the humbucker housing, you now have a parallel dummy coil off to the side that is not also generating voltage. Since only one of the two coils is producing voltage, the voltage drops way below what it would be split, or parallel with both coils generating a voltage. In the practical test I linked to, there was a 4.7dB drop, and so I'd expect a similar substantial drop in the case of a parallel dummy coil. I wonder about the Ilitch dummy coil, I haven't researched it yet, but it gets good reviews and I'm wondering what parallel or series consequences is has on the passive circuit of a Strat. thanx a lot for the fine explanation! About Ilitch the cost is prohibitive. I better go grab some entwistle or eyguitarmusic, or for that money even bite the bullet for some Dimarzio Areas. |
|
|
|
Post by antigua on Jan 4, 2018 3:56:57 GMT -5
Well my guess was wrong, it turns out that with the screw coils of the JB and Jazz in parallel, for the same level of external excitation, the JB coil produced less voltage than the Jazz coil. The lower red line is the JB, the higher black line is the Jazz: no load:  with test load:  The inductance of the Jazz and JB screw coils in parallel measured 1.11 henries. It's sinking in slowly, but based on the explanations above, I take it the laymen's explanation would be that, when in parallel mode, if the higher inductance coil is also the voltage source, then it's rather heavily loaded down by the lower inductance coil beside it. In the opposite case; when the lower inductance coil is the voltage source, it is not as heavily loaded down by the higher inductance coil beside it. It also appears this trend holds with LTSpice when using the standard three part pickup model, where there are two coils representing a humbucker, with each coil having an AC source, and the voltage being switched between the two using a step command and two tables. The green line represents the "JB" AC source producing a voltage, and the blue line represents the "Jazz" AC source producing a voltage. This first screen shot shows the series configuration. The curve doesn't match the measured bode plot, but even still, the "JB" line is definitely above the "Jazz" line for the most part. This modeled line differs from the measured bode plot in that the modeled version shows the difference increases with frequency, where as the series bode plot in the first post shows a uniform difference that is not frequency dependent. I tried moving some components around, but it didn't make much difference.  Now the model is changed around so that the two coils are in parallel, and now the blue line, the "Jazz" coil, is well above the green "JB" line. And with this parallel model, the curves also match a lot more closely align with the measured bode plot, as there is a uniform degree of separation that doesn't change with frequency.  |
|
|
|
Post by wgen on Feb 16, 2018 10:58:35 GMT -5
I'm writing here again to ask something which I think is quite strictly related to the situations already discussed.
I'd like to better understand which is the pickups tonal response when you have two pickups wired in parallel, let's take the two typical Telecaster single coils with the pickup switch in the middle, but; imagine you have the neck pickup with a brass cover, which severely dampens treble frequencies, like one of those cheap Tele knockoffs, while the bridge pickup is your typical Tele Bridge uncovered.
So, how much would the eddy current losses coming from the neck pickup dampen the highs of the final result, considering that you also have the bridge pickup selected, which is wired in parallel to that (remember, I'm talking about the in-between position of the switch in this case)..?
Thank you very much in advance!
|
|
|
|
Post by antigua on Feb 16, 2018 18:30:20 GMT -5
I'm writing here again to ask something which I think is quite strictly related to the situations already discussed. I'd like to better understand which is the pickups tonal response when you have two pickups wired in parallel, let's take the two typical Telecaster single coils with the pickup switch in the middle, but; imagine you have the neck pickup with a brass cover, which severely dampens treble frequencies, like one of those cheap Tele knockoffs, while the bridge pickup is your typical Tele Bridge uncovered. So, how much would the eddy current losses coming from the neck pickup dampen the highs of the final result, considering that you also have the bridge pickup selected, which is wired in parallel to that (remember, I'm talking about the in-between position of the switch in this case)..? Thank you very much in advance! According to member "ms", and my best understanding, eddy currents that act upon the magnetic field of the coil itself manifest as parallel resistance, which is why they dull the resonance, the same way the parallel resistance of the volume and tone pots due, except that it's frequency dependent like a reactance, and I'm not sure whether or not it impacts the phase of the signal, either. But based on that understanding alone, yes, eddy currents from the neck would decrease the overall Q factor of the neck and bridge in parallel or series. This is something that can be tested, I'll have to put it on the to do list. |
|
|
|
Post by Yogi B on Mar 12, 2018 2:29:22 GMT -5
I've had this rattling around my head since you posted, in particular this sentence stood out:Note that the resonant peak is the same, which proves that the inductances of the coil combine to create a single dominant resonance.
