Modeling an electric guitar with LTSpice

[antigua]

Here in the future, electrical circuits can be simulated with software before anyone ever has to worry about the practical matter of how they will be made. As electric guitars are literally electrical circuits, they too can be simulated with software.

Guitarists spend a lot of time experimenting with different pot and capacitor values, and even spend a lot of time and money trying out various guitar pickups. To a very large extent, a lot of time and money can be saved by modelling these circuit modifications through software. The trick is just knowing where to start, and that's the purpose of this post.

First of all, what software to use? Software circuit simulators are called "EDA's". My EDA of choice is LTSpice, which can be downloaded here. It's free and has existed for decades. It's user interface will take you all the way back to Windows 95.

This forum's own JohnH has created a dynamic spreadsheet called GuitarFreak that offers this same sort of modelling in a convenient, interactive way. The major difference is that his dynamic spread sheet already has the models built, you just plug in the values and choose options, whereas this walk through will have you making a model from scratch, using free EDA software. 


Pickup Theory

So where do we start? It's all begins with the guitar pickup. For all the magic that supposedly goes into a guitar pickup, the truth is that they are cold, unfeeling "low pass filters with resonance". It's the first and most essentially electrical component in an electric guitar, so the first thing to do is model a pickup.

Helmuth Lemme published a useful representation of a guitar pickup, it can be seen here:


This representation proves to work well for modeling single coil pickups with AlNiCo rods, though for humbuckers and other pickups with steel parts, the model has to be expanded slightly, but that will be detailed later.

To quickly explain what these things represent to the pickup:

The AC source is the moving guitar string. Through Faraday' law of induction, the magnetized string creates an alternating current in the pickup coil.

The "L", or the inductor, is the inductance that comes from the pickup's coil. The pickup is primarily an inductor, and the several henries of inductance that most pickups produce is considered a high value of inductance.

The "C" is the capacitance that occurs due to the close proximity of the wire in the coil. This unwanted capacitance is a phenomena that inflicts nearly all electrical components to one degree or another. When there is more space between the wire, this unwanted capacitance is reduced. Thicker insulation coats on the wire, or a more scattered coil, will create such space, and therefore reduce the capacitance. In reality, it's like there is an infinite number of little capacitors along the length of the coil's wire, but it so happens that we can define it theoretically as a single parallel capacitor.

The "R" is series resistance, defined by the conductivity and the gauge of the copper wire. If the wire were silver, the resistance would be lower. If the wire is thicker, the resistance will be lower.




Getting started

- Download and install LTSpice
- Open LTSpice

- Click File > New Schematic


This next step is an optional, quality of life issue. This is what LTSpice looks like with it's default colors.



I find this hard on the eyes. To improve the color scheme...

- Click Tools >  Color Preferences
- Select "Background"
- Move the Red/Green/Blue sliders all the way right to make the background white, or whatever you like. 
- Click OK

Then to make the wires and plot lines bold...

- Click Tools > Control Panel > "Drafting Options" tab
- Check "Draft with thick lines"
- Click OK

It should look something like this:







Creating a Pickup

The first thing to do is create the AC source, the (~) to the left.

Note that "undo" is the "F9" key.

- Click Edit > Component
- Select "Voltage" from the menu
- Click "OK"
- Place the "voltage" somewhere to the left:



- Right-click on the new "voltage" component
- Click "Advanced"
- Set "AC Amplitude" to "0.2"

This AC source essentially represents a moving guitar string. Next, the coil must be represented with an inductor, a capacitor and a resistor.

- Click Edit > "Draw wire"
- Draw a wire connecting to the ends of the AC source like this, and one wire in parallel to the AC source, like this:





Note: use "Edit > Move" to rearrange elements, and the magnifying glass icons to zoom in or out.

- Click Edit > Inductor, then click ctrl+r to rotate the inductor horizontally.
- Place the inductor along the top wire as seen below.

- Click Edit > Resistor, then click ctrl+r to rotate the inductor horizontally.
- Place the resistor along the top wire as seen below.

- Click Edit > Capacitor
- Place the capacitor on the parallel wire.

The final requirement is to specify a "ground".

- Click Edit > "Place GND"
- Put the component at the far right end of the lower wire.





The next step is to give these components values that are realistic to a guitar pickup. I'll state values that are typical of a Stratocaster pickup.

- Right-click the inductor
- Enter "2.2" in the Inductance(H) input, click"OK" (2.2 henries)

- Right click the resistor
- Enter "5.8k" in the "Resistance" input, click "OK" (5.8 kilohms)

- Right click the capacitor
- Enter "120p" in the "Capacitance" input, click "OK" (120p = 120 picofarads)


All the values are now set:





We now have a model of a Strat type pickup, an RLC (resistance inductance capacitance) low pass filter, that matches the circuit diagram provided by Helmuth Lemme.

In order to simulate this circuit, we need one more thing, a "input impedance", representing the guitar amplifier, or test equipment, and which serves as a "point of view" when evaluating the output voltage by frequency.

- Click Edit > Draw wire, and draw a wire parallel to the capacitor and AC source, far right, as seen below
- Click Edit > Resistor, place the resistor on the new wire
- Right click the new resistor, enter "1000k" int the Resistance input

Here is where we're at:




With an "input impedance" in place, we can now simulate the frequency response of the pickup (RLC circuit).

