Low Impedance Over and Under Pickup Design

[bbsailor]

I want to start a discussion about something we have all taken for granted since the dawn of guitar pickup design: balanced symmetry pickup output.

All pickups that have the magnets located under the strings introduce a type of distortion the we have come to accept as a given. When the string vibrates it, travels in multiple directions and it is mainly the up and down movements that generate the greatest amount of output voltage. When the string vibrates downward toward the magnet it is in the strongest magnetic field producing the highest output and when it travels upward it is farther away from the magnet and it is in a weaker magnetic field, producing a weaker output. If you look at the output of a pickup on an oscilloscope, one half (depending on the phase of the pickup connection to the scope) of the observed pattern is always visibly higher than the other half, assuming that use 0 volts as your reference point. Magnet force generates the induced pickup output voltage but this voltage varies by the cube of the string distance from the magnet so any small change in string vertical position creates a visible lack of symmetry which can also be called distortion. This lack of symmetry also varies by the amplitude of the string vibration so pickups closer to the bridge exhibit less of this but it is still there. Take a look on a scope!

Current transformer pickups use a thin single loop about .125 inches thick and can be mounted both below the string in a typical manner and can be mounted above the string in a new way to restore the symmetry in the pickup output. This over and under dual loop design can also be made in a humbucking configuration. By balancing the magnetic field both above and below the string some new possibilites can be designed to restore full symmetry to the simple notion of a "vibrating string in a magnetic field".

I hope this sparks some more technical discussions.

Joseph Rogowski

[StarryNight]

I'm not speaking here from a learned point of view but if I were to visualize a cross section movement of a string it seems to start oscillating in the direction in which it was struck, then it settles into a more oval shape which doesn't stay on the same axis. I guess I'm wondering if having two transducer fields will make the signal any less "distorted"? at least having one transducer field gives sort of a one dimensional picture of the string movement. Would not having another transducer from above have the same distortion characteristics plus creating more phasing issues? If we filter out the string distortion (i.e. non linear movements) are we getting a true picture of the string movement?

[bbsailor]

I'm not speaking here from a learned point of view but if I were to visualize a cross section movement of a string it seems to start oscillating in the direction in which it was struck, then it settles into a more oval shape which doesn't stay on the same axis. I guess I'm wondering if having two transducer fields will make the signal any less "distorted"? at least having one transducer field gives sort of a one dimensional picture of the string movement. Would not having another transducer from above have the same distortion characteristics plus creating more phasing issues? If we filter out the string distortion (i.e. non linear movements) are we getting a true picture of the string movement?

StarryNight,

1. The early phono cartridges needed to introduce stereo signals that could be derived from a single groove, be compatable with mono cartridges, and offer enough separation to enjoy a relativly (about 25db) separation between the left and right stereo signal encoded in a single groove. See Magnetic cartridge - Wikipedia, the free encyclopedia

2. Having two transducer fields could offer more symmetry espicially if sampling the exact some above and below the exactly same position on the string. Today's humbuckers use two coils but sample the same string vibration slightly displaced by about .75 inches, the distance between the center of both humbucker coils. This slightly displaced sampling adds another form of high frequency distortion.

3. Adding a coil above the string and the under the exact same string position only minimizes the vertical string displacement distortion. To fully sample the complete string motion you would need to add 5 horizontal motion pickups in the spaces between the six strings plus one extra pole piece on each outer string (for a total of seven additional sampling pole pieces for the horizontal movement) to fully sample all vertical as well as horizontal string movement. Ultimately full string fidelity and dynamic movement detection may require this, but first minimizing vertical string movement distortion seems the most practical.

4. The vertical sampling of a string in exactly the same position above and below can only minimize the asymmetry created by only using under string sampling. The question as to whether this is more pleasing to the ear is the subjective evaluation that this discussion is seeking to uncover. Acoustic guitars have a well defined and understood difference in sound quality compared to electric guitar magnetic pickups pickups. The real question is related to the effort necessary to minimize vertical string vibration symmetry distortion and is it worth the effort?

The over and under design is just the start of this analysis.

Thanks for the questions.

Joseph Rogowski

[Rick Turner]

We tried quadrature individual string pickups at one point at Alembic with the pickups being in a "V" formation. You could also put small coils or single turn coils on either side of each string so as not to have anything over the strings, and that could be done in V formation. And this from a guy who makes horseshoe pickups to which the main objection is that huge bit of metal going over the strings which makes it impossible to pick there or damp the strings in that place. This was Sonny Landreth's chief objection to my Model T with the Rick horseshoe. He loved the sound and hated the playability factor.