Looking at the your first diagram (series, unloaded) doesn't seem odd that the Jazz coil's apparent peak is so small compared to the JB coil?
Plus the Jazz coil has lower inductance, capacitance and resistance all of which would normally indicate a higher resonant frequency.
What I believe is causing this apparent behaviour is the inactive JB coil is acting as a notch filter of sorts, the small peak we see in the Jazz is the beginning of it's resonant peak before the filtering of the JB coil interferes. You can just see the very beginning of the Jazz's true resonant peak at the far right. (It's a shame the series plots stop at 10kHz, unlike those for parallel.)
(It's worth noting the reverse would also be true, when the JB coil is driven and the Jazz coil is doing the filtering. It's just that the notch that would likewise occur at around the Jazz coil's true resonant frequency, thus is also missing of the end of the plot.)
Thus I'd expect the combined series unloaded frequency response with both coils driven to actually have two peaks, along with a fairly deep valley between them.
When in parallel I would expect only a single resonant peak in each case (JB driven, Jazz driven, both driven), I would also expect the resonant frequency to be the same across the board (and roughly at the centre of the series peaks). I believe the main cause of this is both coils' capacitances being in parallel with the driven coil(s), whereas in series the un-driven coil's capacitance is in series.
That's mainly the notch affect I'd expect from the filtering from the unused coil, in addition you can see the beginning of a notch in the JB's line at the very end about where I would expect the natural resonant frequency of the Jazz coil . Also note that in LTSpice you've set the frequency axis to linear, not logarithmic! (That probably happened automatically since you're doing a linear AC sweep).
As for the difference amplitude difference between those lines or indeed which line is higher, I wouldn't read too much into it -- we're still missing an important piece of information needed to properly answer that question: the current induced in (and thus voltage produced by) the pickup. In your model the voltage supplied by each pickup (when active) is 10V in each case, yet we know that the JB coil produces ~3.4dB more output from the series tests.
|
|
|
|
Post by antigua on Mar 12, 2018 3:54:12 GMT -5
I've had this rattling around my head since you posted, in particular this sentence stood out:Note that the resonant peak is the same, which proves that the inductances of the coil combine to create a single dominant resonance.
I wasn't specific, but I was referring to the loaded case. I don't care a whole lot about the unloaded case, since it's never met in situ. I should have carried the plot past 10kHz though, I don't remember what my thinking was there. I got into doing this because there was talk somewhere about hybrid coils simultaneously embodying the qualities of two pickups at once, which is not at all how it works, regardless of whether the pickup is loaded or not. In situ, you're getting the same function for both coils, with only the amplitude differing. There's also DiMarzio's "dual resonance" business, which is much more technically dubious, because in that case both coils have the same turn count, and have RLC values that only differ slightly. I'll take another look at the LTSpice model. I don't care too much if the x axis is lineal or log, as it's providing me the same information either way. What's more irritating is the auto scaling y axis, I'll see what appears to be a huge amplitude spike, only to realize the y axis' range just became a lot smaller. |
|
|
|
Post by gitpiddler on Mar 12, 2018 9:21:30 GMT -5
Nice work! My JB meets EVH Tele emphasizes the high strings at low volumes up to 6-7. The bass rolls in between there and 8-9. Above that the ice-pick comes out. Direct-mounted, wax-potted, series-wired it's a brick that loves a cranked amp. Nice to see a graph of what I've heard since '84.  |
|
|
|
Post by Yogi B on Mar 12, 2018 18:44:14 GMT -5
I wasn't specific, but I was referring to the loaded case. I don't care a whole lot about the unloaded case, since it's never met in situ. Yeah I agree, though it's still an interesting thing to observe. A linear y-axis I can work with -- x-axis not so much, the scaling is just too far off for me. Or to have one trace hit essentially zero Volts so the others get squashed too close together -- yeah, it's irritating. There's a couple of workarounds, but they're also not too brilliant. The reason I suddenly jumped on this was I was thinking about applying this sort of logic to model the blending of pickups with volume controls. |
|
|
|
Post by antigua on Mar 12, 2018 22:39:42 GMT -5
Here is the JB + Jazz screw coils in series, extended out farther. In a no load context. The resonance of the Jazz coil does cause a distinct impedance. And here is another plot, looking at the loaded peak of the JB's screw coil with and without the Jazz coil in series. Putting the Jazz screw coil in series bring the peak down by about 550Hz, and lowers the Q factor a bit. |
|
|
|
Post by wgen on Jul 25, 2019 13:03:16 GMT -5