- Click "Simulate" > "Edit Simulation Cmd"
- Select the "AC Sweep" Tab
- Enter "100" in the "Number of points per octave"
- Enter "1" in the "Start Frequency" input
- Enter "20000" in the "Stop Frequency" input
- Click "OK"
- Place the text box anywhere in the schematic, here I have it at the bottom:



The text ".ac oct 100 1 20000" is an LTSpice directive, which tells LTSpice what sort of test conditions we want to put the circuit through. We want to test most of the audible range, from 1Hz to 15,000Hz, in 100Hz increments.

- Click Simulate > Run

When you first run the simulation, you will see an empty black window appear, and it will look something like this.





- Resize the windows to be taller and/or wider
- Click Window > Tile Horizontally
- Click on the "1000k resistor", this will cause the upper window to show a voltage-by-frequency plot from this location within the circuit

At this point, a plot should appear:



The black plot is now showing two things:

1) a solid green line, showing the voltage seen at the "input" over a range of frequencies
2) a dashed green line, showing the phase of the voltage and the current

For tonal analysis, only the solid green "amplitude" line is needed. To get rid of the phase plot, click on the dB scale to the right of the plot, a popup window will appear, click the button "Don't plot phase", and the dashed green line will disappear:



This shows us something interesting in it's own right, as this represents the EQ profile of a guitar pickup with 2.2 henries inductance, 5.8k resistance and 120 picofarads capacitance, which is very close to the electrical values of the Fender 57/62.

This curve arises from is the fact that, below a particular frequency, the inductance wins out over the capacitance, and then at that particular frequency, the capacitance and inductance interact with each other, or "resonate", and create a greater impedance than an ideal inductor would achieve by itself. Then, above that frequency, the parasitic capacitance takes over, and the inductor is no longer able to provide an inductance, so the impedance, and therefore voltage, drop off thereafter.  Check out this web page regarding low pass filters with resonance for more insight into what makes this particular response curve interesting, and defining of "electric guitar tone".  An electric guitar's tone is characterized by the fact that it's harmonics are boosted slightly at that resonant frequency, and then almost completely chopped off above that point, as it falls quickly out of audible range. This is why "50's guitar technology" persists to this day, because that particular technology inadvertently gave rise to a tonal aesthetic which everyone has come to know and love. 

If you place the cursor over the peak, you can see that the resonant peak calculates out to 9.6kHz. A fun thing about this is that you can now change the inductance and capacitance values, and get a feel for how much the resonant peak moves with a pickup has 3 henries of inductance, or twice as much capacitance, etc.



Modelling a Guitar Rig

We're not done yet though. You never hear a guitar pickup with a frequency curve like this for two big reasons: 1) the volume and tone pot add parallel resistance in relation to the pickup, about 125 kilohms total (two 250k ohm pots in parallel), and 2) the guitar cable adds parallel capacitance, which might land around 500 picofarads, though this varies widely from one guitar cable to the next.

In order to simulate a real rig, we have to add fake pots and a guitar cable.

- Return focus to the circuit
- Zoom out to make more room
- Click Edit > Draw Wire and extend the lead wire out to the right
- Click Edit > Move, draw a box around the "ground" symbol and move it to the end of the wire:




- Click Edit > Move, draw a box around the wire and 1000k resistor, and move both simultaneously to the far right:



You now have space for more parallel components in between the "pickup" and the "guitar amp"

- Click Edit > Draw Wire, draw three evenly spaced parallel wires in between the "pickup" and the "guitar amp"




Next, we add representative components

- Click Edit > Resistor, and place resistors on the first two new wires.
- Right click the new resistors, give them values of "250k" each (these represent 250K ohm volume and tone potentiometers)
- Click Edit > Capacitor, and place one capacitor on the third new wire.
- Right click the capacitor, give it a value of "500p" (this represents a typical guitar cable)



At this point, there are enough components that they can become confused with one another. The names "L1", "R1" etc. can be renamed by right clicking on the name, and entering a new name. The names have to be alphanumeric, which includes the underscore, or "_". Convention is to use an underscore in place of spaces, which are not allowed.

Here are names I chose for the components:




With LTSpice you can also add generic labels wherever they might be helpful, by clicking Edit > Text.

I've added some labels:





Now when we run this simulation, it should return a plot that more realistically represents what we hear in real life.

Repeating the simulation steps above produces this plot:



Placing the cursor over the peak reveals that the resonant peak is 3.95kHz, and that the resonant peak rises about 5dB ahead of the roll off. This is a realistic value, a Fender 57/62 pickup chops off everything right around 4kHz, when you have them wired into a typical Strat with 250K ohm pots, using a typical guitar cable.

If you want to model a Les Paul, you can actually model two coils (and eddy current losses) for a highly accurate model, but if you want something that is useful without all the hassle, you can simply assign typical PAF values to the single coil pickup model. That means 4 to 5 henries of inductance, 7.5k to 8.5k DC resistance, and about the same 120pF capacitance (note that the capacitance is less critical), and then by changing the tone and volume pot values to 500k each. This will give you reasonably accurate resonant peaks, though the Q factor (how tall the peak is) will be higher than in reality due to the fact that the steel parts, and the nickel silver cover on a typical PAF causes eddy currents, which reduce the height of the Q factor. Where as it's seen above that the Strat pickup rises about 5dB at the resonance, a PAF pickup will only rise above 1dB to 2dB covered, and 2dB to 3dB with the cover removed. Those eddy current losses can be modeled as well, using an eddy current model that I will detail in a later post (it involves an inductor and a resistor in series, parallel to the pickup).





