Magnet Analysis in Consistent Humbucker - Advice appreciated

[gpdb]

Hi All,

I'm a new member of GN2. First I'd like to thank everyone here for the vast amount of knowledge that has been created here, it has been incredibly helpful in my own path towards investigating pickups. A quick note about myself before getting into the charts.

I've always been interested in pickups, but frustrated by the difficulty required to understand what's available, what they sound like, and what they are similar to. I've been attempting to address these problems with a website I've created called Guitar Pickup Database (guitarpickupdatabase.com). The goals of the website are:

1. Provide an easy way to see all available pickups compared to each other, starting with publicly available specs. This allows you to see similar offerings from all brands side by side. This will grow into providing non-public specs like inductance, bode plots, gauss, and anything else that can provide valuable insight into how the pickup will sound. Lastly, the actual sound of the pickup. I have purchased specific guitars that will be used to demo each pickup with a consistent pickup height, and provide the same playing examples through a DI and an amp.
2. Provide a knowledgebase for pickup information, with the intent of blending technical knowledge into it's most simplistic and useful form. The intent is not necessarily myth-busting, but to provide real data for real questions, and not to rely on opinion.
3. Just being up front, money. The website will have affiliate links on each pickup, advertising, etc. The only way I can continue this will be for it to actually make money. If not, it will just be an expensive side project.

The website is up and running and I would love any feedback that can be provided. I started this before I found Antigua's Echos of Mars site which is also fantastic, though I believe our paths are slightly different. The website is still very much being developed, but currently every major pickup brand is on the site. There's also a great sidebar filter that can help you narrow down what you're looking for.

Now, back to what I'm asking help with! 

I recently purchased a PC scope (Velleman 200) and Ken Willmot's Integrator to create bode plots. I took a Dimarzio PAF 59, and created plots with the original magnet (degaussed A5), along with A2, A3, A4, fully charged A5, A8, and C8. You can see my readings in the image below (y-axis is dBV):



I was a bit surprised to see that as the magnet gauss went up, the overall plot went down. However, these were my first plots so I'm not sure if I've done them correctly or not. I was using the standard load of 200K/470pF for each. So my question is a bit nebulous, but I'm curious if these seem... correct. There's a few things I'm aware I probably need to do for the next analysis, which would be creating a jig to keep the frequency generator position absolutely consistent, as well as potentially some EMI that Ken warned me of. 

Anyways - I appreciate any responses and look forward to the conversation. Again, thank you to all who came before and paved the way for this data exploration to exist.  

[gpdb]

The chart wasn't appearing at first, it should be fixed now.

[aquin43]

A probable explanation, which I think has been mentioned here before:

The various magnets have different magnetising force and different permeabilities. The magnetising variations alter the inductances of the coils by changing the working point of the pole pieces. The permeability variations slightly alter the mutual coupling between the coils, again affecting the overall inductance of the pickup.

There is also the difficulty of exactly reproducing the mechanical set up each time.

[antigua]

To be honest, I think the only think anyone really needs to know about any given pickup, aside from the magnet type, DC resistance, and other things which are often disclosed, is the inductance. And the nice thing about that is with the DER DE-5000 you can get really good inductance values for any pickup when set to the 100Hz Rs setting, and it has a built in DCR measurement also.

At 100kHz Cp , the DE-5000 will give a rather reliable capacitance measurement also, but more often than not, much of that capacitance will actually come from the hookup cable instead of the pickup itself, and the capacitance is proportionately small compared to the inductance, so it's just not important anyway. Also you will find that pickups of a given coil geometry all tend to show a similar range of capacitance values, and so you don't learn a whole lot about a pickup that you couldn't have just guessed by looking at it. 

