Pickup info

Pickups, Diy Electronics, Stompboxes, Guitar wiring...
richard37066
Posts: 517
Joined: Sat Dec 11, 2010 11:44 am
Location: Gallatin, Tennessee, USA

Re: Pickup info

Post by richard37066 »

Ed -

I'm certain that you know - after all these many months - that I'm not in the least bit interested in the ubiquitous and traditional magnet/coil variety of pickup. Your semiconductor design has me totally enamored as it is a light year ahead of a similar quest that I made many years ago.

You may have, at least, partially answered one question in a response to Georg earlier in this thread but I'll pose it again in an effort at clarity. To wit: - In a "conventional" magnetic pickup, the length of the string subject to detection is sometimes referred to as the "capture length", "capture aperture" or a phrase quite similar and is determined, in part, by the width dimensions of the pickup elements. In your response to Georg, you indicate that, - in your design - it might be in the neighborhood of a half inch or so. This is much greater than the length of the resistance element in the IC used. Is this reasonably accurate? I find no reference to indicate a possible dimension. The absolute dimension is not critical to my thinking but would be very informative.

In addition, can you give me the approximate depth of the combination of IC package, PC board and solder "bumps" on the bottom - if any? The desire for this is explained below.

Traditional guitar builders place the center of the magnetic pickup approximately 2" from the center of the changer finger or thereabouts. I assume that this was adopted as a result of a perceived notion as to what a PSG should sound like re the relationship of fundamental to harmonic frequencies. A compromise, at best.

It should be obvious to everyone that the placement of any pickup in relation to the center of the changer finger will determine the ratio of the detected amplitudes of the fundamental to the harmonics - the sensing of the nodes and antinodes in the vibrating string. I believe that I'm on firm ground here but, in the simplest of terms, as one moves the pickup further from the changer then the greater the resultant detected amplitude of the fundamental to that of the harmonics. This might even be termed a mechanical tone control (Gasp!). IF - IF - the depth of your assembly is thin enough then I propose to place it between the strings and the neck as opposed to it being buried and screwed within a cavity. With a reasonable range of adjustment - longitudinally - one should be able to satisfy a personal, tonal "sweet spot" - one not needing much tweaking in the search for a personal optimum. I would anticipate the need for relatively fine adjustment but that is simply a mechanical design problem to be tackled at a later time. Needless to say, your configuration lends itself directly to individual gain, EQ and lord knows what for each string thereafter. This, in contrast to having a fixed pickup location and resorting to an array of broad-band amplifier adjustments which may, or may not, produce the desired end result. In my experience - and within my personal preferences - this circumstance has proven to be lacking and very frustrating.

Since there is no magical dimension between strings and neck (fretboard) I view my proposal to be feasible providing that I can shoehorn your design within that space and with reasonable clearances at both top and bottom.

What sayeth you -yay or nay?

Richard
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

Richard…et al; Some info on the semiconductor pickup:

No magnets needed…but can be used if a reason shows up.

The output of the common wound pickup is kN dphi/dt where k is a constant to cover the network effects etc, N is the number of turns, dphi is the change in flux, and dt is the change in time. As the string vibrates in the magnetic field a change in flux is caused…this causes a voltage across the coil. This indicates that the output is a function of the number of turns, and of the amount of flux changed by the vibrating string (dphi/dt). The k will be different for different circuit (magnetic and electrical) configurations.

The semiconductor pickup is made up of a Wheatstone bridge of magneto resistive material in a small “chip”, the output of which is a voltage change that is a function of the angle of the magnetic field…not the magnitude (strength) of the mag field, so it goes down to DC…(hence can also be configured to give non contact position information if used re pedals and levers.) The vibrating string will provide the change in angle.

For the sensor, a change of 45 degrees (for the chosen chip HMC1502) will cause a voltage change of 120mV…add an op amp gain stage to boost this to the level of a common wound pickup output.

