Innovative Synergies

Engineering Dimensionally Alternative Business Analysis 

Electro-Magnetic Guitar Pickups

Analysing Magnetic Fields

(Jan 2003)

Introduction

If the last section didn’t blow your mind, then this section will! This section pulls together much of what we have discussed and rolls this into some Finite Element Moment (FEM) Analysis so the magnetic fields can be "seen" to a very large degree - even if only in 2D.  

Having created a direct mathematical relationship between a couple of very easy to measure electronic measurements on an electromagnetic pickup and a desired timbre, this is a major breakthrough and it spells out what associated electronics is required for a guitar to get a particular spectral range – in other words the one pickup can be used for Cool Jazz, Cool Blues, Rhythm ‘n Blues, Country & Western, and so on – just be changing he load resistor and capacitor values! 

While this may be possible in most cases there are other mitigating factors that need to be considered, like background noise, and why it happens, how to get rid of it and why some picks just never seem to work!

Magnetic Fields Part 1

The typical electromagnetic guitar pickup consists of a permanent magnet with a coil of wire wrapped around this magnet, and the (magnetic) strings in near proximity.  When the (magnetic) string vibrates, it alters the magnetic circuit (specifically the reluctance – magnetic resistance) of the magnet and coil combination and because the permanent magnet’s magnetic potential field is changed, this induces a current flow in the coil of wire. 

The Finite Element Moments (FEM) picture above depicts a typical bar magnet – shown vertically as though you were looking through the end of the pickup - and a part of a string shown above it – shown horizontally.  This is the typical structure for the Strat style (single coil) pickups that have been measured in several different ways in earlier sections of this Website.  Any relative vertical movement of the string will alter the reluctance of the magnetic field. As the sting moves away from the magnet, this will cause a minute increase in magnetic reluctance, which will minutely decrease both the magnetic field potential (P), and the self inductance (L) of the coil/magnet assembly (and visa-versa).  

A Few Mathematical Transforms  

To get a grasp of this mathematically; where P is the Magnetic Potential (or Field Strength), and L is the total Self inductance, this calls for a little multivariable calculus to demonstrate the relationship of the voltage generated in relation to the string movement. 

V(t) = P dL/dt + L dP/dt

This is a very general equation, and it does not even have linear or angular dimensions included into it - but it definitely ties output voltage to changes in Magnetic Potential and changes in the total Inductance. Now the fun begins because we can now analyse the instantaneous voltage produced by the relative motion between the pickup and the vibrating string

As the string cyclically oscillates (normal or vertical to the plane of the pickup magnets), it cyclically changes the magnetic resistance (reluctance) of the pickup magnets' magnetic field, and cyclically changes the inductance of the pickup; and we now know that the maximum voltage (either positive or negative) is when the string is traversing the zero (or quiescent / stable)  condition.  Remember when playing on a swing, this is also near perfect sinusoidal motion, and the fastest point is when you go through the swing's resting position, but you come to a temporary halt at a forward limit and a backward limit - just like the guitar string!  So now we know why the maximum voltage is produced when the string is traversing the stable point.  

While doing this small signal analysis, we have several other conditions that are currently in our favour because these conditions can be considered as time invariant and therefore we can neglect them and the others can be considered as linear this makes small-signal analysis fairly easy.  

Just as a matter of gymnastics - look at the magnetic field above and realise that the string could have been placed below the pickup and the results would be identical.  Alternatively, the pickup could have been put in upside-down and it would work equally well.  It may not look that pretty - but it will work equally well (simply because the magnetic field is symmetrical about the centre of this magnetic structure)!

The first step that we need to do is to transform our time related maths into frequency related maths, then we can move from a single frequency to a linear spectrum, and then from a linear spectrum to a logarithmic spectrum (because that's the way our ears work)!

Changing from the time domain to the frequency domain is a little tricky - because in mathematics the frequency domain is in radians, and radians are rather poorly understood little creatures.  Basically if you take a circle and measure the diameter, then wrap the diameter around the circumference of the circle; then you will need about 3.14 something diameters to equal the circumference. This constant relationship between circumference and diameter of any circle is called Pi.   