The bode plots, and the resonant peak you derive are mostly a reflection of the L and C, where the L a large factor and the C is a smaller factor. The only other thing it reveals is eddy current related Q factor losses, and similar to capacitance, you will tend to see the same patterns for a given class of pickup, because they will all have similar metal parts in their construction. Fender pickups with AlNiCo 5 poles will give a higher Q than ones with lower grades, but once you've measured one, it's like you've measured them all. Where it's really useful is when you have a metal cover over a pickup, and you want to know if the manufacturer used brass or nickel silver. The use of brass will show an obvious low Q factor, you often won't see a resonant peak at all. Most premium pickup makers will use nickel silver, but sometimes you see something odd, like brass covers on high cost Telecaster neck pickups, or pickups from China that make or may not have used nickel silver.

Another place where bode plots are really useful is when documenting active pickups like EMGs, because EMG's have circuitry in them that give them an unpredictable frequency response, and because they are almost always sealed with epoxy, a bode plot is the closes you will get to intuiting what sort of EQ circuitry they've buried in the goo, and it does have a very real impact on the sound you will get from that active pickup. In fact I just got some low cost active pickups on eBay, and I was really disappointed with what I saw in the bode plots, a roll off up to 200Hz and past 2kHz, with a 6dB/oct roll up and roll off, and I could glean from that, that I'm not going to even waste my time putting them into a guitar, and so I just ordered an other set that will hopefully have a better response profile. EMG actually provides bode plots of their pickups, but they're buried in the product literature for each pickup, and so not easy to compare. 

But given that of all the metrics, L is by far the most dominant value, if one knows the inductance of a pickup, the Q factor and C and the resonant peak can be inferred by just known what the type of pickup is, and from that you can make and educated guess how it will sound in a guitar.

It's so easy to get the L value with the DE-5000 that I've thought about going to a used guitar shop with the DE-5000 and asking if I could quickly gather the inductances of whatever pickups they had on hand, just to build up the database of values, but the nearest used guitar shop that has loose pickups is far from where I live, so I haven't made that trip yet.

[gpdb]

Feb 9, 2022 4:21:10 GMT -5 aquin43 said:

A probable explanation, which I think has been mentioned here before:

The various magnets have different magnetising force and different permeabilities. The magnetising variations alter the inductances of the coils by changing the working point of the pole pieces. The permeability variations slightly alter the mutual coupling between the coils, again affecting the overall inductance of the pickup.

There is also the difficulty of exactly reproducing the mechanical set up each time.

Thanks for the explanation aquin43. I didn't use a consistent setup for this analysis other than just eyeballing putting the frequency generator in the same place each time, but I'm working to create a 3d printed jig that will create that consistent placement.

[gpdb]

Feb 9, 2022 20:45:34 GMT -5 antigua said:

To be honest, I think the only think anyone really needs to know about any given pickup, aside from the magnet type, DC resistance, and other things which are often disclosed, is the inductance. And the nice thing about that is with the DER DE-5000 you can get really good inductance values for any pickup when set to the 100Hz Rs setting, and it has a built in DCR measurement also.

At 100kHz Cp , the DE-5000 will give a rather reliable capacitance measurement also, but more often than not, much of that capacitance will actually come from the hookup cable instead of the pickup itself, and the capacitance is proportionately small compared to the inductance, so it's just not important anyway. Also you will find that pickups of a given coil geometry all tend to show a similar range of capacitance values, and so you don't learn a whole lot about a pickup that you couldn't have just guessed by looking at it. 

The bode plots, and the resonant peak you derive are mostly a reflection of the L and C, where the L a large factor and the C is a smaller factor. The only other thing it reveals is eddy current related Q factor losses, and similar to capacitance, you will tend to see the same patterns for a given class of pickup, because they will all have similar metal parts in their construction. Fender pickups with AlNiCo 5 poles will give a higher Q than ones with lower grades, but once you've measured one, it's like you've measured them all. Where it's really useful is when you have a metal cover over a pickup, and you want to know if the manufacturer used brass or nickel silver. The use of brass will show an obvious low Q factor, you often won't see a resonant peak at all. Most premium pickup makers will use nickel silver, but sometimes you see something odd, like brass covers on high cost Telecaster neck pickups, or pickups from China that make or may not have used nickel silver.

