And now for something completely different...

[Greg Holmberg]

Last year I began a journey to understand guitar pickups. I read some of Helmuth Lemme's articles, and was surprised to learn that they can mostly be described by their frequency response, which is a fairly simple curve (very different from the frequency response of an acoustic guitar, for example).

Of course, no manufacturer publishes the frequency response of their pickups, which is disappointing, since it makes it difficult to compare pickups. This leads to a situation in which guitar players describe pickups in subjective terms, much like wine drinkers. There's almost no way to know if you would like one before you buy it.

This leads the life-long "tone quest" many guitar players find themselves on. They buy and sell many products trying to find the magic combination. Pickups, overdrive pedals, guitar amps, guitar speakers. Many never come to the magic pot of gold at the end of the rainbow. They just keep buying products forever. This is great for the manufacturers, not so great for the players. It's clear why it's not in the pickup maker's best interest to publish the frequency responses of their products.

So I ask myself, what is really happening when a guitar player is trying out these combinations of products? Although there are several effects going on, I think the main one is the superposition of the frequency responses of the pickups, the pre-amp circuit, the guitar speaker, etc. Each has it's own unique frequency curve, and when you combine them you get a net frequency response that is interesting.

So, essentially, when players try different combinations, they're just taking devices with hard-coded EQ settings and combining them, hoping for something good. The same result could of course be achieved with EQ devices that are more flexible. Instead of buying hundreds of products over your lifetime, you could just buy a multi-parametric EQ and get all the frequency responses. You could easily measure known good tones with a microphone, audio interface, and a computer, and then set the parametric EQ to match.

Now, I could try to do this today with a pickup with a known frequency curve, and add/subtract from that to get what I want. However, this becomes a bit difficult with the large resonance peaks of many pickups, and then it becomes impossible when the pickup completely cuts off signal above a certain frequency.

It would be a lot easier if I had a pickup that has a flat frequency response, at least up to the highest frequency I care about. I could then feed that into my computer and use software EQ to get any tone I want. I learned from Helmuth's articles that a pickup's resonant frequency can be pushed up outside the range of human hearing, leaving a flat response in the range I care about. He called it a "low impedance" pickup, and it had already been done by Les Paul in the 1960's.

Les Paul's "Recording" Guitar (another interesting article by Helmuth) had a very low impedance pickup, with an inductance of 0.008 Henries and a resonance peak over 200 kHz!  EQ was in hardware not software (his "decade" switch of capacitors), but basically the same idea. A guitar that can mimic any other guitar because it starts out neutral. Here are some recordings Les Paul made with the guitar demonstrating the different tones: side A, side B.

So I had Helmuth build me such a pickup. Here it is:



Name: JZ2
Design: humbucking using two half-width coils—a “split” pickup as in a Fender Precision Bass or the G&L Z-Coil guitar pickup. Single point of sensing, so as not to lose or distort certain frequencies.
Balanced output (three wires) to reduce noise in the cable. Isolated from ground. Can be used unbalanced (two wires) also.
Dimensions: 70 x 38 x 12 mm, 2.75 x 1.50 x 0.50 inches.
Cover: black plastic. Copper shielding inside.
Mount: end of fretboard.
Space to strings: 3 mm recommended.
AC inductance: 0.3 Henries.
DC resistance: 1.6 kΩ.
Output voltage: 15-20 mV_RMS = 21-28 mV_PK.

This is not intended to be used directly into a guitar amp. It would require a pre-amp to change both the voltage and the impedance. I wired it to an XLR jack (balanced), and I connect it to an audio interface (microphone input). I connect that to an iPad and add EQ and other effects using the BIAS FX 2 software from Positive Grid, and play through full-range studio monitors to get those higher frequencies.

I tested this pickup with Ken Willmott's Integrator device, a MOTU M2 audio interface into a Mac, using the Room EQ Wizard software.



Minimum load in red (10 MΩ, 10 pF): 15.44 dB @ 17.0 kHz.
Load in green (200 kΩ, 470 pF): 12.07 dB @ 10.26 kHz.
Inductance test in blue (10 MΩ, 4700 pF): 4270 Hz.