The pickup spacing beneath the strings can be the same “two quarters” as used for the common pickup spacing. The green board shown on the Photobucket site is 3+3/8” long X 1+1/2” wide X 3/32” thick if headers for connection are not used, and about 7/16” if headers and connectors are used.

The six channels of the pickup take up 2+1/16” for strings spaced at 11/32” apart. The board is fabricated as a snap apart for the channels and other sections so that each channel may be used as a position sensor. There are no solder bumps on the back. The board could be configured to be about 1” wide if desired. The 1+1/2” width dim was chosen so that the board would fit the Sierra interchangeable pickup module., hence the BEAST.

The circuitry is designed to be used as either an AC, or DC sensing element depending upon how the board is populated.. The AC sensing is used for the pickup, and the DC is used for the position sensing function. The Wheatstone bridge has a resistance of 5K ohms. The pickup is powered by 3VDC…low current.

Common pickups have an output that is greater as the string vibrates toward and away from the magnets; the semiconductor pickup has an output that increases as the string vibrates across the chip.

I have an alternative wound design that is only 7/16” thick that I used to slide to any position beneath the strings. It is a bifilar wound and differentially terminated Lo Z configuration to be used with the CYBER III box that I did up for SIERRA about two decades ago.

Because the pickup market is larger for the standard guitar, the next build will be for string spacing of 3/8”. This will allow slanting it a bit to use with narrower string spacings such as PSGs.

As the bar is moved up and down the string(s), the location of the nodes and nulls of the summed vibrations change locations re the “2 inches from the bridge/changer fingers” also change. How wide a swath of these vibrations sensed by the common pickup depends upon the mag circuit design, and the proximity of the poles to the string (big tonal change here). Look at the mag field mapping of the field above the strings and at the poles of the pickup shown in the Photobucket file re pickups.

For those that are concerned with the mag field damping the string vibration (lowering sustain), the semiconductor pickup does not do that.

Edp
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

“To have resonance or not to have resonance”. This was asked in another thread that related mostly to the body and mechanism contribution to “tone” and “sustain”. Here we ask it as related to pickups and strings.

Music wire strings are the accepted norm for the PSG. There are few types of music wire, but there are a number of different processing methods…each resulting in its’ own sustain and tone characteristics, minimal difference though there be compared to the difference in pickups

The common wound pickup comes in a variety of shapes and sizes. Several types of magnets may be used to create the magnetic circuit. Most of the variations are in the return path of the magnetic field. Screws are used along magnetically conductive cups/blades, with different numbers, types, and shapes of magnet.

The common types of magnet are Alnico (several compositions), Barium Ferrite (BaFe), and several types of “rare earth”. The Alnicos are usually ground into cylinders. The length to diameter should be 3 to 1 for stability. The BaFe are usually ground into bar form.

It should be apparent from the previous post(s) re the effect of wire and winding that the electrical resonance of the pickup can be controlled by amount and size of wire applied.
If increased resonance is desired, large diameter wire should be used as the R to L ratio will be lower. If decreased resonance is preferred, small diameter wire should be used as the R to L ratio will give a flatter spectral (frequency range) response.

When the pickup is placed beneath the strings, the magnetic field produces a “window of susceptibility” within which the vibrating string(s) will cause a change in magnetic flux, thus generating a current to flow in the coil, and a varying voltage across the coil as seen/measured at the ends of the windings. The variation in the voltage is a function of the frequencies of string vibration that fall within the window of susceptibility.

The window of susceptibility is a function of the pickups magnetic circuit, and its’ orientation and proximity to the strings.

Now for the show and tell:
Fig #1 shows the “test bed” = the BEAST.

Image

The BEAST was constructed to include a longer scale length (about 30”), an integrated changer and keyless/gearless tuner, changer placed on the players left, shallow angles for string attachment, narrow and interchangeable materials for bridge contact area, no neck block, dual interchangeable tapped pickups, a drop in fret board with room beneath it for circuitry, an aluminum body with mica covering, more pedals and levers than sanity requires (10 and 7), and a constant 11/32” string spacing from bridge to nut.