Real-life results compared to modeling

There is an important difference to be aware of between modeling and real-life results. In this modeling, the AC voltage is a consistent amplitude throughout the sweep of tested frequencies, but since a guitar string "induces" current through magnetic induction, and inductors cause impedance, and impedance increasingly "impedes" as frequency rises, there is a +6dBV per octave impedance slope in real life, that instead looks perfectly horizontal in these fixed-amplitude simulations. This means the voltage is not flat as it appear in the model, but actually slopes upwards with frequency. 

The oscilloscope generated graph below shows both types of plot; the "raw" gray plot which starts very low, and slopes towards the resonance demonstrates the +6dB/octave transfer of a guitar pickup, while the "integrated" red plot line has been compensated (using a device designed by Ken Willmott) with a -6dB per octave filter in order to flatten it out. This "flat" plot is how the LTSpice models will portray the circuit. In real life, that +6dB slope exists, and it has the effect of greatly decreasing the bass response, and increasing treble response, relative to, say, a piezo pickup, which has a flat response by nature.





It is actually worthwhile to consider the transfer characteristic of pickups with that +6dB/octave slope removed, as it is with the red line, and as it is in the LTSpice model, because it allows us to see the equalization of a pickup in isolation. Every electric guitar pickup will exhibit this +6dB/octave slope, and since it is not unique to a given pickup, the slope only serves to obfuscate what makes one pickup, or wiring scheme, different from the next. In other words, by extracting the +6dB slope, we can more easily see what makes a pickup different from another. It is also easier to identify things like eddy currents, because the losses will contrast against a horizontal plane, rather than a sloped angle. 

It would be nice if LTSpice allowed for a voltage that increased or decreased with test frequency, so that we could model both types of plots, integrated and non-integrated, but LTSpice does not support it.
 


As far as modeling accuracy goes, just as a matter of demonstration, the plot below was made using an actual Fender 57/62 pickup set. The higher "unloaded" peaks correspond to the first model, where the guitar cable, tone and volume pot was left out, and the lower "loaded" plots correspond to the modelling that did include those things. There are three overlapping plot lines for all three pickups in the set. The values shows in the plot below vary slightly from the modeled values, as the modeled values were a little different from the actual values, but JohnH has shown on a number of occasions that model used here curve matches actual measurements to an extremely high degree, especially when the eddy current damping is added in, which will be detailed in the next post.




Here is an example where JohnH overlayed his calculated curves modeled through GuitarFreak with the actual measured bode plot:


Based on the extent to which these curves overlap, you can get a sense of how well the modeling approximates a pickup's actual electrical performance.  They say that humans can only discern volume differences greater than 1dB in amplitude. The vertical scale is 1dB per division, and you can see that the deviation is hardly a fraction of a decibel, so there is good cause to have faith in pickup modeling, as an alternative to doing things the hard way, with a soldering iron, and a junk box of assorted components and values.






Dynamically altering component values...


Once you have created a model of a pickup, or an electric guitar, a powerful capability you now have is to compare multiple component values, and see the results side by side.

For example, suppose you want to see how much the resonant peak increases in amplitude when you change the 250k tone and volume pot to 300k, 500k or 1 meg. The trick is to change the values of the components whose values will change in a "variable", meaning the value will become several different values during the simulation. You would then use the LTSpice "step" command to specify those different values: 250k, 300k, 500k and 1 meg.

- Right click on the value of the tone and volume resistors, change the values to {POT_VAL} in both resistors.
- Click Edit > SPICE Directive, and enter this text:

   .step param POT_VAL list 250K 300K 500K 1meg
  
- Click "OK", and place the text somewhere

As you can see, "POT_VAL" is the name of the variable, and the SPICE directive will cause the simulation to "step" through these values.

- Click Simulate > "Run"

- Click on the 1000k "imp" resistor at the far right side.

- Click on the "phase" scale on the right side of the plot and click "Don't plot phase"




You now see several plots, each with different colors. The bright green trace represents the first value, 250k, and it has the lowest peak. The highest peak corresponds to the 1 meg value, and so this is visual and mathematical proof that using 1meg pots will make a guitar brighter than 250k pots.

If you click Tools > Color Preferences, you can view and edit the trace colors:








Dynamically altering multiples component values...

Often times you want to edit two or more parameters at once.

Suppose you want to simulate how well a treble bleed capacitor value will retain treble as you turn down the volume knob. The problem is, a potentiometer is like two resistor value changing in tandem; the portion to the left of the wiper, and the portion to the right (voltage divider setup). The step command as shown above only provided a single value, but we need two values in order to model the two sides of the volume pot at different volume settings.

The solution is an LTSpice "list" and it's "table" feature. Suppose you name your resistor variables "POT_UPPER" and "POT_LOWER" to represent the two halves of the potentiometer, and you want to plot six different points along the volume knob turn, with a different division of resistance at each point, the command would look like this:

  .step param COUNT list 1 2 3 4 5 6
  .param POT_UPPER table( COUNT, 1,1, 2,50k, 3,100k, 4,150k, 5,200k, 6,240k )
  .param POT_LOWER table( COUNT, 1,250K, 2,200k, 3,150k, 4,100k, 5,50k, 6,10k )

The first line says that you want to step through six values, using a variable called "COUNT".

The second line says you want to create a table that maps all six steps of COUNT to specific values. These values will represent various resistances on one side of the pot, and assign those values to "POT_UPPER"

The third line does the same thing as line two, but this time the values are inverted, in order to represent the other side of the pot, and those values are assigned to "POT_LOWER".

Note that zero can't be used as a resistor value in SPICE, so "1 ohm" is used as a stand-in value.

