Another Run At Medium Z (around 1 Henry) Pickups and Preamps

[charrich56]

I think there's a "sweet spot" in the pickup design space that can work very well with an active preamp and active voicing control to provide versatility and relatively low cost for an onboard system.

So, here are the basic ideas, and a sample schematic for the voicing preamp, for some discussion. I'm just going to list the pieces and key assumptions and comments are welcome.

1. 500 milliHenry to about 1.25 Henry pickups of conventional design (humbucking preferably.) Nominal 1 Henry. About 1K to 2.5 K DC resistance. Self -resonance looking into a non-capacitive buffered input will be about 10 KHz or thereabouts. As a wild guess, typical wire gauge will be in the 39AWG to 40 AWG range for most bobbins.

2. This means reduced output, less than conventional pickup winds by about 12 dB.

3. 12 dB is a good gain range for a single JFET Class A preamp. It allows a single low cost preamp stage with good drive, current consumption, and dynamic range, and good signal to noise with the right JFET. We are about in the 30 to 75 mV input signal range. This signal level is also OK to drive a simple JFET buffer for the capacitor bootstrap voicing control.

4. I have covered the variable resonance control in another post. Basically using a buffer, a variable capacitor can be simulated with a pot and a fixed capacitor. This allows single pot control of the load capacitance of the pickup and therefore allows you to "dial up" the resonant frequency of the pickup over a fairly large range (1-1/2 to over 2 octaves) depending on a few other things.

Changing the Q value in the same way is possible but you need more parts, and a relatively fixed Q can be designed in (about 6-12 dB peak amplitude at resonance) in the simplest circuit to emulate a normal pickup loaded by tone/volume and some cable capacitance. So we are going to just design in a reasonable value of Q and leave it fixed in this first pass design.

The voicing may be turned on and off or switched between different resistors and capacitors to give various ranges and/or resonant "presets" to the musician to control his/her tone onboard the instrument.

5. With the "sweet spot" termination resistance loading the PU , when the voicing is switched out, instead of a voiced or midrange-resonant pickup, there is just a 3-6 dB boost at 10 KHz, it's pretty flat below that, and we have a very nice full range sound.

My experimental assumption is that it might be musically unnecessary to provide any output from the pickup over 10 KHz, even for an "acoustic-ish" sounding hi-fi output. Most of what people think of as highs, are roughly around 5K- 7.5K. If you are using a "normal" guitar amp, there's no or very little speaker output above 5-7 KHz.

A small boost at 10 KHz might be OK but the voicing can be additionally switched (not shown on the schematic, but pointed to and hinted at) to provide just a variable lowpass rolloff if needed.

6. So if all this works out, we don't have to go to the trouble of low impedance and extremely low output to get the versatility of high fidelity output (for our purposes,) up to or a bit over 10KHz, combined with the ability to voice the pickup to be "close" to the sound of almost any pickup of the same basic construction and magnetic circuit.

7. Funny enough, if you wind a pickup this way, I think you end up pretty close to the old Alembic pickup formula.

I'm posting the schematic just as a reference point for discussion and sharing the idea. (There are surely lots of different ways to do this, but this is the KISS approach , has the right set of tradeoffs for onboard use, and some proven ideas are used here (e.g. JFET preamps are euphonic on guitar). Don't let the simplicity fool you into thinking this design won't sound good. ) This circuit can swing over 3 Volts P-P so there is really quite a bit of dynamic headroom. I have this in CircuitLab and may publish the public URL at some point.

Please comment on the thread if you are interested! Thanks, -Charlie

[bbsailor]

I think there's a "sweet spot" in the pickup design space that can work very well with an active preamp and active voicing control to provide versatility and relatively low cost for an onboard system.

So, here are the basic ideas, and a sample schematic for the voicing preamp, for some discussion. I'm just going to list the pieces and key assumptions and comments are welcome.

1. 500 milliHenry to about 1.25 Henry pickups of conventional design (humbucking preferably.) Nominal 1 Henry. About 1K to 2.5 K DC resistance. Self -resonance looking into a non-capacitive buffered input will be about 10 KHz or thereabouts. As a wild guess, typical wire gauge will be in the 39AWG to 40 AWG range for most bobbins.