All of these “differences” from the standard PSG could have served to make a disaster re attack, decay, and timbre. The test evidence for the results vs. 32 other PSGs was presented in graphic form (Frequency Spectrum Analyzer) on the other Forum several years ago. The FSA results gave the spectrum graphics as a function of the signal present at 0, 2, 4, & 8 seconds after string excitation.

Vibrations in the various parts of the body (front apron center and ends, end plates, string termination blocks, changer mount, legs etc.) were taken using the contact mic from the Peterson VS-II tuner fed into the FSA…the point being that the body vibrations are caused by stealing energy form the string(s) vibration spectrum thus shortening the time response of the string(s).

Did we have more or less overall resonance, where in the spectrum was it, was it pleasing or not as compared? Pleasing or not is a personal thing…the rest is quantifiable and the numbers are repeatable.

Fig #2 shows the neck pickup on the BEAST as used here…two rows of magnets, one row of return screws, tapped coils…wire sizes unknown.

Image

This is a 14 string pickup. For any given number of turns, it will have more resistance and capacitance, and more inductance than a six, ten, or 12 string pickup built using the same principles…hence its’ electrical resonance will be different. This pickup was built by Danny Shields.

Fig #3 shows the magnetic field at the surface of the pickup with the strings above the sensing material
.
Image

Here we see a magnetic map of the magnetic field for the pickup shown in Fig #2. The pickup to string distance is about the standard street wisdom value (two quarters) . The sensor material is in contact with the magnets. The magnet diameter is 3/16” and the string spacing is 11/32”. The magnet rows are 1” apart, and the return screws are ½” from the magnet rows. The resulting fields at the magnets and screws are seen in the light green shading. The real question is, how much of this field reaches the strings and where is it located? Let’s look at Fig #4 for an answer.

Fig #4 shows the magnetic field with the sensing material moved to the top of the strings.

Image

This is the same as Fig 3 except that the sensor material has been moved to the top surface of the strings showing the magnetic field as it applies to the strings. Notice that the fat strings are “flux hogs”, and the poor little G# (third from the left) looks like it is starving to death…how can these produce a balanced signal? The output is proportional to the amount of flux disturbed as a function of time. The small strings have a higher frequency of vibration and therefore disturb the flux faster than the fatter strings Thus evening out the outputs

You may notice that the end strings are in a slightly different magnetic environment than the adjacent ones.

Susceptibility is a term used in magnetics to define a magnetic property. You may notice that the field changes with separation from the magnets and also focuses in on the strings. The length of the strings falls within a “window of susceptibility” wherein if the string vibrates it will cause a current in the pickup that is proportional to the flux density and the velocity of vibration. There are many velocities of vibration in any string (= the harmonics). The shape of the field(s) at the point where the strings intercept it will have an effect upon the timbre (summed harmonics). String to magnet spacing changes will change this relationship, and also the overall amplitude. It will also affect the attack.

The excited string(s) will vibrate in a circular way = toward and away from as well as across the magnets. A harmonic may vibrate across at the same time that the fundamental vibrated toward and away from the magnets etc.. Because the sensitivity is greater toward and away from the magnets, the result is a churning in the harmonic relationships as heard. This can be seen in FSA real time traces. All magnetic circuits do not have the same axial sensitivity hence they may have a difference in “liveliness” of sound. Then there is the manner in which the strings are terminated that can affect the axial rotation issue.

Fig #5 shows 4 pickups and the amount of room that they take up on the neck of the BEAST.

Image

Here we see the method used for the bridge and string attachment on the BEAST (our test bed) = shallow angles, narrow contact point on the interchangeable bridge piece, and three types of pickups to be discussed. Also shown is the switchless method of choosing the usage of the two tapped pickups = slide back and forth to combine or isolate. The use of switches is reserved for choosing the taps and phases of the pickups. The neck pickup is on, the bridge pickup is off.

Fig #6 shows a different angle on Fig #5.