We need this relationship so that we can relate a linear dimension (the diameter) with a frequency (rotational) dimension (the circumference).  As the radius is the distance from the centre of a circle to the circumference, this is a very nice dimension to use (and it also happens to be half the distance of the diameter)!  This circumference / radius relationship becomes 2* Pi, or about 6.28, and now we are getting somewhere although it may not be obvious - just yet - but look! 

We now have a circulatory relation to a linear function  - but it is in radians, and this is frustrating, and we need to make it practical, and this is the missing link.  If we roll the wheel (circle) a full circumference, we have traversed 2 * Pi radians (radius equivalents) to do a full cycle, so there are 2 * Pi radians per cycle. 

Now, if you subtend a right-angled triangle from a point an a circumference to a horizontal line passing through the centre and include the radius (of course), then the angle at the centre from the horizontal will follow the Sine rule as you rotate that point around the circumference, and in reference to the horizontal the point will trace out a sine wave (if the rotations are consistent with time).  So now we have the next link to tie a time related function to a rotational function. In a general form this link can be written as: 

Omega = Radians  per second

Omega = 2 * Pi * Frequency                      Where Frequency is Cycles per second (or Hertz (Hz))

We are now coming out of the forest and into the clearing, because Cycles per second has a more well known unit called the Hertz (after Heindrich Hertz)

f(t) = A(max) * Sin(Omega(t) + Ø)              Where Ø is an angular difference

f(t) = A(max) * Sin(2 * Pi * Frequency + Ø)          

or 

f(t) = A(max) * Sin(2 * Pi * F + Ø) 

Applied maths works in a funny way.  We firstly observe the situation and then fit various mathematical models until we get one or more that fit the approximations and then we refine it - and that is what we do here!  We know that basically the wave shape of a stringed instrument is a damped Sine wave, and this is a very good starting point.  We also know that the voltage generated is 90 degrees (or Pi/2) from zero degrees so we can use a Cosine function as the first approximation.

As 

V(t) = P dL/dt + L dP/dt

So

V(t) = P * A(Max) * Cos (2 Pi F) + L * B(Max) * Cos (2 Pi F)

There is a lot of common terms here so this can be grossly simplified to:

V(t) = K * Cos(2PiF)

Where K is a sensitivity constant (that is constant for small signals.)  For all those non-believers hook a guitar to a scope and gently thumb-pluck a string near the centre of its span and watch the near perfect sine wave for a short time.  Case almost closed.  

We know that the idle or steady state for a guitar string is when the string is not moving, and there is no sound being generated.  We also know that the instant the string is plucked it has its nominal maximum exertion form the idle state.  We also know that in nature all life decays exponentially - so it makes sense that after the string is plucked, it decays to its' stable state with a defined time constant that is directly related to the 'sustain' and this has a direct inverse relation to the energy lost in holding the string to the fret.  This is why open strings subtended by the bridge and machine head (nut) have a much longer sustain time than a string subtended by the bridge and a fret (by a finger).  The finger absorbs a small amount of energy that dulls the tone and considerably shortens the sustain time constant.  In fairly simple maths, a decaying step is described by:

Output (t) = Max * Exp(1 - t/Tc)      Where Tc is a known time constant

In putting these two equations together we have a rather simplified approximation for a damped sine wave from a plucked string:

 V(t) =  (Max * Exp(1 - t/Tc)) * K * Cos(2PiF)

There is little point in getting any deeper in this area as it will not prove anything substantial - but it does close the case for the first round of small signal analysis!

As the movement of the string is small, then the change in reluctance can be taken as linear with relative string positioning.  Virtually all guitar pickup analysis stops here (usually much earlier), but in reality, we are just starting to get a better appreciation of what is really going on - and now is the time to move on to slightly larger signal analysis.

That "Warm" Sound - from Large-Signal Analysis

If the string movement causes the change in reluctance to be non-linear, then the time-dependent voltage generated will also be non-linear with relative string positioning.  