Another place where bode plots are really useful is when documenting active pickups like EMGs, because EMG's have circuitry in them that give them an unpredictable frequency response, and because they are almost always sealed with epoxy, a bode plot is the closes you will get to intuiting what sort of EQ circuitry they've buried in the goo, and it does have a very real impact on the sound you will get from that active pickup. In fact I just got some low cost active pickups on eBay, and I was really disappointed with what I saw in the bode plots, a roll off up to 200Hz and past 2kHz, with a 6dB/oct roll up and roll off, and I could glean from that, that I'm not going to even waste my time putting them into a guitar, and so I just ordered an other set that will hopefully have a better response profile. EMG actually provides bode plots of their pickups, but they're buried in the product literature for each pickup, and so not easy to compare. 

But given that of all the metrics, L is by far the most dominant value, if one knows the inductance of a pickup, the Q factor and C and the resonant peak can be inferred by just known what the type of pickup is, and from that you can make and educated guess how it will sound in a guitar.

It's so easy to get the L value with the DE-5000 that I've thought about going to a used guitar shop with the DE-5000 and asking if I could quickly gather the inductances of whatever pickups they had on hand, just to build up the database of values, but the nearest used guitar shop that has loose pickups is far from where I live, so I haven't made that trip yet.

I do agree inductance is probably the most useful - and unfortunately the one that isn't reported by pickup makers all that often. I do have inductance values in my database for the pickups that I have found it for, but it's not available for all of them. I'm currently using a PEAK LCR45 for measurements, it only goes down to 1khz for inductance. I'm happy to purchase that meter if it makes that big of a difference, though that will be my 3rd meter! 

You hit the nail on the head for why I'm really interested in bode plots - comparing active pickups to passives is something that hasn't been documented anywhere and is a complete unknown to most people. That, and also people are very visual. I believe a bode plot will be a great comparison tool to explain how different specs affect the overall sound of a pickup.

I was hoping you'd respond to this post because I have a question for you - when you test pickups, do you change the height of the frequency generator when you're measuring a pickup with a cover vs. without? I was just doing some testing by putting non-ferrous material in between the generator and the pickup, and the height of the generator changes the chart. Since a cover would do the same, do you account for this, and if so, what is the thickness of a cover? This would be super helpful for me as I'm working on the 3d printed jigs right now. Also, do you make a similar change for single coils with pole pieces of different heights, or is it more about how close the generator is to the coil than the pole pieces?

[antigua]

Feb 13, 2022 15:25:20 GMT -5 gpdb said:

Feb 9, 2022 20:45:34 GMT -5 antigua said:

To be honest, I think the only think anyone really needs to know about any given pickup, aside from the magnet type, DC resistance, and other things which are often disclosed, is the inductance. And the nice thing about that is with the DER DE-5000 you can get really good inductance values for any pickup when set to the 100Hz Rs setting, and it has a built in DCR measurement also.

At 100kHz Cp , the DE-5000 will give a rather reliable capacitance measurement also, but more often than not, much of that capacitance will actually come from the hookup cable instead of the pickup itself, and the capacitance is proportionately small compared to the inductance, so it's just not important anyway. Also you will find that pickups of a given coil geometry all tend to show a similar range of capacitance values, and so you don't learn a whole lot about a pickup that you couldn't have just guessed by looking at it. 

The bode plots, and the resonant peak you derive are mostly a reflection of the L and C, where the L a large factor and the C is a smaller factor. The only other thing it reveals is eddy current related Q factor losses, and similar to capacitance, you will tend to see the same patterns for a given class of pickup, because they will all have similar metal parts in their construction. Fender pickups with AlNiCo 5 poles will give a higher Q than ones with lower grades, but once you've measured one, it's like you've measured them all. Where it's really useful is when you have a metal cover over a pickup, and you want to know if the manufacturer used brass or nickel silver. The use of brass will show an obvious low Q factor, you often won't see a resonant peak at all. Most premium pickup makers will use nickel silver, but sometimes you see something odd, like brass covers on high cost Telecaster neck pickups, or pickups from China that make or may not have used nickel silver.