Image

This as a different angle on the 4 pickups (3 types).

The short white one is a bifilar (two wires wound at once and bonded together) using a Barium Ferrite (BaFe) rod magnet. The wound wires are a red and green pair. The start red and finish green are connected as a center tap. The result is used with an instrumentation amplified front end in differential form. This method gives a Hi Z input, and maximum common mode rejection (think hum elimination. The wire used is #48 gauge = smaller than usual. Check the wire data table on the Photobucket site for the resistance difference between #48 and say #36. The #48 wire provides a lower R to L ratio (lower Q) therefore a flatter frequency response. This approach gives a thin pickup. One use is to slide it between the steel’s strings and the fret board to see/hear the result pickup position re string length.

The short green board contains one configuration of the “semiconductor” pickup. Thin is in. More later.

Fig #7 shows a BaFe bar magnet re the strings.

Image

Here is a shot of the BaFe bar magnet. Length to breadth is not an issue re BaFe. Notice the grind mark across the bar. This will cause loss of magnetic surface properties as to much heat was generated during surface grind. Details will be seen later Figs.

The magnetic circuit may be controlled using BaFe by grinding slots etc. to focus the mag field re the strings. There are many other variations in the possible uses…one shown later in this blurb.



Fig #8 shows The magnetic field from the BaFe bar with the sensing material on the surface of the bar magnet.

Image


Here is the field from the BaFe magnet with the sensor material in contact with the magnet. Here you can see the magnetic result of the aggressive grinding on the bars surface.

Fig #9 shows the magnetic field from the bar magnet with the sensing material at ¼” spacing above the BaFe bar.

Image

Here we see the magnetic field with the sensor spacing ¼” above the magnet’s surface. This would be a change in the susceptibility window = timbre change.

Fig #10 shows a pickup made with BaFe bar magnets.

Image

Here is a pickup using BaFe magnets and #48 bifilar wire.

Fig #11 shows the mag field from the BaFe pickup with the sensing material on top of the pickup, and the strings ¼” above.

Image

Fields at the surface of the BaFe bar with the strings ¼” above.

Fig #12 shows the mag field from the BaFe pickup with the sensing material on top of the strings which are ¼” above the magnet.

Image

Field from the BaFe bar as seen with the sensor on top of the strings and the strings ¼” above the magnet.

Fig #13 shows the 3 pickups and the amount of space that they take up beneath the strings…the third pickup (far left) is the semiconductor pickup.

Image

For those that wish to know what can fit where, here is the comparative spacing re the pickups and the strings. The string to fret board distance is about ½”.

Fig #14 shows BaFe magnet against string bottoms, sensor against string tops. Pole toward strings.

Image

Now for a bit of fun to stimulate your imagination: This shows the mag fiels with the BaFe bar against the bottom of the strings and the sensor on top of the strings. The magnetic poles are facing toward and away from the strings. Compare this with the next Fig where the mag poles are facing along the neck.

Fig #15 shows BaFe magnet against string bottoms, sensor against string tops. Poles along neck.

Image

Here we have the same setup as in Fig #14, except the bar has been rotated so that the poles face along the neck. Lots of room to play with the susceptibility window using just one bar…how about using more?

Fig #16 shows the two BEAST pickups, sensor against the magnets, strings above.

Image

The neck pickup is made by Danny Shields using two rows of Alnico magnets with a row of return screws.
The bridge pickup is made by DiMarzio with two rows of return screws and what looks like BaFe magnets. Compare the window of susceptibility differences.

Fig # 17 shows the two BEAST pickups, sensor in contact with the top of the strings..

Image

The neck pickup is made by Danny Shields using two rows of Alnico magnets with a row of return screws.
The bridge pickup is made by DiMarzio with two rows of return screws and what looks like BaFe magnets. Compare the window of susceptibility differences.