I'm writing here again to ask something which I think is quite strictly related to the situations already discussed. I'd like to better understand which is the pickups tonal response when you have two pickups wired in parallel, let's take the two typical Telecaster single coils with the pickup switch in the middle, but; imagine you have the neck pickup with a brass cover, which severely dampens treble frequencies, like one of those cheap Tele knockoffs, while the bridge pickup is your typical Tele Bridge uncovered. So, how much would the eddy current losses coming from the neck pickup dampen the highs of the final result, considering that you also have the bridge pickup selected, which is wired in parallel to that (remember, I'm talking about the in-between position of the switch in this case)..? Thank you very much in advance! According to member "ms", and my best understanding, eddy currents that act upon the magnetic field of the coil itself manifest as parallel resistance, which is why they dull the resonance, the same way the parallel resistance of the volume and tone pots due, except that it's frequency dependent like a reactance, and I'm not sure whether or not it impacts the phase of the signal, either. But based on that understanding alone, yes, eddy currents from the neck would decrease the overall Q factor of the neck and bridge in parallel or series. This is something that can be tested, I'll have to put it on the to do list. Dear Antigua, this message just to ask you if you ever had the time to do a test with two coils in parallel, or in series, to see how much the eddy current losses of one coil only might impact the final sound and respinse. Thank you! |
|
|
|
Post by antigua on Jul 26, 2019 14:32:22 GMT -5
According to member "ms", and my best understanding, eddy currents that act upon the magnetic field of the coil itself manifest as parallel resistance, which is why they dull the resonance, the same way the parallel resistance of the volume and tone pots due, except that it's frequency dependent like a reactance, and I'm not sure whether or not it impacts the phase of the signal, either. But based on that understanding alone, yes, eddy currents from the neck would decrease the overall Q factor of the neck and bridge in parallel or series. This is something that can be tested, I'll have to put it on the to do list. Dear Antigua, this message just to ask you if you ever had the time to do a test with two coils in parallel, or in series, to see how much the eddy current losses of one coil only might impact the final sound and respinse. Thank you! Modelling with the "Tele Tucson" spice model has shown that a transformer is an effective analogy for eddy currents, which is like a parallel short across the pickup that increases with frequency. The parallel short is across both the neck pickup with the cover, as well as any pickup connected in parallel, based on this model. This means the brass cover on the neck pickup will decrease the treble response in the bridge pickup as well. If it were true that eddy currents were a series resistance, the eddy currents in the second coil would cause an increase in treble in the primary, because the second coil acts as a shunt upon the first, and so if the eddy currents increased the resistance of the second coil, it would become less of a shunt at those higher frequencies, and so overall the treble would increase. Since both modeling and reason suggest that eddy currents are a parallel load, treble should drop overall. I forgot to test this out the last time I had the setup in place, but it's a simple test; put two single coils in parallel with the coils several inches apart, then put the excitation coil over one coil and measure the frequency response of the network. Then, put a brass plate (something big an conductive) over the other co, and repeat, and see how the response curve is effected by the presence of the eddy currents in the second coil. |
|
|
|
Post by straylight on Jul 26, 2019 22:02:47 GMT -5
This has me thinking, is some of the intent of Dual Resonance to have a pair of coils from a neck and a bridge pickup that commbine together nicely in parallel either with each other or with a single coil? Similar inductances between coils on a pickup but different DC resistances suggests same wind count and different wire gauges. I don't think I have a set to hand with matching resistances for neck higher resistance wire and bridge lower resistance wire combinations, but they converge some.
I'm not sure what makes an optimum match, I'd guess matching impedances and matching signal voltages which gets tricky to simulate as the amplitude of string vibration is less at the bridge?
I think, but I haven't really set out to invesigate by winding many pairs of coils mismatched by increasing amounts, that in series mismatched coils give me a wider, flatter resonat peak. However I've been making my wind count mismatch mirror my core mismatch so fillister screw coils are wound more than slug coils and if i'm using double set screws in a traditional cover (so longer set screws on the "screw" side) the screw coil is underwound I've matched inductances. I'm getting the results I want for humbucker voicings, but I've not really thought about pickup combinations. I probably should as I really like the neck parallel and inside coils configurations typical of positions 2 and 4 of HH ibanez guitars.
Modelling parallel combinations, particularly with coils that aren't next to each other looks like a whole new can of worms. HSS guitars suddenly got interesting as I'm looking at each coil in the bridge humbucker I'm currently using being about 3H, compared to very overwound strat styled singles at ~3H and ~2.2H for the neck and middle singles which in series would be a decent inductance to approximate a neck humbucker and I'm wondering if there's a noticable difference in building my bridge humbucker like two strat singles with alnico rods and more winds.
|
|