Because the air gaps in these magnetic transducers are relatively big, and the relative string movements of fundamental mode vibrations are sometimes also relatively big, then for most fundamental string vibrations, there is a non-linear reluctance/movement relationship and the wave shape from the transducer in these cases will be non-linear.  

Because of the “distance squared” relationship between the magnetic potential field and the relatively moving string, the output will be a slightly tilted sine wave that looks more like a sawtooth than a clean sine wave.  This might be difficult to understand - until you realise that the output is greater while the string is deeper into the magnetic field and lesser as it is lesser into the magnetic field - and remember that the voltage is a direct product of the instantaneous relative movement of the string in the magnetic potential - not the instantaneous position of the string in the magnetic field. (But they are related!)

If you have ever wondered why an electric guitar often sounds like a violin then here is the answer:  The violin's bow, when drawn across the strings causes the strings to hop and recover on their resonant frequency - producing a pronounced sawtooth waveform.  This is very similar to that of an electric guitar - except that a violin has more timbre (is richer in harmonics). 

For those that are mathematically inquisitive, the Laplace (pronounced la-plarze) Transform of a sawtooth waveform consists entirely of even harmonics, and that sounds very warm.  I am not going to do the maths on this as it is available from a multitude of sources.  Like bees around honey - musos generally love warm notes.    

There are a few problems with putting the pickup very close to the strings in that the strings can be 'pulled' by the intrinsic magnetic field, and/or they can buzz when the extremities of the movements actually hit the pole faces in the pickups.  The 'pulling' of the strings causes non-harmonic notes to be generated that are slightly flat - and most musos hate these sounds!

So an electric guitar can and does sound like a violin but substantially without the abundance of warm even order harmonics that come from a violin being 'caressed' with a bow.  The musos want much more (even harmonics) from an electric guitar and this can be done in the amplifier with relative ease.  This is another topic and I will cover this in detail in another area - devoted to amplifiers - and what really happens in there!  

Interactive Issues

The problem with the simple pickup is that the magnetic field is very leaky and because of that, not only is the field very susceptible to external interference like florescent lights but the windings have to have an incredibly high number of turns to pick up the fluctuations in field strength, and this results in a relatively high self inductance and self resistance. 

Typically these single coil transducers have a self-inductance of about 3 Henrys and an internal resistance of about 6000 ohms – based on a typical 10,000 turn winding, and they produce typically 100 mV rms.  To get a normalised approach to this,
the Ar is about 6,000 / (10,000 ^2) = 60 u ohms,
the Al is about 3/(10,000^2) = 30 nH,
and the time constant is about 0.5 msec or seen another way, about 250 Hz.

As we now know that these transducers appear as a voltage source with a coil in series, then the upper frequency response knee can be set by the load resistance.  If 1 Mohm is used as a reference, then the upper knee frequency is R / (2*PI *L) or 159/L kHz.  In this case of 3 H it is about 53 kHz.  It is more common to have a 200 k ohm load so the knee is much closer to 10.6 kHz but in practice this will not happen because of excessive capacitance!

To compound the matter, the upper frequency response is limited by self-capacitance – typically about 70 pF – bringing the resonance to about 8 kHz.  With some cable capacitance this resonance can come down remarkably.  Say 1500 pF in cable leads and the self-resonance is 1.959 kHz and with a 0.022 uF ‘Tone’ capacitor, across the cable the self-resonance is back to about 600 Hz. 

Now with a little understanding on why these external; forces cause so many apparently interactive frequency  / spectral response issues it is not wonder that so many of these pickups are so ‘twiddle’ fussy – it takes a long time to ‘get the right setting with a guitar - even on stage! This is one reason why some pickups are ‘so ugly’ – and not the looks of them. 

Single Coil, 6 Rod Magnets - Magnetic Fields 

Looking at a standard six magnet single coil pickup as though we are looking across under the strings (x-z projection), we get Finite Element Moment analysis picture in 2D that shows the six magnets and the intense magnetic fields that leaks between them: 

                   

 

These pictures form an interesting scenario as the six magnets have a lot of fields lost between them in magnetic leakage.  It also indicates that even if the pickup were inverted then the sensitivity could be very similar to the top side.  This will help to explain why in general single coil pickups are very sensitive to external fields like fluorescent lights and SCR dimmed lights with nearby power cables.  