All the above graphics, and more are available on my Photobucket site.

http://s75.photobucket.com/albums/i287/ ... 20SENSORS/

Maybe we will do the mag field mapping in another axis…then again maybe not.
richard37066
Posts: 517
Joined: Sat Dec 11, 2010 11:44 am
Location: Gallatin, Tennessee, USA

Re: Pickup info

Post by richard37066 »

Ed -

You know that I'm fascinated by all things technical but, in viewing your graphics, I couldn't help but wonder about a couple of things.

In no particular order:

• Probably because of its simplicity - passive operation - the magnetic pickup is the overwhelming choice amongst builders. Is there really that much to be gained in the tweaking of the animal - in terms of the response re the redistribution/shaping of the fields, etc? 'Twould appear as though a dramatic improvement were needed before widespread acceptance by a preponderance of "average" ears. Not an order of magnitude, mind you, but an indisputable superiority over the garden-variety pickups now available.

• The downside to your semiconductor pickup is the necessary addition of power supplies and electronics which is of no concern to me. I welcome anything which exhibits higher performance than what we now have. This is a lead-in to another question: - Have you made a judgement - however subjective - as to the aural qualities of the semiconductor pickup versus those that you have installed in the Beast? Use any adjectives that come to mind - it'll get sorted out in due time.

• Your "sensor" is neat! A form of metallic/magnetic thin-film device? Haven't run into it before. Intriguing.

In spite of my fixation with the semiconductor pickup, I welcome such posts as this.

Richard
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

Richard...

1. What is to be gained is an better understanding of why they do what they do, plus things like how to make visual the Mag Sus Window as a function of separation, mag circuit etc and correlation with performance as seen by the FSA.

2. Re power for the Semiconductor pickup = 3VDC or better. Very low Power. More stuff to be added that will be most convenient with on board battery power...It is the way the world will go including wireless to amps or computer etc.

2.5 Re Aural quality...methods 1 & 2 are to my taste. Lots more configurations to try.

Where is this going?...Comparison of the various pickups using the methods applied in the 32 PSG evaluation, except this time with recorded data so that it can be looped for FSA analysis. This way we will have the sounds, the FSA results, and a good timing reference = metronome. Set the metronome to 120 Beats per second = 1/2 second between clicks during playback, and plot the volume vs time for "sustain". MP3 or sim'...maybe video.

The recording will be via Sonar X1.

My other time consuming activity is mapping out the logic, and signal flow for the ADC/DSP/DAC chain. The math model, and macro controlled test is in Excel.

Edp
richard37066
Posts: 517
Joined: Sat Dec 11, 2010 11:44 am
Location: Gallatin, Tennessee, USA

Re: Pickup info

Post by richard37066 »

Ed -

Thanks.

Awaiting the next installment.

Richard
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

Before passing on to the actual signals from the BEAST and it’s assortment of pickup positions and configurations, let’s touch upon the various types of losses involved.

An excited (picked/plucked etc.) string vibrates in a magnetic field. The vibration disturbs the magnetic flux (flow). A coil positioned within the field has a current induced into it from the disturbed flux. A voltage is caused to exist at the ends of the coil that is a function of the velocity of the vibration, the amount of flux disturbed, the number of turns in the coil, the resonance of the coil, and the load placed across the coil. This is your signal.

Your PSG signal decays for a number of reasons. Overall it is energy loss…no perpetual motion here. The strings are under tension. They are physical filters with a reasonably high Q. The resonant frequencies are primarily determined by the strings’ material, length, diameter, and applied tension.

The string is conductive and moving in a magnetic field, therefore there is an amount of Eddy current generated in the string. Eddy current (Foucault current) generates some heat in the strings…also in the magnets. This heat is removed from the strings by their motion thru the air. This is a minor energy loss in the PSG. The higher the conductivity and the greater the field plus the higher the vibration frequency, the greater the Eddy current losses.

The excited string has a variety of harmonics in its vibration. These bend the string in various directions. This bending generates heat in the string…again mostly dissipated by the strings motion thru the air. Consider that the string must be longer when vibrating than when still. The string will tune higher when vibrating a greater distance (picked hard) than when approaching its static condition (still).

