moving the curve

[Gyros]

Check the attached picture please. There are 7 jazz bass bridge pickups in the plots. 6 are much the same, but 1 has the peak at a lower frequency. This one is a fender pickup, the others are my attempt to come close. Although I do not necessarily want to reproduce that fender pickup, I would like to understand what to do to tweak my pickup to get the same response. Just for the learning experience. I have kept all factors similar to the fender: coil height, magnet dimensions, wire type, amount of turns (checked dc resistance to get near). What I have been playing with is wire tension and tpl. Does anyone have a suggestion for what I can do to move my plots to the left, so to say?

[Helmholtz]

The slightly lower resonant frequency means a higher inductance or capacitance.

DC resistance is not an accurate indicator of turns number, as wire diameter can vary and a looser wind requires more wire, giving a higher DCR for same number of turns.

Inductance depends mainly on turns number and magnet type/size.

PU capacitance is influenced by insulation type and thickness as well as how the PU is wound.

To get closer I would increase the turns number by maybe 5%.

But actually I think the difference is negligible and originals might vary by a similar amount.
When loaded by controls and cable such minor differences often vanish.

[Mike Sulzer]

You need to measure inductance. The best thing to do is buy a meter, but since you already can measure resonant frequency, you could use a technique where you put capacitances across the pickup and make multiple measurements of the resonances and solve an equation. You need some experience to get that to work.

Put a turns counter on your winder, and find out how many turns you should need.

[Chuck H]

We don't know for certain what you have done to try to "come close".

If you have matched wire gauge turns and poles then you should have the right inductance. My guess is that something about the Fender pickup has a higher capacitance. As Helmholtz said, this could have to do with insulation thickness or winding specifics. Another guess... Being a "commercial product" the Fender pickup may have thinner wire insulation or may be wound more professionally on a better machine (ie: tighter with less scatter in the winding) or both.

You could always just add a small value cap across the winding on any pickup you have made to very effectively simulate this.

EDIT: Also in line with Helmholtz considerations... The difference is negligible. Especially when you consider the affect that the rest of the overall guitar circuit will have. Which includes the volume and tone controls and even the instrument cable. The difference in the location of the resonant peak will be MUCH greater switching between a 10' and 20' cord to the amp than the difference shown in your graph between the pickups you've made and the Fender pickup.

[bbsailor]

This topic reflects on the concept of the “typical electric guitar sound”. This comes from most single coil pickups having the resonant peak point near 3KHz and changing slightly with the length of the coax adding about 30pf of capacitance per foot. Passive pickups needed enough turns to generate enough voltage to drive two amplifier characteristics.
1. amplifier input impedance and loading by on-instrument volume and tone controls.
2. amplifier voltage sensitivity due to preamp design.

In order to generate this voltage, pickups needed to pack enough wire turns close enough to the primary string magnetic field to produce the required output level. This is what made AWG 42 and 43 the most common wire sizes.

With single coil pickups typically using a 250K ohm volume pot you will find that typical pickup resistance is about
40 times lower than the pot value, which I call the “rule of 40” to easily estimate the number of needed turns. Winding style, tight or loose, wire insulation type and thickness will each add another variable to keep in mind when trying to target a particular design, output level and sound.

Another variable is the chosen pickup magnet type and size along with whether you use a metal pickup cover which may alter the sound due to eddy currents being generated in the metal cover.

Joseph J. Rogowski

[Mike Sulzer]

Yes, the cable capacitance is the dominant capacitance, and so it tends to "buffer" changes in pickup capacitance. That is, a difference you see in a measurement without the cable is less significant when using the cable because the difference is a smaller fractional part of the total.

If the difference is the inductance, well, there is nothing buffering that, and so it would matter more. But Gyros wants to understand the reason, not necessarily duplicate the Fender. I think that requires some measurements.

[Helmholtz]

With single coil pickups typically using a 250K ohm volume pot you will find that typical pickup resistance is about
40 times lower than the pot value, which I call the “rule of 40” to easily estimate the number of needed turns. Winding style, tight or loose, wire insulation type and thickness will each add another variable to keep in mind when trying to target a particular design, output level and sound.

Regarding PU loading the DCR is rather insignificant.
What matters is the impedance.
A typical single coil PU (2.5H) has a (mostly inductive) impedance of around 50k @ 3kHz, i.e. without resonance.
At resonance the PU impedance can approach 1M.

[bbsailor]

Regarding PU loading the DCR is rather insignificant.
What matters is the impedance.
A typical single coil PU (2.5H) has a (mostly inductive) impedance of around 50k @ 3kHz, i.e. without resonance.
At resonance the PU impedance can approach 1M.

Yes, the pickup alone has what you stated. However, add a 10 ft coax and you add about 300pf of capacitance in parallel which has a 176,838.8 ohm reactance at 3KHz. This is the load that is in parallel with the volume pot 250K ohm value and additional tone pot loading. This is why some might want to try making an active FET buffer in the jack end of the guitar cable with the power section being at the amp end either using a battery or plug in power supply. This FET has about a 5 to 10 meg ohm pickup load which isolates the pickup from the cable capacitance since the FET output impedance is much lower going into the amp input jack.