Another place where bode plots are really useful is when documenting active pickups like EMGs, because EMG's have circuitry in them that give them an unpredictable frequency response, and because they are almost always sealed with epoxy, a bode plot is the closes you will get to intuiting what sort of EQ circuitry they've buried in the goo, and it does have a very real impact on the sound you will get from that active pickup. In fact I just got some low cost active pickups on eBay, and I was really disappointed with what I saw in the bode plots, a roll off up to 200Hz and past 2kHz, with a 6dB/oct roll up and roll off, and I could glean from that, that I'm not going to even waste my time putting them into a guitar, and so I just ordered an other set that will hopefully have a better response profile. EMG actually provides bode plots of their pickups, but they're buried in the product literature for each pickup, and so not easy to compare. 

But given that of all the metrics, L is by far the most dominant value, if one knows the inductance of a pickup, the Q factor and C and the resonant peak can be inferred by just known what the type of pickup is, and from that you can make and educated guess how it will sound in a guitar.

It's so easy to get the L value with the DE-5000 that I've thought about going to a used guitar shop with the DE-5000 and asking if I could quickly gather the inductances of whatever pickups they had on hand, just to build up the database of values, but the nearest used guitar shop that has loose pickups is far from where I live, so I haven't made that trip yet.

I do agree inductance is probably the most useful - and unfortunately the one that isn't reported by pickup makers all that often. I do have inductance values in my database for the pickups that I have found it for, but it's not available for all of them. I'm currently using a PEAK LCR45 for measurements, it only goes down to 1khz for inductance. I'm happy to purchase that meter if it makes that big of a difference, though that will be my 3rd meter! 

You hit the nail on the head for why I'm really interested in bode plots - comparing active pickups to passives is something that hasn't been documented anywhere and is a complete unknown to most people. That, and also people are very visual. I believe a bode plot will be a great comparison tool to explain how different specs affect the overall sound of a pickup.

I was hoping you'd respond to this post because I have a question for you - when you test pickups, do you change the height of the frequency generator when you're measuring a pickup with a cover vs. without? I was just doing some testing by putting non-ferrous material in between the generator and the pickup, and the height of the generator changes the chart. Since a cover would do the same, do you account for this, and if so, what is the thickness of a cover? This would be super helpful for me as I'm working on the 3d printed jigs right now. Also, do you make a similar change for single coils with pole pieces of different heights, or is it more about how close the generator is to the coil than the pole pieces?

If I select 1kHz on my LCR meters, the inductance for humbuckers and other steel poled pickups tends to read much lower than expected, so for that reason I don't think 1kHz adequate. The 1kHz test frequency is high enough to induce eddy currents, and those eddy currents result in the inaccurate reading. If you get the DE-5000, that should be the last meter you ever need. I still buy new ones for the fun of it, but I actually use the DE-5000 on the regular. Another forum recommended the Extech 380193 about a decade ago, and still to this day people find that recommendation and buy one, which is sad because they're rather outdated and overpriced now. 

It would be great if you could bode plot strange pickups, but for Fender and Gibson style pickups, it would probably just waste your time. If takes me about a half to measure a single pickup, for the time of setting up the rig, getting software configured, waiting for it to do it's thing, then measuring the results, adding text labels and uploading the image. It's just so much more work than quickly getting the inductance value of a pickup, and in most cases doesn't really tell me anything I didn't already know. 

> I was hoping you'd respond to this post because I have a question for you - when you test pickups, do you change the height of the frequency generator when you're measuring a pickup with a cover vs. without?

I adjust the output voltage so that the measurements will have a common baseline, because the actual voltage in real life is dependent on how far the pickup is from the guitar strings, and so if the purpose of the plot is to compare EQ, then it's more idea that the amplitude be normalized. For example, the cover will make it so the probe doesn't get as close to the pickup, it will cause a drop in dBV, but thats more of a testing error than a real difference. But suppose the cover did cause a legit drop in voltage, you could raise the pickup by 1mm and gain the voltage back, easy enough, but the real question at hand is how the cover impacted the EQ curve, since that's not so easily compensated for.