2. This means reduced output, less than conventional pickup winds by about 12 dB.

3. 12 dB is a good gain range for a single JFET Class A preamp. It allows a single low cost preamp stage with good drive, current consumption, and dynamic range, and good signal to noise with the right JFET. We are about in the 30 to 75 mV input signal range. This signal level is also OK to drive a simple JFET buffer for the capacitor bootstrap voicing control.

4. I have covered the variable resonance control in another post. Basically using a buffer, a variable capacitor can be simulated with a pot and a fixed capacitor. This allows single pot control of the load capacitance of the pickup and therefore allows you to "dial up" the resonant frequency of the pickup over a fairly large range (1-1/2 to over 2 octaves) depending on a few other things.

Changing the Q value in the same way is possible but you need more parts, and a relatively fixed Q can be designed in (about 6-12 dB peak amplitude at resonance) in the simplest circuit to emulate a normal pickup loaded by tone/volume and some cable capacitance. So we are going to just design in a reasonable value of Q and leave it fixed in this first pass design.

The voicing may be turned on and off or switched between different resistors and capacitors to give various ranges and/or resonant "presets" to the musician to control his/her tone onboard the instrument.

5. With the "sweet spot" termination resistance loading the PU , when the voicing is switched out, instead of a voiced or midrange-resonant pickup, there is just a 3-6 dB boost at 10 KHz, it's pretty flat below that, and we have a very nice full range sound.

My experimental assumption is that it might be musically unnecessary to provide any output from the pickup over 10 KHz, even for an "acoustic-ish" sounding hi-fi output. Most of what people think of as highs, are roughly around 5K- 7.5K. If you are using a "normal" guitar amp, there's no or very little speaker output above 5-7 KHz.

A small boost at 10 KHz might be OK but the voicing can be additionally switched (not shown on the schematic, but pointed to and hinted at) to provide just a variable lowpass rolloff if needed.

6. So if all this works out, we don't have to go to the trouble of low impedance and extremely low output to get the versatility of high fidelity output (for our purposes,) up to or a bit over 10KHz, combined with the ability to voice the pickup to be "close" to the sound of almost any pickup of the same basic construction and magnetic circuit.

7. Funny enough, if you wind a pickup this way, I think you end up pretty close to the old Alembic pickup formula.

I'm posting the schematic just as a reference point for discussion and sharing the idea. (There are surely lots of different ways to do this, but this is the KISS approach , has the right set of tradeoffs for onboard use, and some proven ideas are used here (e.g. JFET preamps are euphonic on guitar). Don't let the simplicity fool you into thinking this design won't sound good. ) This circuit can swing over 3 Volts P-P so there is really quite a bit of dynamic headroom. I have this in CircuitLab and may publish the public URL at some point.
[ATTACH=CONFIG]37099[/ATTACH]
Please comment on the thread if you are interested! Thanks, -Charlie

Charlie, and others interested in this subject

Guitar pickups have a well known history using many thousands of turns of very fine magnet wire to create enough output to drive tube based amps of the early days of the electric guitar. Add to this the capacitance of the magnet wire and the coax capacitance and you create a resonant hump in the frequency response in the 2KHz to about 4.5KHz range which creates the traditional "electric guitar sound".

One problem with this high Z pickup design is the noise pickup as single coil pickups also pickup noise from a variety of sources. Just hold and reorient your single coil guitar near a computer to hear what this means. Humbucker pickups were invented to remedy this problem but at the expense of altering the single coil sound but still having an electric guitar sound.

Your analysis needs to consider one more critical aspect of pickup design: "signal to noise ratio". As you state that your medium impedance pickup is -12dB that means that is has about one fourth the number of turns of a passive high Z pickup. Pickups have historically occupied critical space between the heel of the neck and the bridge, requiring about a .75 inch deep hole to mount the pickup with all its wire. This hole is in a critical location where the neck joins the body and the remaining string energy, after the pluck or strum, reaches the pickup. This medium Z pickup offers up the potential to make a very thin pickup, typically less than the thickness of a fingerboard, about .25 inch thick without the need to cut a hole in the guitar body to mount a high Z pickup.

I have always looked to find ways to minimize the noise pickup in a pickup and thus have more options to make wider bandwidth and flatter pickups (electronically and physically) to match commonly available inputs available to musicians. These inputs include the traditional 1M ohm high Z input of the guitar amp but also the 150 ohm to 300 ohms input impedance of a low Z XLR input that is typically rated at about 2.4K ohms to act as a bridging to a low Z microphone. By keeping the pickup noise low the extra gain of the XLR circuit can provide an nice boost but also have an extended frequency response well beyond the typical 5KHz cut off of most electric guitar amps.