- Click Edit > Resistor, ctrl+r to rotate it flat, and place it beside the "vol_pot" resistor.
- Right click the resistor, name it "vol_pot_up" and set it's value to {POT_UPPER}
- Rename the existing "vol_pot" to "vol_pot_down", and set its value to {POT_LOWER}

It should look like about this:


- Click Edit > SPICE Directive, enter the list and table directives above, click OK and place it somwhere.
- Click Simulate > "Run"
- Click on the 1000k resistor at the far right side.
- Click on the "phase" scale on the right side of the plot and click "Don't plot phase"

The result should look like this:




You can see that the step and tables directive is below the circuit, the volume pot is represented by two resistors, and the plot shows how the frequency varies as the voltage is divided between the pickup and ground. The green line is the first step, where most of the voltage from the pickup is sent to the amp, this is "10" on the volume dial. You can see a bump at the resonance, but notice that the bump disappears in all but the lowest plot line, which correlates with the heard experience of the volume pot diminishing the treble response of the pickup.

So let's add a treble bleed capacitor and see what happens..

Click Edit > Draw Wire, and draw a wire up and over the "vol_pot_up" as shown in the circuit below.
Click Edit > Capacitor, ctrl+r to rotate the capacitor, and place it on the new wire
Right-click the capacitor and name it "treble_bleed" and give it a value of 500p (500 picofarads)
Click Simulate > Run, and click on the far-right 1000k resistor.

The result should look like this:



Notice now that not only is there a high peak at the resonance for all of the different plots lines, but some of them are even higher than the green peak "at 10 on the dial". This demonstrates visually and mathematically what a treble bleed is doing in order to retain treble throughout the sweep of the dial. The fact that the peaks vary in size is a testament to the fact that treble bleeds can be hard to perfect. Different capacitor values and different combinations of capacitors and resistors in series and parallel will yield difference curves. In general, you would want values that roughly yield the same shape of peak for every plot line. Notice how, with this 500p treble bleed, just before you hit "zero" on the volume, there is a lot of treble coming through, so this cap value might cause the lower portion of the sweep to sound too bright.

If you are able to guess within reasonable certainty the electrical values on your pickups, you can use this dynamic modelling to preview the dynamic behavior of different treble bleed values, or tone and vol pot values, or different tone capacitor values. Even if you don't know the exact pickup values, you can assume 2.5 henries for a Strat pickup, 5 henries for a PAF type humbucker, 100p capacitance for each, and the resistance is usually easy to determine, and from those general values, you can figure out what values or wiring scheme is right for you.




Modeling a humbucker wired in series...


To model a humbucker in series, all you have to do duplicate the "pickup" portion of the circuit, attack them together with "wire", and connect the ends to the rest of the circuit. Here is what this looks like:



This resulting plot shows a resonant peak of 3kHz, which is realistic for a typical PAF neck humbucker.

The step taken to get here from a simple single coils schematic were

- Click Edit > Move and grab everything left of "vol pot", and drag it upwards.
- Click Edit > Duplicate, draw a box around the pickup, and move the mouse down below the pickup, and click again to place a duplicate copy of the pickup below the first pickup.
- Click Edit > Draw wire, and connect the two pickups/coil to each other in series, and the other end of each to the other half of the circuit, as shown.

The two pickups now represent the screw and the slug coil of a humbucker, so click to rename all the components "screw" and "slug" as seen in the screen shot.

The resistance values should be reduced to 3.5k to 4.5k, or half of whatever actual humbucker's resistance comes out to. Inductor values of 2.2 henries are still realistic for a PAF type coil, because the steel slugs and screws give them a higher inductance for a smaller coil. The capacitance should be about half for each coil, so I set that to 60p for each. I also set the vol and tone pots to 500k to mimick a Less Paul wiring scheme. Note that no tone cap is shown in the diagram since it has almost no effect where there is such a high resistance in series with it.

This diagram is very useful for previewing the effects of coil splitting, and trying out exotic splitting schemes, such as splitting with a capacitor:



In the diagram above, you can see that I've added a 4 nanofarad capacitor called "split_cap", with one end between the two coils, and the other end to ground, and the result is an interesting "mid scoop" or "band stop" at 2.3kHz, which makes for an interesting new tone. Changing the cap value moves the band stop, and this modelling is very helpful in determining which capacitor values will have which effect for a pickup of given values.





In conclusion...

Pretty much any crazy wiring scheme can be evaluated with LTSpice before you commit to the time and cost of actually rigging in real life. The draw back here of course is that you can't hear a visual frequency plot, but over time you will come to relate what you see to what you have heard, and know what you are hearing by sight, and vice versa.

You can also vastly improve the accuracy of the modelling if you are able to determine the exact electrical values, such as the resistance and inductance of the pickup, the capacitance of the guitar cable, and the input impedance of your guitar amp. The capacitance of the pickup can be determined through various means, but more often than not, the guitar cable's capacitance will be so high as to overpower the capacitance of the pickup. In general, Stratocaster and Tele bridge pickups have capacitance values between 90pF and 140pF, while Tele neck pickups have capacitances closer to 200pF. 

[Deleted]

I remember spice from Univ, back in the 80s. Sir, hats off to you. Keep up the great work.

[newey]

Antigua-

In reality, it's like there is an infinite number of little capacitors along the length of the coil's wire, but it so happens that we can define it theoretically as a single parallel resistor.

I assume you mean "capacitor", not "resistor". Since this is a step-by-step tutorial for the uninitiated, typos should probably be corrected for clarity.

But, yes, great work! I won't move this thread, as it does relate to pickups certainly, but I think you should probably repost this in the "References" section as well.