As you can see, the resonant frequency is very high, and in the range of my hearing (about 13 kHz--I'm old!), it's essentially flat.

In the following recording, the pickup is installed in an inexpensive fully hollow archtop (laminated top, strings are D'Addario Half Rounds, 10-46), at the end of the fretboard, and then wired directly to the audio interface (no tone or volume controls), into an iPad and recorded with Garage Band. No EQ, compression, reverb, or any other effects were added. You can hear that it's a very quiet pickup--no hum at all. There are eight segments, the first five are played with a pick, and the last three are played with fingers. I pick up near the neck, never down by the bridge. Mistakes were made, but I think you can get an idea of the tone.

JZ2 recording

By itself, it sounds uninteresting--kind of like a boring acoustic guitar. Even beyond an acoustic guitar, which ends at about 12 kHz.

But it's a great starting point for manipulating into different shapes. The "Guitar Match" feature in BIAS FX 2 seems to do a pretty good job with this pickup applying EQ adjustments to mimic certain pickups. Alternatively, I can easily cut off frequencies above various levels, and add a hump at a certain point, and get a variety of common tones. Personally, I prefer a clean Jazz tone like a P-90, and I can get that by cutting off above 4 kHz and boosting the 2-3 kHz range. Maybe boost a little in the bass around 100 Hz. However, if I want to sound like an acoustic guitar, the frequencies have not been lost, and with a few adjustments, I can have that sound too.

Helmuth also tested the pickup. His tester has a minimum load capacitance of 47 pF plus some unknown amount from a 30 cm cable.  The first two graphs show the JZ2 tested with a 300 kΩ resistor and various capacitors:

Position	Additional capacitance pF
11	0
10	470
9	1000
8	2200
7	3300
6	4700
5	6800
4	10000
3	15000
2	22000
1	33000

The third graph shows the JZ2 tested with no additional capacitors (so 47 pF), but various resistors:



So, collecting all the data from both tests:



What I find interesting about these graphs is that one could move the resonant peak down to almost any frequency one wanted with the right capacitor (as Les Paul did), and one can lower the peak dB with the right resistor.

I think you could have a single guitar that could mimic most pickups by applying a set of resistors and capacitors to the pickup. Combine that with a pre-amp that has a shelving-style treble and bass controls (Baxandall), and converts the voltage and impedance to something a guitar amp wants, and you have a guitar that can sound like any other guitar.

I'm surprised that some guitar electronics vendor out there hasn't already done this, but then again, it might spell the end of the "tone quest", and that's not good for business!

[gckelloch]

Any fan of Monty Python is OK in my book. That was good work you did. Les Paul knew the score for sure. The late Bill Lawernce once wrote an article on the issue of guitar amps all being Hi-Z now and the limitation that presents. He became chief engineer at Gibson around the time the Les Paul Recording guitar was developed, so he may have at least had something to do with the pickup design. Bill did design other Gibson guitars and/or their pickups to follow, including the: L5-S, L6-S, Les Paul Signature, Howard Roberts models, and Ripper & Grabber basses.

If I understand correctly, very Lo-Z pickups will have some low-end loss when plugged into a Hi-Z input resulting in a relative lack of bass tightness/clarity. Very thin coil wire can compensate, but there are physical limits on how thin a wire gauge can be wound even by a machine. I’m not sure how the output voltage of a pickup can really be judged since it depends on how it's driven and the thermal limits of the coil. Surely the thermal limit of the coil is much higher than the value you listed?

That being said, your pickup sounds very nice with great highs. I wonder if the coils are wired in parallel to achieve the low inductance? I am also curious as to what's in the core -- AlNiCO of some type? Bill Lawrence once made a low wind count 0.5H L600 (Mini-HB with AlNiCO poles) for Jazz guitarist Grant Greene.