To all intents and purposes the fields look much stronger in the ends and much weaker in the centre - but this is an illusion caused by only looking at one slice which is right in the middle of the pickup.  If we were to twist the analysis by 90 degrees and look end-on through the middle of the pickup then we would see only one of the six magnets as they would all be in line, like the picture below:

  

This picture shows the cross-section of the Stratocaster or other single coil pickup and the picture on the right hand side shows just how sensitive the pickup is along the string.  By the 30 mm position off-centre from the pickup the field from the lower end of the magnet effectively cancels the detected string movement and as we move further, the sensitivity comes up to about 32 dB then gradually falls away with distance.  

Lateral Sensitivity

So now we know that the magnetic field strength has a direct bearing on the part of the vibrating string, and it is nominally limited to about 15 mm either side of the pole-piece centres. If we go back 90 degrees and look across the strings then the relative sensitivities are like the picture below:

The picture above is rather interesting as it shows the relative sensitivity of a range of pickups along the face of the pickups, where the strings are numbered 1 through to 6 and for thoroughness the spacing was extended outside the face area to include strings 0 and 7.  What has to be understood is that relative levels are not easily distinguished and a difference of 3 dB is virtually undetectable.  The outer-most strings appear to be a slightly less sensitive on all these pickups (be they single coil of humbucker) and surprisingly, these sensitivities are relatively consistent when normalised on the 4th string as a common point. 

We can now see the similarity of the pickup sensitivity compared with the string position and the magnetic field ‘slice’ through the centres of the six pole pieces.  (Imagine the strings are over the tops of the pole pieces!)  Where the field wraps itself around the ends, the sensitivity would be weaker for vertically moving strings, but for horizontally moving strings the sensitivity will be a little greater – and this may be a sensitivity/position compensating factor

Single Coil, 6 Iron Rods, 1 Flat Magnet - Magnetic Fields.

The next version of Stratocaster single coil design pickup types has a ferrite magnet and six soft iron pole pieces

 

In this case the (ferrite) magnet is sitting under the six soft iron pole pieces and the strings sit a few mm over the soft iron pole pieces.  The 2 Dimensional FEM analysis picture on the right hand side really needs some explanation to tell what is going on:  The six soft iron pole pieces are central to the picture in a row, and each rod does not appear to have their top ends closed.  The (ferrite magnet is sitting under the six soft iron poles and this magnet has the North Pole on the top (touching the six pole pieces) and the south pole is the bottom face

There are no strings shown in this picture but if they were to be shown then they would like six office pins sticking out of the picture just over each soft iron rod.  

What is confusing about this picture is that the magnetic field appears to be intense at the ends of the flat magnet and virtually no intensity in the middle.  The issue is that the 2D picture is going through the middle of the magnet and rods, and in the middle of the magnet there is very little static magnetic field.  The picture below shows the same pickup but from an end-on (90 degree change in view) perspective at the middle of the pickup: 

 

 

 

You can see here that there is an intense magnetic field around the edges of the flat bar magnet, and that the North pole (in this case) is extended up through the soft iron rod towards the string.  The big change in magnetic intensity is around the flat ferrite magnet, and there is not much change in magnetic intensity over the poles - where the strings are!  This magnetic lack of intensity at the required area is the reason why these types of magnetic pickups are not as sensitive as their six magnet rod cousins.  It would be a fair guess that these pickups probably also suffer more from external magnetic noise sources than their six-magnet cousins.

Single Coil, 6 Iron Rods, 1 Deep Magnet - Magnetic Fields.

As a variation on the thin flat ceramic magnet at the base of the pickup, a much deeper magnet was positioned in place.  Effectively this makes no difference at all as the magnetic field is now extended over the deeper magnet, and again the biggest magnetic differential is at the base of the pickup away from the strings so this will be rather insensitive as it's flat magnet cousin. 

 

Single Coil, 1 Deep Magnet - Magnetic Fields.