I have managed to deal with string vector phase with some of my piezo pickups where the coupling to the piezo ceramic is so close that it integrates pressure mode with vertical vibrations and shear mode with horizontal vibrations.

You could balance out hum by flipping phase of every other -45 degree coil and ditto on every other + 45 degree coil. Yeah, it's a lot of coils for a guitar, but they don't have a lot of wire! Might even come up with something printed on a flexible circuit card...

[Mike Sulzer]

If you look at the output of a pickup on an oscilloscope, one half (depending on the phase of the pickup connection to the scope) of the observed pattern is always visibly higher than the other half, assuming that use 0 volts as your reference point. Magnet force generates the induced pickup output voltage but this voltage varies by the cube of the string distance from the magnet so any small change in string vertical position creates a visible lack of symmetry which can also be called distortion. This lack of symmetry also varies by the amplitude of the string vibration so pickups closer to the bridge exhibit less of this but it is still there. Take a look on a scope!

How do you know that the asymmetry that you see is not due to even harmonics in the string vibration?

The distortion must exist, but I doubt that it is very big. For one thing, the field is not pure dipole, and so it does not fall off as fast as one over r cubed.

One way to check this is to produce as low a level of harmonics as you can in the string vibration an see how much asymmetry there is. The best way I know to do this is to strike a 2nd harmonic on the no. 1 E string (touch over the 12th fret, pick over the location of the "24th"), and then look after the transient dies out using the neck pickup. I do not see asymmetry when I do this.

[bbsailor]

How do you know that the asymmetry that you see is not due to even harmonics in the string vibration?

The distortion must exist, but I doubt that it is very big. For one thing, the field is not pure dipole, and so it does not fall off as fast as one over r cubed.

One way to check this is to produce as low a level of harmonics as you can in the string vibration an see how much asymmetry there is. The best way I know to do this is to strike a 2nd harmonic on the no. 1 E string (touch over the 12th fret, pick over the location of the "24th"), and then look after the transient dies out using the neck pickup. I do not see asymmetry when I do this.

Mike,

The initial attack of a string has the greatest amount of harmonics but these harmonics quickly fade away. However, the ear is very sensitive to the initial attack of a string, espicially when the guitar or bass is trying to fit into very short musical spaces in a rhythmic manner. Guitar Pickup Signal

Rick Turner even recognized this asymmetry by sharing with us how he experimented with this issue at Alembic. "We tried quadrature individal string pickups at one point at Alembic with the pickups being in a "V" formation".

Look at the initial attack of a string using the neck pickup where the string movement is the greatest, then look at the bridge pickup. You will see the greatest asymmetry with the neck pickup because the string has the greatest mechanical swing for the first few cycles farther away from the bridge. It is still present at the bridge pickup but to a lesser extent. The voltage output of a guitar pickup follows the physicsl of an induced voltage generator and is very sensitive to the strength of the magnetic field. True, this harmonic quality fades quickly over time but the ear is still a very sensitive receiving instrument and can detect these subtle sound characteristics. That is why many posts on this forum discuss why this or that magnet, metal cover, wire insulation thickness or insulation material sound better, worse or different. This just proves that science and objective observation can explain sound qualities that we can hear.

As always, you pose interesting questions.

Joseph Rogowski

[bbsailor]

We tried quadrature individual string pickups at one point at Alembic with the pickups being in a "V" formation. You could also put small coils or single turn coils on either side of each string so as not to have anything over the strings, and that could be done in V formation. And this from a guy who makes horseshoe pickups to which the main objection is that huge bit of metal going over the strings which makes it impossible to pick there or damp the strings in that place. This was Sonny Landreth's chief objection to my Model T with the Rick horseshoe. He loved the sound and hated the playability factor.

I have managed to deal with string vector phase with some of my piezo pickups where the coupling to the piezo ceramic is so close that it integrates pressure mode with vertical vibrations and shear mode with horizontal vibrations.

You could balance out hum by flipping phase of every other -45 degree coil and ditto on every other + 45 degree coil. Yeah, it's a lot of coils for a guitar, but they don't have a lot of wire! Might even come up with something printed on a flexible circuit card...

Rick,

Your observations are very insightful. We could use small magnets between adjacent strings plus one magnet on the two end strings with single coil loops of heavy wire looped between the strings (and on the two ends) connected to a current transformer to pickup the full horizontal string movement with better symmetry. Another experiment is to obtain a ring magnet from K&J magnetics with the string going through the center hole and a metal loop forming a current sensor. This whole assembly could be fabricated into a bridge assembly to sense the movement of the string much like a stylus armature in a phono cardridge. By changing the wire size of each string loop and with experimentation, the tonal voicing of adjacent string pairs could be altered to emphasize desireable harmonics from those string pairs.