[JohnH]

Yes this is great stuff indeed. When I get home, Ill chip in about the different free modelling program that I use, which is 5Spice. It has as interface that is more 'noughties than 'nineties.

Also, there is heaps to add about more explicit modelling to match test data, but in the meantime the GuitatFreak post has further info.

[antigua]

Dec 16, 2016 6:02:29 GMT -5 newey said:

I assume you mean "capacitor", not "resistor". Since this is a step-by-step tutorial for the uninitiated, typos should probably be corrected for clarity.

Thanks, fixed. 

[ms]

An alternative to using spice:

Programming languages are quite powerful and easy to use now.  For example Python can be used in an interactive mode, like Matlab, for example, and arrays, single numbers, real and complex, all can be mixed in expressions.  This makes it really easy.

Once in interactive Python, the six commands below are all it takes to get the attached plot.




The first line brings in the file pupc, which has the code shown at the end.
The second line gives the little project a name, (ap stands for Antigua's pickup) and defines the start frequency, the end frequency, and the number of frequency points.  No need to worry about the frequency after that. Then we need the serial and shunt impedances to put into the method for finding the response of a voltage divider.  the values come from Antigua's plot just below "Getting Started".  fZL finds the impedance pf an inductor vs. frequency given its inductance; likewise for fZC for a cap.  vdiv finds the response using its input as the series and shunt impedances.  Then it is plotted.

In [1]: import pupc

In [2]: ap = pupc.pupc(20., 20.e3, 500)

In [3]: Zser = ap.fZL(2.2) + 5800.

In [4]: Zshu = ap.par(ap.fZC(120.e-12 + 500.e-12), 250.e3, 250.e3, 1.e6)

In [5]: res = ap.vdiv(Zser, Zshu)

In [6]: ap.plres(res, 'Antigua\'s pickup')

This is the code in pupc that is imported above:

from matplotlib.pyplot import *
import numpy as np
import time


# Starting and ending frequency and number of frequencies
class pupc():
    def __init__(self, fs, fe, fnum):
        self.f = np.linspace(fs, fe, num = fnum)

# Method to return impedance of a capacitor
    def fZC(self, C):
        return -1.j/(2.*np.pi*self.f*C)

# Method to return impedance of an inductor
    def fZL(self, L):
        return 1.j*2*np.pi*self.f*L

# Method to parallel 2 or more impedances
    def par(self, Zf, *Zs):
        admit = 1./Zf
        for Z in Zs:
            admit += 1./Z
       return 1./admit

# Method to find the response of a voltage divider
    def vdiv(self, Zser, Zshu):
        return Zshu/(Zser + Zshu)

# Method to make a plot of the results of vdiv
    def plres(self, res, str):
        semilogx(self.f, 20.*np.log10(np.absolute(res)))
        grid()
        xlabel('Frequency (Hz)')
        ylabel('Response (0 = unity gain)')
        title('{0:s} {1:s}'.format(str, time.ctime()))

[antigua]

Thanks for detailing exactly how that's done. Did you use Matlab? One thing that has held me up is the cost and licensing for Matlab. Even supposing it's $150, which isn't insignificant, just having to email them to get a quote, and figure out the license, and worrying about the fact that the license is term limited, is kind of stressful when you're capacity is as a hobbyist. I love free things more for the fact that they're convenient, even more than the lack of cost. 

I can see how the scripting approach would simplify automation, and permit the +6dB slope in the AC source, and all sorts of other things that would go beyond the capabilities native LTSpice, and SPICE in general. I think the real sell with LTSpice, though, is the ability to drag and drop components, and not even have to be aware of how they components and their values mathematically relate, or even a scripting language, aside from some a handful SPICE directives. That, on top of the fact that it's free, greatly reduces the barriers to entry. So if we were ranking the options by of "ease of use" versus power, I'd go GuitarFreak, then LTSPice, then interactive Python. 

[ms]

Dec 16, 2016 16:03:48 GMT -5 antigua said:

Thanks for detailing exactly how that's done. Did you use Matlab? One thing that has held me up is the cost and licensing for Matlab. Even supposing it's $150, which isn't insignificant, just having to email them to get a quote, and figure out the license, and worrying about the fact that the license is term limited, is kind of stressful when you're capacity is as a hobbyist. I love free things more for the fact that they're convenient, even more than the lack of cost. 

I can see how the scripting approach would simplify automation, and permit the +6dB slope in the AC source, and all sorts of other things that would go beyond the capabilities native LTSpice, and SPICE in general. I think the real sell with LTSpice, though, is the ability to drag and drop components, and not even have to be aware of how they components and their values mathematically relate, or even a scripting language, aside from some a handful SPICE directives. That, on top of the fact that it's free, greatly reduces the barriers to entry. So if we were ranking the options by of "ease of use" versus power, I'd go GuitarFreak, then LTSPice, then interactive Python. 

No Matlab in there at all; it is all Python, and it is all free.  Matplotlib is a Python package; it is a plotting package that works very much like the plotting in Matlab.  I think different people prefer different approaches. That is good; checking by different means is essential.  My main point here is that programming it from scratch is a simpler now the it was in the past.  If one wants even more simplicity, the six lines of code I used could be put a single routine with the list of component values as arguments.  That would not be best for me, but I think this approach could be made as simple as the others if anyone wanted to.