Another alternative is to look at a single magnet instead of several pole pieces.  To date most pickups seem to be stuck on the approach of a pole piece (or two) pole pieces for each string – and the approach seems questionable.  Take the case of a single long magnetic strip across six strings, the representation could look like this: 

        

 

If the North is the top-side and the South is the bottom-side (and it is easier seen if expanded by clicking on it)! On the left hand and right hand ends, the magnetic field is more intense as it wraps itself around the ends to make a circuit – this is happening all along the sides too – but this depiction does not show that - but this is shown in the right hand picture!  Compare this to the pictures above this one and it has to be appreciated that the 6 rod designs above have ‘very busy’ magnetic fields leaking back around the pole pieces – and not going anywhere near the strings.  If the length (seen here as height) is increased, then the field intensifies, making it appear stronger at an equivalent (string) position. 

This type of pickup should be comparatively sensitive and because of that, the coil windings could be minimised so the resistance and inductance would be reduced, making the connectivity to tone and volume controls that much easier, and provide a very wide range of tonal range.

 This would also explain why vertical hum buckers (Kinman variety) appear to be less sensitive (but remember the test apparatus is an electromagnetic field and not a moving string), than a multi-rod single coil variety.  

Single Coil, Two Lateral Magnets (P90) - Magnetic Fields

The P90 pickup is at the best an enigma to explain as it has two magnets in opposition to (facing) each other into a mild steel block and six iron screws to set the "sensitivity" for the strings to the pickup.  

This picture over-justifies the performance capability of the P90 as the centre mild steel square rod is not much thicker than the flat magnets, so the field does not really have that much "carry" into the strings.  This analysis does, however open up a large range of innovative ideas for restructuring pickups to be both efficient and sensitive.

Dual Coil, Lateral Magnet (Humbucker) - Magnetic Fields

Now onto hum bucker structures.  These have an entirely different structure, where the magnetic field is somewhat contained and it is this reason why they buck hum so well.  Or is it?  One approach is to use a common joining magnet as in US Patent 2896491 (Seth E Lover, 22 June 1955) and in US Patent 5399802 (Steven L Blucher 21 May 1995) and this has a magnetic representation like that below, and this field is symmetrical along the strings.  Structurally these two are very similar but the latter patent includes two soft iron strips to link the poles of the magnet to the adjustable pole-pieces.  To all other intents and purposes the magnetic fields are virtually identical as below:

        

This one takes a little explaining, so use the picture on the right hand side as a guide!  The magnet is lying down across the bottom-middle (North to the right, and South to the left) with two iron strip plates attached at each end of the elongated strip magnet.  In the two iron plates there are six vertically mounted iron screws with bit heads on them. (The heads are near the top of this representation.)  Above the screw heads are the strings, hence the ‘stretched’ magnetic field above the assembly.  The two coils sit ‘flat’ around the screws.  As the very large majority of the magnetic field is via the screws, and through the strings, this provides an efficient magnetic path with the strings – without unduly pulling them – and plenty of room for the two coils winding areas.  The bottom nipples are the extensions of the screw threads! 

As the permanent magnetic fields are in opposite directions through the coils facing the strings, this forms a relatively closed magnetic path, so any change in reluctance will cause an equal voltage to be created on each coil, and the coils are usually connected in series so the signal is additive.  As the magnetic field is somewhat constricted it is this advantage that provides the humbucking effect – hence the term ‘hum bucker’!  

Dual Coil, Single Vertical Magnet (Humbucker) - Magnetic Field

As an alternate to the lateral magnet some manufacturers came up with a similar design that only has one vertical magnet in the construction - together with two lateral coils.  These humbucker designs work surprisingly well considering the rather poor engineering involved in most guitar pickups. 

 

In this case the six bar magnets are mounted on the left hand side and a thin soft iron plate is the relatively high reluctance path to the right hand side where six soft iron rods form the return path towards the strings.  It is fairly obvious that the soft iron plate is close to magnetic saturation and the rods on the right hand side do not carry near the flux density of that in the six magnet rods on the left hand side.  There are two coils, each wound around the two sets of six co-linear rods.