The challenge is introducing a guitar design that does not require a pickup hole in line with the neck but offers enough of a physical advantage to motivate folks to attempt this new design that is not based on traditional high Z pickup cutout footprints (single coil, humbucker, P90, etc).

The traditional guitar amp has little output beyond 5KHz. If anyone wants to hear what this 5KHz limit sounds like, just listen to music on any AM radio that has their frequency allocation at 10KHz intervals but that translates into a plus and minus 5KHz maximum high frequency to not exceed the FCC frequency and bandwidth allocation.

The maximum primary frequency on a guitar is about 1Khz on the upper frets of the high e-string. Only the energy of the harmonics are left and the question that remains is: Is there anything worth listening to in this upper harmonic register?

One thing that I have found is that when there is less pickup frequency peaking in the 2KHz to 4.5KHz range, the guitar sounds a little more acoustic if the response can be extended out to near the 10KHz range. Here is where medium Z pickups may come into their own unique place in modern guitar design. Add to this the ability of medium Z pickups to be modeled to have an electric sound as well as an acoustic electric sound, this topic opens the door for considering the evolution of the guitar and weighing the value of energy preservation in the vibrating string.

I hope this stimulates some more comments?

Joseph J. Rogowski

[charrich56]

Joseph,

Thanks very much for your comments, and your posts on this forum have been an inspiration on my journey for several years now.

That's an interesting item you brought up-- that medium Z can actually mean a physically smaller (flatter) pickup which allows better energy transfer through the neck/body/top join since there isn't a large hole there. I don't think you could beat a single loop ultra low-Z pickup there for height but several people have posted about thin flat single coils and humbuckers.

I really like low-Z designs as well, and the variable voicing trick can work with them too. But it looks like this medium-Z recipe is going to work pretty well.

I think you just posted a vote for not needing over 10 KHz response even if you want a glassy, bell-like acoustic type sound with lots of high harmonics and good pick transients.

Could someone comment on whether real amplified acoustic guitars need much over 10K to sound good with piezos, or mic'ed?

[charrich56]

Here's some plot images from Circuitlab swept frequency simulation runs for the pickup and preamp. The circuit is exactly as shown in the schematic with the exception of Plot 3, where R2 is set to 20K instead of 0.

(Please note that the pickup DC resistance of about 2.3 K ohms is not visible in the schematic because it is put in as an ESR parameter in the inductor L1.)

1. Voicing (SW1) off.


2. Voicing (SW1) on, stepping the Tone pot through its range. C3 is constant; we are only varying the Tone pot.


3. Voicing (SW1) on, stepping Tone pot through its range, R2 is set to 20K instead of 0. C3 is constant as before. This produces a simple "treble" or lowpass control.


The Q of the pickup/preamp system was set for fairly low peaking at resonance.

-Charlie

[Stratz]

Charlie, a bit off topic, I hope you dont mind.

What software are you using for the scope?

Also, can anybody please explain to me exactly what the dBV scale on the left represents and how it relates to the frequency readings on the bottom. I think it's a voltage measurement of some sort but I dont know how to read it.
I apologize for my ignorance and going off topic but I've been trying to figure out what the dBV scale represents for some time now.
Thanks.

[charrich56]

Stratz,

I'm using CircuitLab, which is a Web-based (but paid subscription) electronic circuit simulator/modeler like Spice.
www.circuitlab.com

It allows you to draw all sorts of DC and AC circuits and then simulate their response to signals. It has libraries of parts where you can "plug in" the parts and values you want and see how the whole circuit will work pretty accurately before you build the circuit.

The plots are AC sweep frequency plots that can be done with Circuitlab easily where we can vary any value of resistance, capacitance, voltage, etc. of parts in the circuit , sweep the input signal frequency over a range, and see the effect.

The dbV scale is a logarithmic relationship of output voltage to input voltage of a circuit. There's another definition relating to comparing to a standard 1 volt level in audio but here, Circuitlab uses it just to compare two points in a circuit. So a straight piece of wire has 0 dB gain (20 log Vout/Vin = 0 because log 1 = 0.) 6 dB gain measured at the output is twice the AC voltage that was input, -6 dB is half the input, etc.