The wire insulation type and thickness affects the coil turn to turn capacitance along with how much winding scatter is used to attempt to reduce this capacitance. All of this is done to find what sounds better to the ear, particularly when playing live in a group and have a better way to blend than just turn up the volume on the amp.

A raw pickup that is 2.5H will have a reactance of 47,100 ohms at 3KHz measured directly at the pickup output cable or wire. Then, the pot loading and coax length and loading will add additional tonal variations with the resonance peak Q dropping as well as the resonant frequency dropping.

The best measurements to make are done at the pickup output direct, at the guitar output jack and finally at the end of the guitar cable to see how those measurements change with pot loading and coax cable loading.

Pickup winding techniques will affect the direct pickup output sound first then the pot loads and coax cable loads add and may mask some more subtle winding variations.

Joseph J. Rogowski

[Gyros]

First off, thanks everyone for offering your thoughts. Let me clarify with some more details.

The plot is one with simulated cable capacitance. I used the Integrator of Ken Willmott and Rightmark software. The integrator has a load/noload/impedance switch. This plot was with the switch on load. I picked that one, because that would probably best match with what I heard.

I did measure impedance and capacitance. Here is the compared bridge pickups of the fender, and one of my own.

Fender Jazzbass bridge
I=3.57H
C=135pF
Rdc=7.87Ohm

Mine
I=3.19H
C=103pF
Rdc=7.82Ohm

The coil height is the same. It is both PE wire, mine 42awg, my guess is 42 for the fender as well. I think the amount of wire is very similar, maybe a tiny difference, but that wouldn't explain the big tonal difference. I notice that the fender coil looks fuller, like it was wound with thicker wire. But PE is always single build, as far as I know. And I don't think they used 41 wire. So there may be something going on with insulation thickness. Maybe today's PE is different than in the past. The magnet spacing is a tiny bit different. This makes the Fender coil just a slight bit larger. But I don't think that that causes so much of a difference.

I tried different wire tensions and different tpl's. They have a bit influence on the tone, but not that much.

Looking at the numbers, I see that the induction and capacitance of the Fender are considerably higher. I think it may have to do with the magnets. I assume it is A5, but how can I tell? Could it be A2?
Or, maybe the specs of the A5 are different. I have seen many different values for BHmax, Hcb, Hcj and Br of A5, depending on the supplier. Has anyone ever tested these differences?

[Mike Sulzer]

" I think it may have to do with the magnets."

Yes, I think differences in the magnet permeability are the most likely cause.

[Helmholtz]

The plot is one with simulated cable capacitance. I used the Integrator of Ken Willmott and Rightmark software. The integrator has a load/noload/impedance switch. This plot was with the switch on load. I picked that one, because that would probably best match with what I heard.

Fender Jazzbass bridge
I=3.57H
C=135pF
Rdc=7.87Ohm

Mine
I=3.19H
C=103pF
Rdc=7.82Ohm

It seems the integrator setup loads the PU with a capacitance of around 500pF to simulate the effect of the guitar cable.
So total capacitance is around 600pF.
Difference between PUs is just 32pFor 5% of total capacitance.
The 32pF difference alone will make for a resonant frequency shift of 2.6% or 85Hz.

Now the difference between inductances is 0.38H or 11%.
This causes another fres shift of 6% or 200Hz.
So the influence of inductance is much stronger.

I understand you don't count turns but wind to same DCR.
As said above that might result in a different number of turns and consequently different inductance.
You should wind to same inductance instead.
That will require around 6% more turns and will increase the DCR by a few hundred Ohms.
The difference in DCR is inaudible.

[Gyros]

You should wind to same inductance instead.
That will require around 6% more turns and will increase the DCR by a few hundred Ohms.
The difference in DCR is inaudible.

I have done that already, and actually it is included as one of the 6 plots of my own designs . It didn't result in moving the curve to the left. I am more and more inclined to suspect the magnets. I will buy some different magnets and do some more tests. Thnx!

[Mike Sulzer]

I have done that already, and actually it is included as one of the 6 plots of my own designs . It didn't result in moving the curve to the left.

Then there is something wrong with your measurements. It appears that the effect of the capacitance on the differences is small, and so it must be the inductance. Increasing the inductance should lower the resonant frequency. (This is assuming AlNiCo magnets for pole pieces. With steel pole pieces, the eddy currents can result in a lower "effective inductance" at the resonance than the inductance value measured at low frequencies.)

[Helmholtz]

What is the inductance of your "overwound" sample?

[Gyros]

The overwound one is 3.32H, and it has about 4% more windings. The difference is actually quite audible. It has more low midrange and the attack gets a little bit more direct. This is of course what always happens with more windings, but even a little bit like 4% is already quite noticeable. If I were to keep adding windings until I get to 3.57H, then I will have to go up to 10% more windings. Then the DCR will have increased so much that it is really no longer close to the Fender. So this leads me to think that it's not windings. It must be a magnet issue, I think.

I have ordered some A5 magnets with higher values for Br, H and BHmax. That will be an interesting comparison.