The exception is if I measure bridge and neck pickups, then I normalized the pickup to itself, but I don't normalize the pickups with respect to each other, because then the difference in amplitude reveals how much greater the dBV is for the "overwound" bridge pickup. That measurement is accurate because all of the increase can be attributed to the additional turns of wire on the bridge pickup. Most of the time, if the neck pickup is 7k and the bridge pickup is 8k, the difference in wire / inductance causes a barely noticeable difference, in the area of 1dBV. Bug if you have a set like the JB / Jazz, where the bridge pickup has about twice the inductance of the neck pickup, then you see a more substantial difference of 3dBV to 4dBV, and that can be interesting to observe. (it goes to show that for a "hot" set like the JB/Jazz, increased voltage is important, but for other vintage style sets, the overwound bridge is probably more about rolling off the treble response). Another upside is that the same bode plot can have both the neck and bridge pickups, and it's easier to tell which is which since they've vertically offset from one another. 

[roadtonever]

Feb 13, 2022 16:25:05 GMT -5 antigua said:

for that reason I don't think 1kHz adequate

Isn't part of the usefulness of inductance readings that you can derive a pickups resonant frequency from it?
It would be interesting to see the results of reading inductance at varying frequencies on a bunch of pickups and see which is closer between predicted vs measured resonant freq.

[antigua]

Feb 13, 2022 17:25:11 GMT -5 roadtonever said:

Feb 13, 2022 16:25:05 GMT -5 antigua said:

for that reason I don't think 1kHz adequate

Isn't part of the usefulness of inductance readings that you can derive a pickups resonant frequency from it?
It would be interesting to see the results of reading inductance at varying frequencies on a bunch of pickups and see which is closer between predicted vs measured resonant freq.

The resonant frequency moves around depending on the capacitance of the guitar cable, and the pickup lead cable. With Strat pickups, sometimes people twist the pickup and other wires together, to make it look neater, that actually increases the capacitance, too. But the inductance is essentially permanent, so from the inductance you can calculate the resonant peak with a situationally dependent capacitance. All pickups have some coil capacitance also, but the nice thing about this is that for a give type of pickup, the capacitance is fairly easy to predict, since it's so heavily dependent upon the coil's height versus it's width, and wiring configuration. PAF style humbuckers have very low capacitance due to the series wiring, usually around 50pF, up to 100pF for the high inductance models, less than 50pF for Jazz Master pickups because the coils are so flat and wide, 100pF for Strat pickups because they're so thin and tall, and 200pF for Tele neck pickups because they're even more thin, and tall.  So if I know the inductance, almost all of the mystery about a pickup of a known construction, is gone. 

For the inductance measurement of pickups, you want the lowest frequency possible, because eddy currents increase with frequency, and eddy currents are not an intended part of the measurement. It's like weighing someone while they're wearing snow gear. I'm not sure what sort of applications call for measuring inductance at higher frequencies, but I would guess non-audio applications and inductors of lower value and likely much higher Q factors. A guitar pickup is not really and inductor, it's a transducer, so it can't really be treated like a regular inductor, and the best way to deal with the non ideal nature of a guitar pickup is to test at the lowest frequency the LCR meter will allow. Also, if you calculate the inductance of a pickup bases on resonant peak and a large known value of C, the L you get will most closely match the L given by the LCR meter at 100Hz as  opposed to 1kHz. 

[gpdb]

Feb 13, 2022 16:25:05 GMT -5 antigua said:

Feb 13, 2022 15:25:20 GMT -5 gpdb said:

I do agree inductance is probably the most useful - and unfortunately the one that isn't reported by pickup makers all that often. I do have inductance values in my database for the pickups that I have found it for, but it's not available for all of them. I'm currently using a PEAK LCR45 for measurements, it only goes down to 1khz for inductance. I'm happy to purchase that meter if it makes that big of a difference, though that will be my 3rd meter! 