There are some interesting points about this structure in that the magnetic field is far from symmetrical but in saying that, the stray fields - like those in the other lateral humbucker diagrams in comparison to the single coil designs - are comparatively well contained and therefore have the humbucker "tick of approval" - no matter how poor that is! Most of these humbucker designs are thrown together with plated soft iron so the magnetic leakage is often much worse that what is depicted here!  

If the soft iron plate is replaced by a thicker plate then the structure of the magnetic field takes on a more symmetrical shape:

The magnet retains much the same flux density, but the flux density in the now thicker plate is not saturated and the rods on the right hand side are now with a higher flux density, so this simple improvement should make a significant improvement in the sensitivity of this type of pickup. 

Dual Coil, Dual Vertical Magnet (Humbucker) - Magnetic Field

As another alternate to the lateral magnet some manufacturers came up with a similar design that only has two vertical magnets in the construction - together with two lateral coils.  These humbucker designs also work surprisingly well considering the rather poor engineering involved in most guitar pickups. To get an understanding of this structure, here are two magnets (one north at the top the other South at the top) near a string:

This is not much different that a pair of single magnet pickups put beside one another.  The magnetic fields are interacting and pulling into each other to dramatically reduce the amount of external interference.  Once a thin iron sheet is placed to join these fields then the large majority of the magnetic fields passes through the soft iron to again dramatically reduce the external magnetic field. 

As shown here the vast majority of the stray field is captured and linked to the adjacent rod magnets leaving very little external fields other than that near the strings. 

Note there are three areas where the strings are drawn to the magnets (the low magnetic areas are just outside the line of the magnets). 

Lateral / Humbucker Sensitivity

The above is a representation of the y-z plane through the middle of a typical hum bucker pickup coil arrangement.  There are a few very obvious differences with this and the single coil structure.  There are two magnetic structures relatively close to the strings (on top), and with the soft iron joining plate (across the bottom) holding both the magnets (left hand side) and soft iron pole pieces (right hand side) this makes a relatively ‘closed’ magnetic circuit that has a far more limited extraneous magnetic field than the simple ‘single coil’ approach.  This is another reason why this structure is a much better hum bucker than a single coil style pickup.  The magnetic field is virtually symmetrical, but with pickup coils on both pole pieces, variations in magnetic field strength (flux) caused by relative (predominantly vertical) string movement will induce onto the coils. 

Looking along the strings for sensitivity measurements it is obvious that the sensitivity falls off after about 20 mm from the centre of the structure (between the two pickups), so we know that the pickup will react to strings near the pole pieces/magnets – in a very similar fashion to the single coil pickup – but over a slightly wider length.  

This is very similar to the Hum Buck 01 pickup used in these experiments and it has series connected coils giving a total resistance of about 8.45 k ohms (dc) and a total inductance of about 4.59 H at 1 kHz.  With the coils connected in series aiding, this pickup will have a smooth (or muffled) sound suitable for cool Jazz or Blues.  Alternatively if one pickup coil is used by itself, it will have a response similar to that of a single coil pickup, but because of the extra ‘metalwork’ the inductance is marginally greater, so it will still have a slight muffled sound.  By putting the second coil in additive parallel, the external field will be bucked, the internal resistance will be halved, and the internal inductance dropped by about a third, and this should be a very lively pickup. 

Comparative Sensitivity

Looking across the fret-board with different pickups all placed nominally 10 mm from the exciter coil, the following graph gives an interesting picture of relative sensitivities of various pickups.  

 

This is very similar to an earlier graph, but not normalised at the maximum points, and it generally shows that longer magnets end up with a more sensitive result - not taking into account the number of turns!  Considering the wide manufacturing approaches, it is surprising that most pickups fit into a very common area, and the overall difference in sensitivities is about 9 dB, and when considered that the distance from string to magnet is a critical factor for absolute level, then these differences could well be less than 6 dB.  Each 'family' of pickups have curves that very closely align.

Going one step further by looking along the strings with different families of pickups, another set of non-normalised curves emerge as below:

This is particularly interesting as the Strat, and the hum bucker wired as a single (HMB01S) form a nice pair with the extra metalwork of the hum bucker putting a 'bump' on one side of the sensitive curve (caused by the second pole piece).  The two 'notches' are caused by the fields interacting as they go through an angle where the phase reverses. 