Freq response isn’t the only guitar pickup characteristic. There’s also how the magnetic circuit, coil dimensions, and wire gauge affect signal asymmetry and either create “false” harmonics or alter the lower to upper harmonic ratio for each note = note timbre. Adjustable pole screws is a way to alter note timbre with a given coil dimension. I’m not sure how false harmonic generation is affected, but I assume it is at least dependent on coil proximity. Antigua showed how magnetic string pull also affects note harmonics via intermodulation and increased higher harmonic content, and I suggested a solution to increase higher harmonics in lower coercivity pole pickups without increasing intermodulations via pick design and material. There may also be some level of freq range “ducking” as the magnetic field in a string vibrates closer to a (dai or para)magnetic metal in close proximity to the coil, but I have no way of verifying that.

Hate to sound like a broken record, but Bill Lawrence knew of and addressed these things as best as possible with his Microcoils. The high-end can be further extended above the usable range of any known guitar speaker either without bass loss using his “Q-filter” or with significant bass loss and no resulting mid-dip when wired as an L-filter.

You should also know there is some phase-cancelation in the space between the coils in a split-coil design like that. Coil overlap may address it, while affecting the aperture window in the overlap region. The Zexcoil slanted-blade hum-canceling design is probably the best way to keep both tone & timbre consistent for each string.

[aquin43]

The inductance is a bit high* to feed a 200 ohm mic input channel on a typical preamp which may have an input impedance of 2k to 3k. Although modern transformerless input stages could have much higher input impedances, values around 2k are used both for historical reasons and because that is the optimum load for most unpowered mics.

I see also that the value of R/L for the pickup is 5k33. This is typically around 2k5 for the usual run of pickups.

* 2k and 300mH --> 1k06Hz

[ms]

Did you try your pickup into an instrument input?  I think it has enough output so that it would work.

The output impedance of a pickup is a complicated function of frequency, and the load it looks into with a normal guitar amp input is much higher at most frequencies. The amp input resistance can cause a little damping near the resonance, and its capacitance could lower the resonant frequency a bit, but neither is  a huge effect.  In your case, the load of the amp would do almost nothing, and so your very high frequency resonance might need a bit of damping, depending on the materials used in the pickup. If you use a volume control on this guitar, 25K would be a low enough value so that the tone would not be affected by its setting; that is, cable capacitance would have a very small effect.  Also 25K would be high enough so that it would not affect the pickup much, except to damp the (ultrasonic) resonance, if it is significant.

[aquin43]

In my previous post I mentioned R/L as being rather high. The significance of this is how it affects the behaviour of the conventional tone control. Usually there is a resonant peak of a few dB at around 400Hz when the tone control is fully down. Some like this, some don't but it is a feature of the electric guitar as we know it. With R/L in the 5k region, this doesn't happen.

800 ohms would be a better value for R in this case, requiring a heavier wire gauge. It isn't a trivial problem since L falls as the square of the turns while R falls directly with the turns so R/L tends to rise as L is reduced and bobbin space may limit the useable wire gauge.

[antigua]

May 27, 2021 22:44:08 GMT -5 Greg Holmberg said:

So I had Helmuth build me such a pickup. Here it is:

When did Helmuth Lemme make this pickup? Is this something that happened recently, or a while ago? I didn't realize that he made pickups.

I think there's two way of looking at it, one is your way, that you can approximate a lot of pickups by tailoring the resonant peak, or another way, which says that if a capacitor can mimic the sound of other pickups, then they were never that different in the first place, and marketing goes a long way to make it seems as if there is a lot of difference, for example, by saying that one 2.4H pickup is good for blues, and another is good for jazz.

That being said, since the resonant frequency could be anything, there's theoretically a huge number of possibilities, but we can only distinguish so much variation. On another forum, someone asked how many audibly distinct pickups you could create by varying the resonant peaks, and while that's subjective, one thing that is noteworthy is that for any given pickup manufacturer, they will only offer three, maybe four variations for a given pickup. For example, they might have a "vintage", then a "blues hot" and then a "high gain" model, for any given type of pickup, which suggests that amongst the infinite possibilities, you can choose only about three points along the spectrum, and have them sound distinct. 