You hit the nail on the head for why I'm really interested in bode plots - comparing active pickups to passives is something that hasn't been documented anywhere and is a complete unknown to most people. That, and also people are very visual. I believe a bode plot will be a great comparison tool to explain how different specs affect the overall sound of a pickup.

I was hoping you'd respond to this post because I have a question for you - when you test pickups, do you change the height of the frequency generator when you're measuring a pickup with a cover vs. without? I was just doing some testing by putting non-ferrous material in between the generator and the pickup, and the height of the generator changes the chart. Since a cover would do the same, do you account for this, and if so, what is the thickness of a cover? This would be super helpful for me as I'm working on the 3d printed jigs right now. Also, do you make a similar change for single coils with pole pieces of different heights, or is it more about how close the generator is to the coil than the pole pieces?

If I select 1kHz on my LCR meters, the inductance for humbuckers and other steel poled pickups tends to read much lower than expected, so for that reason I don't think 1kHz adequate. The 1kHz test frequency is high enough to induce eddy currents, and those eddy currents result in the inaccurate reading. If you get the DE-5000, that should be the last meter you ever need. I still buy new ones for the fun of it, but I actually use the DE-5000 on the regular. Another forum recommended the Extech 380193 about a decade ago, and still to this day people find that recommendation and buy one, which is sad because they're rather outdated and overpriced now. 

It would be great if you could bode plot strange pickups, but for Fender and Gibson style pickups, it would probably just waste your time. If takes me about a half to measure a single pickup, for the time of setting up the rig, getting software configured, waiting for it to do it's thing, then measuring the results, adding text labels and uploading the image. It's just so much more work than quickly getting the inductance value of a pickup, and in most cases doesn't really tell me anything I didn't already know. 

> I was hoping you'd respond to this post because I have a question for you - when you test pickups, do you change the height of the frequency generator when you're measuring a pickup with a cover vs. without?

I adjust the output voltage so that the measurements will have a common baseline, because the actual voltage in real life is dependent on how far the pickup is from the guitar strings, and so if the purpose of the plot is to compare EQ, then it's more idea that the amplitude be normalized. For example, the cover will make it so the probe doesn't get as close to the pickup, it will cause a drop in dBV, but thats more of a testing error than a real difference. But suppose the cover did cause a legit drop in voltage, you could raise the pickup by 1mm and gain the voltage back, easy enough, but the real question at hand is how the cover impacted the EQ curve, since that's not so easily compensated for.

The exception is if I measure bridge and neck pickups, then I normalized the pickup to itself, but I don't normalize the pickups with respect to each other, because then the difference in amplitude reveals how much greater the dBV is for the "overwound" bridge pickup. That measurement is accurate because all of the increase can be attributed to the additional turns of wire on the bridge pickup. Most of the time, if the neck pickup is 7k and the bridge pickup is 8k, the difference in wire / inductance causes a barely noticeable difference, in the area of 1dBV. Bug if you have a set like the JB / Jazz, where the bridge pickup has about twice the inductance of the neck pickup, then you see a more substantial difference of 3dBV to 4dBV, and that can be interesting to observe. (it goes to show that for a "hot" set like the JB/Jazz, increased voltage is important, but for other vintage style sets, the overwound bridge is probably more about rolling off the treble response). Another upside is that the same bode plot can have both the neck and bridge pickups, and it's easier to tell which is which since they've vertically offset from one another. 

Guess I need to order a new meter. I think you're right because my inductance readings have always been lower than I would expect, I was comparing my readings for a SD Jazz to yours on your website, and I was only getting 3.73 when yours was in the mid-4's. 

As far as the bode plots go, I understand where you're coming from about their usefulness, but at the same time as a tool to convince someone how a pickup will sound, teaching someone what inductance means and how that sound relates to the number is a bit harder than showing a chart and a sound clip. Regardless, I plan to provide all of them. 

How do you know what the voltage needs to be adjusted to though? Are you just normalizing between two pickups you test, or is it normalized for all pickups you've ever tested? 

For inductance, if it can tell you what the potential resonant frequency is, does it also tell you something about the Q factor? 