The second pair of curves is particularly interesting as these are the hum bucker connected as a series (HMB01SS) and as a parallel (HMB01PP).  Both have exactly the same sensitivity as each other, and also about 8 dB less than the single arrangement.  Effectively, the along string sensitivities agree with each other, and considering that this is near the noise floor for this test arrangement, the non-alignment near the null on the left hand side is barely a concern.  What is interesting is that the maximum levels are virtually identical and with a bit of appropriate engineering, switching in the buck coil as a parallel will give the Country & Western 'twang' as this will have half the internal resistance and about two thirds the inductance of the single coil.  

A little Myth that needs exploring:  If it is not already obvious, the two side-by-side coils in a Hum Buck arrangement are not highly coupled, and in fact they are rather loosely coupled to each other and putting them in parallel does not quarter the inductance as you would expect if they were tightly coupled. The 'twang' comes from the internal inductance being lower in relation to the volume pot to how it was before and this should be a strong sign that output frequency response is heavily controlled by the load on the pickup - including the cable and amplifier input impedances.   

Conclusions

By performing finite element analysis on a range of pickups, this has shown that there is a high degree of correlation between the measured results for relative level and the magnetic structure.  It is not the absolute magnetic field strength that matters but more the change in magnetic field strength with the movement of the vibrating string that is the prime cause for producing voltage from the coil. 

The position and path of the magnetic field therefore is of a prime importance and this explains why 'Strat' pickups with longer magnets seem to be more sensitive than 'Strat' pickups with shorter magnets - as the longer bar magnets appear to be relatively closer to the strings - because of the shape/extension of the polar fields from the bar magnets. 

The ferrite bar single coil type pickups - with their soft iron pole pieces make up what is effectively 'longer' magnets than the simple bars and although these pickups seem to have a generally slightly higher inductance; these pickups appear to be more sensitive than the simple 'Strat' design.

The Dual (Lateral) Coil  hum bucker structures with the magnet centrally located and with pole-pieces to guide the magnetic path through the strings have a naturally far more efficient magnetic path than the single coil 'Strat' and 'Ferrite bar with pole pieces', and it is this structure that also assists with hum bucking, as the stray magnetic circuit is considerably circumvented by the much tighter magnetic circuit.  Because these magnetic fields are somewhat enclosed, this is probably the reason why these pickups in dual coil mode appear to be less sensitive than the single coil pickups by about 8 dB.  The Hum Bucker pickup in single coil mode operates almost identically to the 'Strat' and Ferrite based single coil pickups, and it is because of the extra pole pieces in the second coil of the Hum Bucker assembly do create a secondary magnetic circuit, that the 'along the string' figures do not match exactly the 'Strat'  and Ferrite single coil structures. 

The dual coil Hum Bucker when connected in Series (aiding) and in Parallel (aiding) both came out with the same level (terminated in 1 M ohm), and this lends itself for considerable further development. Not only is the level consistent, but the 'along the line' suppression very consistent, and that leaves external impedances (volume, tone controls, cables etc) as major players to 'colour the sound' from these pickups.  With series aiding connected coils, these pickups tend towards the Jazz/Blues dull/muffled sound, and with parallel (aiding) connected coils, these same pickups lend towards the rock/country sharp/twangy sound, and all this can be dome by changing the load resistance and capacitance to pre-known conditions - removing the need to excessive adjustments - without a reference. 

There are several other magnetic structures that could have been tested, but it seems that most manufacturers seem to be stuck on the rule of one or two pole pieces per string gives an aesthetically pleasing product - and that is what sells! 

The very simple tests done here have shown that with a little structured analysis, pickups could be manufactured to be far more sensitive that they currently are and that the need for external hum-bucking could be considerably reduced.  With this in mind, it might be worth looking at the polar sensitivities of various pickup assemblies, and better understand just how sensitive these pickups are to external electro-magnetic fields.

 

Copyright © Malcolm Moore, 2003.   Comments and Corrections are welcome