Some pickups like the "Little 59" promise to give a P.A.F. type sound in a Strat, and all they're really doing is delivering a very low resonant peak, and asserting with marketing that it should sound like a P.A.F., but the placement of the pickup in a Les Paul is such a big factor in the sound people recognize as "P.A.F." that a lot of people comment that a Little 59, high inductance Strat pickup, is not convincingly close to a real P.A.F. in a Gibson guitar. I think the same problem would apply here, you can assert that cap switching yields many pickups variations, but unless you have a guitar that allows the pickups to physically move around, like a Turner Model 1, it's not really going to nail the same timbre. But I don't mean to sound like a hater of the idea, I put dial cap selectors in a lot of my Strats, I think it's a great mod. 

I also don't know that a pickup has to be low-Z to have a flat response. If a pickup has an inductance below 1.5 henries, generally it's resonant peak will be above 5kHz, even with a longer guitar cable, and that will result in a mostly flat response through the operable range of the guitar amp and speaker. I think the bigger advantage of the low-Z pickups is that they are virtually noiseless. Some noiseless active pickups, like the Fishman Fluence appears to be little more than low-Z pickups with tone shaping built in, to give a high-Z sound. 

[Greg Holmberg]

May 30, 2021 12:17:49 GMT -5 antigua said:

May 27, 2021 22:44:08 GMT -5 Greg Holmberg said:

When did Helmuth Lemme make this pickup? Is this something that happened recently, or a while ago? I didn't realize that he made pickups.

I think there's two way of looking at it, one is your way, that you can approximate a lot of pickups by tailoring the resonant peak, or another way, which says that if a capacitor can mimic the sound of other pickups, then they were never that different in the first place, and marketing goes a long way to make it seems as if there is a lot of difference, for example, by saying that one 2.4H pickup is good for blues, and another is good for jazz.

I also don't know that a pickup has to be low-Z to have a flat response. If a pickup has an inductance below 1.5 henries, generally it's resonant peak will be above 5kHz, even with a longer guitar cable, and that will result in a mostly flat response through the operable range of the guitar amp and speaker. I think the bigger advantage of the low-Z pickups is that they are virtually noiseless. Some noiseless active pickups, like the Fishman Fluence appears to be little more than low-Z pickups with tone shaping built in, to give a high-Z sound. 

He made it in June 2020. He has a few pickups for sale here.  But he can make you whatever you want.

He also made me:

• a "C-switch" with 11 capacitor positions
• an active pre-amp (changes both voltage and impedance so I can connect to a regular guitar amp)
• Baxandal-style shelving treble and bass control.
• an inverter, so I can still output a three-wire balanced signal after all the above.

He has all this because he repairs Les Paul "Recording" guitars, as he describes in this magazine article.

I'm sure that 1. there's a lot more to a pickup's sound than just it's frequency response (but I'm somewhat ignorant of what that would be), and 2. There's more to an electric instrument's sound than just it's pickup.  Although I will say that I've heard videos on Youtube of electric guitars built from glass, stone, concrete, styro-foam, and salt, and they still sound like electric guitars!  Sure, there are differences, but compared to the difference made by pickup choice, guitar preamp circuit, amp speakers, and twists of bass/mid/treble knobs, the body material was insignificant (IMO).  About equal to one notch on the treble knob on the amp.

I was aware that guitar amp speakers don't produce much above, say, 6 kHz (as one can see on Jensen's web site).  In which case, having a peak above 10 kHz won't be heard.  In my case however, I was looking for a recording pickup that I would only play through full-range studio monitors or a PA system.  This way I can also get an acoustic-like sound.

This pickup certainly is virtually noise-less, both because of the low-Z and also split-coil design with balanced output.

I'm enjoying EQ'ing it into different sounds.  It's fun to use with software modeling of effects and amps.  I haven't even wired up the electronics he built me, I'm having so much fun.

Greg

[gckelloch]

Is it AlNiCo II core?