So the plots you see are "frequency response" plots where we can see how much gain we have (y axis) while sweeping over a range of frequencies (X axis).

[Stratz]

Thank you very much Charlie!
Quick question if you dont mind.
Say you measure two hi Z humbuckers.
One measures 6khz/-12dBV and the other measures 6khz -10 dBV. What does the difference in the dBV tell you.

This is where I get hung up. I can measure similar coils and come up with nearly the same resonant peak but the dBV is different.

Thanks again.

[charrich56]

Stratz,

If you are getting those measurements from an oscilloscope or some other piece of audio measurement gear, the dbV you get from a single pickup for example, is a voltage ratio referenced to 1 volt RMS AC. So the first pickup output voltage is antilog (-12/20) or .251 volts AC RMS. The other pickup has antilog(-10/20) or .316 volts AC RMS. The second pickup has greater output than the first.

Here's a good page that explains these:
Understanding dB

[Stratz]

That explains a lot. Thanks very much Charlie!

[charrich56]

Just a quick note for the original post:

A "vintage" wound humbucker of 4.0 Henries or less and about 8K Ohms or less will work with this technique if the pickup is wired to put the coils in parallel. The self resonant frequency won't be quite 10 KHz but still may be up there past 7-8 KHz, giving an uncolored midrange (depending on the amp tone stack of course) with voicing switched out, and lots more treble right before the speaker's cutoff, to work with through a conventional guitar amp.

A single coil-sized humbucker which is not super "hot" (think SD Vintage Rails or Duckbucker type PU, about 9K-10K DCR) should also work well with this setup when wired for parallel coils. That's one direction I'm pursuing for the current Strat-type demo (proof of concept) instrument I'm working on, until I have some time to wind my own single coil sized rails with Mojotone parts.

[Mike Sulzer]

Nice work Charlie. Medium impedance pickups are one of my favorite methods to get good flexibility. I have never tried the boot strap tuning method, preferring to just switch in several different capacitor values. I also put resistors in parallel with the larger Cs to get the sound I want. This gets a bit complicated; you have reduction in Q from the series resistance of the pickup, but also from any parallel resistance. But the two have differing effects; as the C goes up the relative effect of the parallel resistance drops, and the Q rises. It might be interesting to move some of the parallel resistance to be in parallel with the bootstrapped C. Not sure what that would do, but it might tend to give a more constant Q.

I think your pickups might be flatter than you say when you have switched out the bootstrapped C. 220pf seems high for the capacitance of a medium impedance pickup, even a bit high for a typical PAF type humbucker.

Let me know what you think of these comments. I think there is always more to be discussed and experimented with.

I think there's a "sweet spot" in the pickup design space that can work very well with an active preamp and active voicing control to provide versatility and relatively low cost for an onboard system.

So, here are the basic ideas, and a sample schematic for the voicing preamp, for some discussion. I'm just going to list the pieces and key assumptions and comments are welcome.

1. 500 milliHenry to about 1.25 Henry pickups of conventional design (humbucking preferably.) Nominal 1 Henry. About 1K to 2.5 K DC resistance. Self -resonance looking into a non-capacitive buffered input will be about 10 KHz or thereabouts. As a wild guess, typical wire gauge will be in the 39AWG to 40 AWG range for most bobbins.

2. This means reduced output, less than conventional pickup winds by about 12 dB.

3. 12 dB is a good gain range for a single JFET Class A preamp. It allows a single low cost preamp stage with good drive, current consumption, and dynamic range, and good signal to noise with the right JFET. We are about in the 30 to 75 mV input signal range. This signal level is also OK to drive a simple JFET buffer for the capacitor bootstrap voicing control.

4. I have covered the variable resonance control in another post. Basically using a buffer, a variable capacitor can be simulated with a pot and a fixed capacitor. This allows single pot control of the load capacitance of the pickup and therefore allows you to "dial up" the resonant frequency of the pickup over a fairly large range (1-1/2 to over 2 octaves) depending on a few other things.

Changing the Q value in the same way is possible but you need more parts, and a relatively fixed Q can be designed in (about 6-12 dB peak amplitude at resonance) in the simplest circuit to emulate a normal pickup loaded by tone/volume and some cable capacitance. So we are going to just design in a reasonable value of Q and leave it fixed in this first pass design.