Current transformers may offer a different approach to fitting magnetic pickups in previously unavailable spaces.

Joseph Rogowski

[David Schwab]

Joe, I don't know if you have ever read the two Bartolini patents (3983777, 3983778), but he was concerned with something similar. His idea was that rod shaped magnets picked up a lot of side to side motion of the string, and that an acoustic sound board was excited by the up-down part of the string motion. So he tried to devise a way to make a low flat field over the pickup, reasoning that the up down motion would be picked up more because of the lines of flux being parallel to the top. He called these high asymmetry (Hi-A) pickups. Did it work? I haven't used many of the old Hi-A pickups, but the one I have has a unique tone.

There was another interesting pickup invented by jazz guitarist Attila Zoller (3588311), that attempted to shape the magnetic field to get a certain tone.

Another thing this reminds me of is Ned Steinberger and his orchestral instruments. He has piezo transducers mounted in the bridge, and had to do them at right angles (I think) for each string. One was good for pizz playing, but sounded bad with bowing, and vice versa. This is because of the way the string moves when it's bowed.

[bbsailor]

David,

About the mid 1980s I was working with a local South NJ guitar maker, Fred Fatovich and helped him design and build his own hand made guitar pickups. He was a very fine craftsman and I received my own custom guitar for my efforts helping him. I attended a mid 1980s Guitar Expo in Madison Square Garden, where Fred (Fedor) Fatovich had a booth to show off his guitars. I was there helping him in his booth. Attila Zoller stopped by the booth and I had a chat with Attila about pickups but he did not reveal anything new to me. http://www.dallasguitarrepair.com/fatovich2.jpg

Joseph Rogowski

[David Schwab]

That's a nice looking guitar!

Here's how the Zoller pickup is set up.

[Rick Turner]

It's not even just a matter of asymmetry; an important issue is string motion phase/harmonics. String motion is planar, but the plane of vibration slowly rotates around the string at rest center, and I suspect that the plane of vibration at any given moment may be different at different harmonics. The only phase coherent string motion (to use the term loosely) is the tension modulation in the "Z" axis, and that tension modulation is at 2 x f from fundamental all the way up. I made an experimental "Z" axis pickup when I worked for Gibson in an attempt to halve the time it takes for guitar synths to lock onto pitch. It was very promising, but I got no support from Gibson management in pursuing this. Of course one of the issues with that whole thing is that the first few cycles of string motion are largely enharmonic noise.

[Mike Sulzer]

String motion is planar, but the plane of vibration slowly rotates around the string at rest center, and I suspect that the plane of vibration at any given moment may be different at different harmonics.

This does not look planar to me, but maybe I am viewing it wrong:YouTube - Guitar string vibration. Why do you think it is planar? If so, a standard guitar pickup, which is primarily sensitive towards and away from the guitar, should have an output modulated by the rotation rate. I do not hear that.

[Rick Turner]

I watched that video and didn't find it very useful.

Spend a bit of time looking at string motion with a synchronized strobe light. You'll see planar motion for any given harmonic, though the phase of the fundamental and harmonics will not match. It's all over the place and a bit crazy, and the plane of string excursion rotates around the "Z" axis of the string.

This is one of the interesting differences between our world of plucked strings and the world of bowed strings where the bow has a strong degree of control over the planar drive of the strings. This is also why it's been hard for some designers to come up with piezo pickups for upright bass that respond equally to arco and pizzicato string excitation.

Another thought...

One of the things that many of us pickup makers often lose sight of is that the classic pickups...high impedance with all the response anomalies, etc., have become the base line for guitarists and bass players. Screwy is normal, and when we try to "scientifically improve things" we do run the danger of attempting to be the tail wagging the dog. High impedance pickups have characteristics based on LCR that many players find desirable. Improving things too much is possible!

[David Schwab]

I made an experimental "Z" axis pickup when I worked for Gibson in an attempt to halve the time it takes for guitar synths to lock onto pitch. It was very promising, but I got no support from Gibson management in pursuing this. Of course one of the issues with that whole thing is that the first few cycles of string motion are largely enharmonic noise.

I bet they kicked themselves later, because they actually worked on and patented some hex pickups for synth and MIDI stuff.

[Mike Sulzer]

I watched that video and didn't find it very useful.

It looks like it is rotating around to me. You have to think about how that motion projects onto 2D.

If it is a rotating plane, how come you do not hear the changes in amplitude as it rotates in and out of the direction of maximum pickup sensitivity?

The harmonics have higher frequencies than the fundamental. Maybe that explains what you are seeing with the strobe light.