[wgen]

Thanks man for this step by step explanation.
I will definitely try this, I'm mainly a bassist and I'd be really interested in understanding what's the tonal response of what I'm using.
It will be difficult though, because for example my main instrument, at the moment, is a Yamaha BB bass, Precision plus Jazz pickups, and all I know is that they are ceramic (mine is the less expensive model...).
It seems impossible to find some useful specs, resistance and inductance first of all.
Also, if I understood something from your superb pickups analysis on the forum, it seems to me that the large geometry of the P pickup, together with the blade plus narrow J pickup, should make for some non conventional capacitance specs...

I hope that was clear enough, I'm not that good with English language...also, I'm a complete noob in this field, but I'll try for sure..!

[antigua]

Dec 17, 2016 7:30:43 GMT -5 wgen said:

Thanks man for this step by step explanation.
I will definitely try this, I'm mainly a bassist and I'd be really interested in understanding what's the tonal response of what I'm using.
It will be difficult though, because for example my main instrument, at the moment, is a Yamaha BB bass, Precision plus Jazz pickups, and all I know is that they are ceramic (mine is the less expensive model...).
It seems impossible to find some useful specs, resistance and inductance first of all.
Also, if I understood something from your superb pickups analysis on the forum, it seems to me that the large geometry of the P pickup, together with the blade plus narrow J pickup, should make for some non conventional capacitance specs...

I hope that was clear enough, I'm not that good with English language...also, I'm a complete noob in this field, but I'll try for sure..!

I have some Fender P-Bass and Jazz Bass pickups I will be looking at later. In general, if the pickup has a ceramic bar and steel poles, it would have around 50% higher inductance than one with AlNiCo poles, so if the DC resistance of the pickup is a typical 7k to 10k, you can assume it will be a lot darker/hotter than a comparable AlNiCo version of the pickup. 

The blades are a wild card, since the exact size of the blade makes a big difference in terms of inductance and eddy current damping, and if it's grounded, it will add a lot of capacitance, also proportionate to how much space it takes up in the core. 

[wgen]

Thanks very much for the reply!
Yes I guess it's not the easiest situation to begin with...
The blade seems somewhat thick, at least if I compare it to other blade pickups I saw on the web, ie for electric guitars.
Also, I guess that the sturdy metal parts which keep the 2 pickups in place (and which are not used with typical Fenders/Fender clones) might have a certain importance in regards to eddy current losses, isn't it?

[antigua]

Dec 17, 2016 14:51:47 GMT -5 wgen said:

Thanks very much for the reply!
Yes I guess it's not the easiest situation to begin with...
The blade seems somewhat thick, at least if I compare it to other blade pickups I saw on the web, ie for electric guitars.
Also, I guess that the sturdy metal parts which keep the 2 pickups in place (and which are not used with typical Fenders/Fender clones) might have a certain importance in regards to eddy current losses, isn't it?

Thick metal parts correlate to high eddy current losses. If the pickups have a low DC resistance, but still sound dark, I would place the blame on eddy currents.

[wgen]

Yes, this bass definitely sounds dark.
Some months ago I replaced the two 250kohm pots with two 1 megohm pots I had, and the bass, while is noticeably brighter than before, still has less top end zing than a mexican J bass I compared it with.
Also, the tone seemed bolder than that of that bass.
But I'd really like to investigate more on this, so I'll definitely try to make some simpler analysis with typical values for P and J pickups...Then, I'll try to do the same with the specs of some overwound models (especially ceramic, if I find something) of bigger manufacturers.
I know that will never be a proper test, but I guess I could somehow make up my mind about what is the ballpark of bass pickups tonal response we are talking about

[sumgai]

wgen,

For the first time since this sub-Forum opened, I must enter into the fray. Congratulations for making that happen.

If you had posed this question before antigua came along, and therefore if you had posted it in the General Wiring questions Forum, we would've told you the following, and I'm gonna tell it to you now:

First, you haven't said a word about what make or model is this bass...

Second, nary a mention of what kind of strings are on it?

Third, .... oh, you get the idea.

In essence, you're paying attention (closely!) to a fellow who is devoting his life to pickups. I can't say how he feels about guitars or basses in general - he hasn't told us anything about himself outside of his life with pickups. And I'm cool with that, don't get me wrong.

But there are other factors that will result in your undesired tone. One is the very wood used in your bass. Another is how (and how securely) the bridge is attached to the body. Third might be the nut, although that's getting to be a bit of a stretch.

Fourth, and my favorite odds-on is your strings. Without another word from you, I'd lay dollars to doughnuts that you're using a flatwound, or a nylon-taped roundwound string. Either that, or your strings have gunk on them dating back to 1986, and you're afraid that if you clean them, all the Mojo Tone will drip right out of your axe.

Before I went any further, I'd visit the folks down at the local music store, and step up to the counter, and give the nice man as much money as he wants for a new set of Stainless Steel strings. This will give you the brightest possible sound out of any instrument, guitar or bass. (Well, I probably wouldn't put 'em on an acoustic, but you get my meaning.) Only then can you be sure that the strings are not the problem, and that you must look elsewhere.

Remember back in the '80s, when we all mounted Leo Quan Baddass bridges, looking for that snappier tone and oodle's of sustain? Well, it turns out that if the stock bridge is mounted solidly, then a Baddass (or any other after-market) bridge is only gonna get you a bit more "tone", and nearly as much sustain.  In point of fact, sustain is not ordinarily something a bass player wants, so that may be of little benefit to you anyway... that'd be up to you.

Try looking at these items first, before you go off the deep end, please.

HTH

sumgai

p.s. And yes, while we Nutz prefer to complicate things as much as we can, there are practical, real-world limits to our budget, both in time and in money.  This is definitely one place were the K.I.S.S. principle should apply, no?