The strings will have a stiffness factor that is a function of the string material, diameter, and any wrap involved (wound strings). The shorter the vibration re string length, the quicker their vibration will decay…higher harmonics have shorter vibration length hence will decay first.

The magnetic field is applied to the strings in the window of susceptibility. Not only is this an Eddy current brake to stop the string from vibration, but it is an attraction (magnetic pull) on the string that also tends to stop the string from vibrating.

For those that think that Eddy currents and string magnetic attraction has an effect: Take your cents measuring tuner….tune the string in the presence of the magnetic field…then remove the magnetic field and see how many cents the string changes.

If the string’s vibration energy stays in the string it will vibrate for a longer time = more sustain. The vibration energy is either absorbed or reflected by the string termination mechanism. The termination material and configuration is a filter. It will reflect certain frequencies and absorb (pass) others. The absorbed energy will then attempt to excite the element to which is attached to self resonance this may be the changer, the nut, the key head, or other part(s). Each body (or other) part has their own preferred modal frequencies of vibration = self resonances. = more filters. The string vibration frequencies may or may not line up with the self resonances of the rest of the system (body, legs, et al).

Now for some more show and tell:

The next four Figures are taken to illustrate principle, NOT detail…No great care was taken re calibration etc The smoothed responses were not adjusted up 10 dB to restore actual peak values.

The actual values of the traces are about those shown +/- a dB. Each small division in Y is one dB. Six divisions up is twice the signal voltage…six down is half the signal voltage.

Two pickups have been shown mounted in the beast. They are tapped pickups. We have a selector switch with three positions = Left, Center, and Right to select taps, or combinations thereof. WE may use either the Danny Shields (DS) or the DiMarzio (DI) pickup by themselves, and in either the neck or bridge position…or we may use both at the same time with either in the neck and bridge position. As the Magnetic susceptibility windows for these pickups were quite different (previous photos) we might expect different performances if just for that reason.

Here is the response to a strum on fret 28 with the strings open (fret 0) using the DS pickup in the neck position by itself, and the DI pickup in the bridge position by itself. Compare these signals with the next Fig.
Fig #18
Image

Here is the response to a strum on fret 28 with the strings open (fret 0) using the DI pickup in the neck position by itself, and the DS pickup in the bridge position by itself.
Fig #19
Image

Here is the response to a strum on fret 28 with the strings open (fret 0) using the DS pickup in the neck position and the DI pickup in the bridge position. Compare the signals in this Fig with the next Fig.
Fig #20
Image

Here is the response to a strum on fret 28 with the strings open (fret 0) using the DS pickup in the neck position and the DI pickup in the bridge position
Fig #21
Image

Fig #22
Image

Many conclusions to come to, and many questions to be asked.

GOOD LUCK!!!!
richard37066
Posts: 517
Joined: Sat Dec 11, 2010 11:44 am
Location: Gallatin, Tennessee, USA

Re: Pickup info

Post by richard37066 »

Ed -

In previously published work, your traces were captured at intervals of one second, i.e., 1, 2, 3, etc. Ditto here?

Richard
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

Richard....not yet...the issue there was "sustain" = spectral decay. Here the issue (at this point) is how the pickups evaluated on a single instrument affect different spectral response shapes.

The next info will be the "E" string(s) single string response for the same pickups/positions. Then we will address the difference (or not) re sustain as a function of pickups and position.

Somewhere in there we will introduce the performance of the semiconductor pickup.

So we will have the performance of several pickup styles in several locations on the same instrument for several functions with data presented in graphic form (not opinions).
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

Although there is much more to be covered on the subject, I am going to drop out for a while. The reason is that the time will be spent focusing on the Semiconductor pickup/position sensor/ADC/DSP project. The pickup and non contact position sensor are just about ready for prime time. The basic blocks for the ADC and DSP just fell into place.

The PHOTOBUCKET site(s) listed above will remain available…with periodic additions for those interested.

Hasta

Edp
Post Reply