The Fluence SC's are stacked designs similar to the Wilde NF series. I've seen one opened up. The two wafer stacks look ~4mm tall and ~1cm wide. They are probably just fairly low inductance Hi-Z. You could get a virtually flat response out to ~15kHz with 1H considering how low the capacitance load would be. The only advantage to the wafer design I see is they don't have to bother winding coils. They do sound like they have stronger fundamentals than 42AWG SC's in the demos I've heard by Greg Koch. The 0.044mm/46AWG Microcoil coils are likely substantially more dense. I still think Zexcoil are the best and most efficient hum-canceling design. I think he uses 44AWG.

Anyway, the JZ2 sounds great to me. Congrats!

[helmuth]

Comprehensive pickup theory here:
www.buildyourguitar.com/resources/lemme

Pickup measuring technique here:
www.planetz.com/wp-content/uploads/2014/10/Pickup_Measuring_Technique.pdf

[ms]

May 30, 2021 18:47:00 GMT -5 Greg Holmberg said:

I'm sure that 1. there's a lot more to a pickup's sound than just it's frequency response (but I'm somewhat ignorant of what that would be), and 2. There's more to an electric instrument's sound than just it's pickup.  Although I will say that I've heard videos on Youtube of electric guitars built from glass, stone, concrete, styro-foam, and salt, and they still sound like electric guitars!  Sure, there are differences, but compared to the difference made by pickup choice, guitar preamp circuit, amp speakers, and twists of bass/mid/treble knobs, the body material was insignificant (IMO).  About equal to one notch on the treble knob on the amp.

Most important: sampling a particular region of the string creates a  certain pattern of harmonics that is a function of the active length of the string.

As a wise person said many years ago:  Put a strat pickup on an an acoustic guitar, and the resulting electric sound is nothing like a strat.

Also, the magnitude of a difference might be small, but it nonetheless might be very significant because it is a very different kind of difference than you can  achieve it by turning a knob or changing a resonant frequency.

[ms]

May 31, 2021 3:37:12 GMT -5 helmuth said:

Comprehensive pickup theory here:
www.buildyourguitar.com/resources/lemme

Pickup measuring technique here:
www.planetz.com/wp-content/uploads/2014/10/Pickup_Measuring_Technique.pdf

Mr. Lemme:

Your writing was the first intelligent material on pickups I found on the web some years ago.

By the way, the web site is not accessible, but the pdf file is. 

[Greg Holmberg]

May 27, 2021 22:44:08 GMT -5 Greg Holmberg said:

So, essentially, when players try different combinations, they're just taking devices with hard-coded EQ settings and combining them, hoping for something good. The same result could of course be achieved with EQ devices that are more flexible. Instead of buying hundreds of products over your lifetime, you could just buy a multi-parametric EQ and get all the frequency responses. You could easily measure known good tones with a microphone, audio interface, and a computer, and then set the parametric EQ to match.

Now, I could try to do this today with a pickup with a known frequency curve, and add/subtract from that to get what I want. However, this becomes a bit difficult with the large resonance peaks of many pickups, and then it becomes impossible when the pickup completely cuts off signal above a certain frequency.

It would be a lot easier if I had a pickup that has a flat frequency response, at least up to the highest frequency I care about. I could then feed that into my computer and use software EQ to get any tone I want. 

...

By itself, it sounds uninteresting--kind of like a boring acoustic guitar. Even beyond an acoustic guitar, which ends at about 12 kHz.

But it's a great starting point for manipulating into different shapes. The "Guitar Match" feature in BIAS FX 2 seems to do a pretty good job with this pickup applying EQ adjustments to mimic certain pickups. Alternatively, I can easily cut off frequencies above various levels, and add a hump at a certain point, and get a variety of common tones. Personally, I prefer a clean Jazz tone like a P-90, and I can get that by cutting off above 4 kHz and boosting the 2-3 kHz range. Maybe boost a little in the bass around 100 Hz. However, if I want to sound like an acoustic guitar, the frequencies have not been lost, and with a few adjustments, I can have that sound too.

In his blog post at Cycfi.com Joel de Guzman essentially makes this same point.

Lo-Z: Let’s Clear a Misconception

I wish him luck in his effort to change minds about pickups with flat responses.  I sincerely hope he succeeds.

Has anyone tested the Cycfi pickups?

Greg