[antigua]

I don't have a myopic view of pickups, and I don't think they ought to be thought of as the deep end when it comes to tone consideration. Another administrator suggested moving this thread out of this sub forum because it might not be pickup-specific enough, that should serve as evidence that I take a broad view of the whole circuit. It's my genuine opinion that ceramic+steel construction leads to such high inductance and such high eddy current resistance that it is a likely culprit in dark tone, not a desire to attribute outcomes only to guitar pickups. 

I had bought a Telecaster copy from China for $150, just to see what $150 would get me, and the neck pickup was ceramic, and it showed such great eddy losses that no amount of hardware changes, or wiring changes would yield a bright tone. See the green and gray lines below:



The truth is that pickups and wiring schemes are more intertwined than pickup makers would like us to realize (I'm not so sure about the hardware argument, though), so even though this is a pickup modelling and testing forum, it ends up being more about electrical engineering when you follow things to their ends. I think it's good to put a focus on pickups though, because you have your typical pot values, 250k, 500k, etc. and typical cap values, 220nF, 470nF, and some others, but then you have a market with hundreds of pickups to choose from, so the overall pursuit of tone knowledge is not specific to pickups, but pickups are very central to the pursuit. A low output pickup like a Filter'tron will result in a bright jangly sound, while a high output monster can make the same guitar sound like punchy and bass heavy. The thing I love about pickups is, we can talk about alder versus ash, or steel versus zinc all day, but an the RLC values of a pickup speak for themselves, on top of being far more defining to the final tone than zinc, steel, rosewood, maple, etc. 

[wgen]

Dec 18, 2016 18:26:48 GMT -5 sumgai said:

wgen,

For the first time since this sub-Forum opened, I must enter into the fray. Congratulations for making that happen.

If you had posed this question before antigua came along, and therefore if you had posted it in the General Wiring questions Forum, we would've told you the following, and I'm gonna tell it to you now:

First, you haven't said a word about what make or model is this bass...

Second, nary a mention of what kind of strings are on it?

Third, .... oh, you get the idea.

In essence, you're paying attention (closely!) to a fellow who is devoting his life to pickups. I can't say how he feels about guitars or basses in general - he hasn't told us anything about himself outside of his life with pickups. And I'm cool with that, don't get me wrong.

But there are other factors that will result in your undesired tone. One is the very wood used in your bass. Another is how (and how securely) the bridge is attached to the body. Third might be the nut, although that's getting to be a bit of a stretch.

Fourth, and my favorite odds-on is your strings. Without another word from you, I'd lay dollars to doughnuts that you're using a flatwound, or a nylon-taped roundwound string. Either that, or your strings have gunk on them dating back to 1986, and you're afraid that if you clean them, all the Mojo Tone will drip right out of your axe.

Before I went any further, I'd visit the folks down at the local music store, and step up to the counter, and give the nice man as much money as he wants for a new set of Stainless Steel strings. This will give you the brightest possible sound out of any instrument, guitar or bass. (Well, I probably wouldn't put 'em on an acoustic, but you get my meaning.) Only then can you be sure that the strings are not the problem, and that you must look elsewhere.

Remember back in the '80s, when we all mounted Leo Quan Baddass bridges, looking for that snappier tone and oodle's of sustain? Well, it turns out that if the stock bridge is mounted solidly, then a Baddass (or any other after-market) bridge is only gonna get you a bit more "tone", and nearly as much sustain.  In point of fact, sustain is not ordinarily something a bass player wants, so that may be of little benefit to you anyway... that'd be up to you.

Try looking at these items first, before you go off the deep end, please.

HTH

sumgai

p.s. And yes, while we Nutz prefer to complicate things as much as we can, there are practical, real-world limits to our budget, both in time and in money.  This is definitely one place were the K.I.S.S. principle should apply, no?

The truth is that I'm a loyal follower of Antigua, the work of whom I discovered somewhere on the web 
Apart from that, I honestly appreciate this "know what you need" approach, if you will.
Believe me if I tell you that I surpassed the "old vs new/flats vs. rounds" stage many years ago. All of that are some important aspects that deserve serious attention while shaping the tone, and I keep all of those things in mind these days, too.
...Didn't really enter into that bridge replacement thing, and hopefully won't do that, for a while at least 
Now, I started searching for informations about how pickups work for quite a while, and while I'm a complete noob with experimenting and dissecting things, I'm really much more interested in understanding.
The modeling analysis proposed here seems to me an excellent starting point for what should really be the only purpose for someone like me.
Discovering the different resonant peaks, and how these can be influenced by different loads, and/or the effects of series/parallel pickups...All this might seem too simplicistic, but for me it's enough...Granted, I read all of those pickups analysis on this sub forum with great interest.
Hope that was clear!

By the way, I thought I did tell something about my bass, which is an older Yamaha BB-414 bass, P+J configuration...but, as I said previously, there aren't many other useful informations, apart from them being ceramic magnet, and the fact that the geometry of these pickups is noticeably different than that of direct Fender P and J pickups replacements...That's the reason why I wasn't more specific about those!

[sumgai]

wgen,

Sorry, somehow I missed your previous posting about your bass being a Yamaha, that's my fault.

And it sounds like you've been 'down the block and around the corner' (tranlated, you're old enough!) to understand the rest of what I said earlier, so those aspects are no longer on the plate - good deal.

---

antigua,

Make no mistake, I think this whole thing is a positive step for those who want to know more in-depth detail about the "miracle occurs here" devices under the hood (pickguard).  Most of the materials presented so far are well explained, and I like to think, understandable by a reasonably curious individual, even if they aren't aiming at an education in Electrical Engineering.