[antigua]

Feb 14, 2022 10:16:35 GMT -5 gpdb said:

Feb 13, 2022 16:25:05 GMT -5 antigua said:

Guess I need to order a new meter. I think you're right because my inductance readings have always been lower than I would expect, I was comparing my readings for a SD Jazz to yours on your website, and I was only getting 3.73 when yours was in the mid-4's. 

As far as the bode plots go, I understand where you're coming from about their usefulness, but at the same time as a tool to convince someone how a pickup will sound, teaching someone what inductance means and how that sound relates to the number is a bit harder than showing a chart and a sound clip. Regardless, I plan to provide all of them. 

How do you know what the voltage needs to be adjusted to though? Are you just normalizing between two pickups you test, or is it normalized for all pickups you've ever tested? 

For inductance, if it can tell you what the potential resonant frequency is, does it also tell you something about the Q factor? 

If you make the bode plots, that's great because there is secondary observations that can be made, like "this company claims the pole pieces are AlNiCo 5 but it has a Q factor that's more common of AlNiCo 2 or 3", so if you have the time and patience for it, it's a nice data point to have for unknown future use. I'd found certain pickups exhibited strange plots beyond 30kHz, and I generally never measured past 30kHz since this is audio application, but the 100kHz plots gave insight into whether a pickup had been hand guided, or machine wound, so you never know how the data might come in handy later. Pickups that are hand guided show additional high frequency resonance as a result of inconsistent spooling, leading to smaller L/C resonances within the coil, that appear well above the resonant peak of the coil as a whole.

To normalize the plots, the default voltage is 4.00 Vpp with the Velleman PCU 200, and I usually just have to bump it up 4.05 or 4.1 Vpp to get the loaded and unloaded plots to match, but I'll set it to whatever it takes to make their baseline equal.  Sometimes I have to start at a higher or lower voltage if the plot is going out of the selected measurement range. If you can keep it at 4Vpp for all pickups, then in theory you can compare the relative outputs of different pickups, as I do with neck and bridge pickups, where everything is the same except the number of turns of wire on the bobbins. But if you have a pickup with a cover, and one without, or one with flat pole pieces and with one where they stick out, then consistent placement of the test coil becomes all but impossible, so the results are not strictly comparable. And later on, if you ever switch to a new test probe, it will probably produce a slightly different field intensity, and so the measurements you made with the old test coil will not compare one-to-one to the tests with the new test coil. I've been measuring pickups for about five years have have gone through about five test coils and three USB bode plotters, so the chance of having a uniform test setup has not been an option for me.

The 200k ohm test load causes the Q factor to drop, and in combination with the eddy currents, the overall frequency-dependent load is interesting, and does things you wouldn't expect, most especially with the Gretsch Filter'tron, which is unique due to the large metal screws and small coils. We were able to model ideal pickups easily and quickly with LTSpice, like Fenter single coils, but we spent about a year trying to figure out how to model a Filter'tron, and then someone came along and make and LTSpice model that modeled it as a three way transformer, and the curves matched perfectly. But again, you see the same pattern with nearly all Filter'trons, and the only place where it become interesting was comparing TV Jones to Gretsch / Fender BHK, because they used different size screws, and that altered the impedance curves in a way that was not something I could have guessed. 

Here's an example, I had normalized the bridge and neck pickup's loaded and unloaded amplitudes at the left side, but diverge as they near resonance. Not normalizing the two pickups together reveals a measly 1dBV output difference, I'd say it's more useful just that they pickup's own plot lines are set apart and not overlapping.  Notice that the combination of 200k ohm test load and eddy currents cause a strange Nike-logo looking plot line. The Q factor drops with the test load, but the absolute amplitude at resonance varies also. The overall trend is that the test load brings the peak amplitude closer to the baseline.  




If the only difference is the inductance, the resonant peak moves, but he overall curve stays the same. Technically the inductance alters the Q factor, because the ratio L/C is part of the equation of Q: 



But when comparing similar pickups, L is generally very large and C is very small, proportionately, so you don't see much variety in Q factors on account of the L/C ratio.