The voicing may be turned on and off or switched between different resistors and capacitors to give various ranges and/or resonant "presets" to the musician to control his/her tone onboard the instrument.

5. With the "sweet spot" termination resistance loading the PU , when the voicing is switched out, instead of a voiced or midrange-resonant pickup, there is just a 3-6 dB boost at 10 KHz, it's pretty flat below that, and we have a very nice full range sound.

My experimental assumption is that it might be musically unnecessary to provide any output from the pickup over 10 KHz, even for an "acoustic-ish" sounding hi-fi output. Most of what people think of as highs, are roughly around 5K- 7.5K. If you are using a "normal" guitar amp, there's no or very little speaker output above 5-7 KHz.

A small boost at 10 KHz might be OK but the voicing can be additionally switched (not shown on the schematic, but pointed to and hinted at) to provide just a variable lowpass rolloff if needed.

6. So if all this works out, we don't have to go to the trouble of low impedance and extremely low output to get the versatility of high fidelity output (for our purposes,) up to or a bit over 10KHz, combined with the ability to voice the pickup to be "close" to the sound of almost any pickup of the same basic construction and magnetic circuit.

7. Funny enough, if you wind a pickup this way, I think you end up pretty close to the old Alembic pickup formula.

I'm posting the schematic just as a reference point for discussion and sharing the idea. (There are surely lots of different ways to do this, but this is the KISS approach , has the right set of tradeoffs for onboard use, and some proven ideas are used here (e.g. JFET preamps are euphonic on guitar). Don't let the simplicity fool you into thinking this design won't sound good. ) This circuit can swing over 3 Volts P-P so there is really quite a bit of dynamic headroom. I have this in CircuitLab and may publish the public URL at some point.
[ATTACH=CONFIG]37099[/ATTACH]
Please comment on the thread if you are interested! Thanks, -Charlie

[charrich56]

Mike,

Thanks for your comments and encouragement! I appreciate that you have done some work with medium Z in the past. One of the beauties of using a circuit simulator like CircuitLab is that you can tweak and experimentally change values, and see the results either in frequency domain or time domain, in a matter of seconds to tens of seconds.

I think that some of the parallel resistor effects you are mentioning, I am just lumping in with the pickup termination resistance since the bootstrap capacitor is also in parallel from the pickup hot to ground. So when I do a set of simulation runs for a certain pickup, I just plug in the real pickup's measured L, C, and ESR/DCR, and tweak away until I get a nice range and even response for all positions of the tone pot, as you can see in the graphs. That steers it in the direction of more moderate and consistent Q across the sweep of the tone pot.

To get better emulations of more "peaky" higher Q pickups is pretty easy depending on the magnetic circuit and eddy losses of what the physical pickup is doing, but for this first pass I'm trying to stay in the same range of loaded Q shown by normal-ish pickups loaded by volume and tone controls, and some sort of dartboard average cable capacitance.

I'm measuring winding capacitance with a DER EE-5000 but started my measurements at 1 KHz (default on the meter). I am now using a 120 Hz test frequency and I will go back and measure my pickup collection to see if what I had measured is different. I am measuring capacitance of all windings connected in series on a humbucker, to shield ground, at the end of the cable.

Am going to get two demo instruments assembled, one S-type three single coil, and one 2-humbucker, and get some players pickin' on them to get some feedback on musicality and how to tweak the idea further.

Even if one isn't going for a radically acoustic-like tone on the highs, the bootstrap capacitor trick is still very interesting both because of tuneability if you want a performance control, but also if you don't, under the hood, you are able to fine tune the pickup resonance to be exactly what you want or what a certain instrument and/or player needs, without having to deal with only having standard capacitor values available and having to solder up a mini-stack of caps, or toss the pickups and try to find/make something better.

The onboard simple variable-frequency (quasi-parametric is just as many letters) mid/resonance control might also be the hot ticket for getting fine tuneability on distortion voicing/tone in crunch to heavier styles, for a given FX/amp chain.

I'm looking forward to some fun times, and as you said, there's a lot of testing and experimentation that could be done, including magnetic circuits and materials. I hope that this thread will open a few eyes up to the idea of going active in a simple way with very definite benefits. Sorry, Bill Lawrence (R.I.P. and what a great guy); batteries shouldn't only be for flashlights.