My friend, I didn't "call you out", I merely pointed out that I personally had no basis on which to make any judgments about you.  Now I know a bit more, so no hard feelings, eh?

Also, you'll note that I do not express any exceptions to some of the ideas and methods you've espoused, predominantly because I know that discussion around differing viewpoints is almost always the basis of arriving at a sound conclusion that will be agreeable to everyone.  However, for the record, I am an EE, retired since before the Millenium, and suffering from major delusions that my memory is still intact.  The wife keeps reminding me of that fact at least once a day, so I do try to bring that "new and improved" mindset with me when I visit The NutzHouse.  I do sometimes go "WTF?", but I keep quiet, because I know that there will be others who will chime in, and that when the dust settles, most likely anything that upset me will be put to rights, and that viewers of this Forum will not go away with false information.

And more than once, I've had to go back either to my college textbooks (the ones I taught from, not the ones I studied). When I cross-check them against the Internet, I find that altogether too often things have changed. My opinion may be that the change wasn't for the better, but I'm still cogent enough to realize that perhaps there was a valid reason that no one asked for my opinion on any particular upcoming/planned change. Sigh.

Please keep in mind that in a nutshell, my goal is not to "prove who's right", it's to avoid anyone making a costly mistake because of what they read here, that's all.  I dread the day that someone feels the need to go on the innerwebs and trash-talk The NutzHouse because of perceived errors in our presentments. 

HTH

sumgai

p.s. No, I don't think that this thread should move somewhere else.  We've derailed it enough (I did that, with my suggestion of the K.I.S.S. methodology), I'm sure it'll get back on track now.

 

[antigua]

wgen, if you ever have those pickups out of the guitar and can take some pictures of the pickups, top and bottom, we can make some good educated guesses. When it comes to eddy current losses and metal housings, all the TV Jones analysis were very illuminating, especially this one guitarnuts2.proboards.com/thread/7785/jones-classics-analysis-review-shootout where you had three of the same target design goal, with slight differences that that could be related to the bode plots. The Strat and PAF analysis are less educational for the fact that they are all so alike. Stratotart's eddy current research has helped a lot, too. 

Also, if you can get the DC resistances, that helps us guess how many winds are on the coil. Add that to the core situation, and we can ball park the inductance. 


***

sumgai, if you have an electrical engineering background and you see any mistakes, feel free to point them out. I don't have such a back ground, so I'm learning on the job. The main reason I'm doing this is because I'm impatient to some degree, and I want to get the information out as quick as possible. I'm personally OK with some margin for error on the basis that a lot of good information with some mistakes is better than no information at all.


A lot of what has been published out there for laymen about guitar pickups is incredibly wrong, written mostly by guys who make and sell pickups, and have no real interest in informing, and even seem intent on misinforming, if it means they can sell a few more pickups. Meanwhile, the info that has been written by physicists and electrical engineers who have no financial stake, other than to sell a dozen books to nerds like me, is very dense, often abstract and academic oriented, and not focused on helping people make purchasing decisions, or debunking specific claims that are proliferating at the moment. There's a huge information gap, and I'm trying to close it as quickly as I can manage. If you can offer any corrections, that's even better.

[wgen]

Dec 19, 2016 4:00:40 GMT -5 antigua said:

wgen, if you ever have those pickups out of the guitar and can take some pictures of the pickups, top and bottom, we can make some good educated guesses. When it comes to eddy current losses and metal housings, all the TV Jones analysis were very illuminating, especially this one guitarnuts2.proboards.com/thread/7785/jones-classics-analysis-review-shootout where you had three of the same target design goal, with slight differences that that could be related to the bode plots. The Strat and PAF analysis are less educational for the fact that they are all so alike. Stratotart's eddy current research has helped a lot, too. 

Also, if you can get the DC resistances, that helps us guess how many winds are on the coil. Add that to the core situation, and we can ball park the inductance. 


***

sumgai, if you have an electrical engineering background and you see any mistakes, feel free to point them out. I don't have such a back ground, so I'm learning on the job. The main reason I'm doing this is because I'm impatient to some degree, and I want to get the information out as quick as possible. I'm personally OK with some margin for error on the basis that a lot of good information with some mistakes is better than no information at all.


A lot of what has been published out there for laymen about guitar pickups is incredibly wrong, written mostly by guys who make and sell pickups, and have no real interest in informing, and even seem intent on misinforming, if it means they can sell a few more pickups. Meanwhile, the info that has been written by physicists and electrical engineers who have no financial stake, other than to sell a dozen books to nerds like me, is very dense, often abstract and academic oriented, and not focused on helping people make purchasing decisions, or debunking specific claims that are proliferating at the moment. There's a huge information gap, and I'm trying to close it as quickly as I can manage. If you can offer any corrections, that's even better.

I re-read that thread, definitely informative...the eddy current losses part related to screws and overall geometry was much interesting.
I'm afraid I don't have the instrument for a proper DC resistance reading;
But I'm planning to do some photos of the two pickups, both at the top as well as at the bottom, at the next strings change.
Thank you very much though!

[newey]

wgen-

Any multimeter will do the trick. LCR meters like antigua is using cost a lot more, but a simple meter can be had at Home Depot for around $15. It's very handy to have and a virtual necessity if you're going to do any rewiring of your guitar. Since you're researching for pickup replacement, it's a small price increment on top of what you'd spend for a new set of pickups.

And, antigua said:

even seem intent on misinforming, if it means they can sell a few more pickups.

They have a word for that. It's called "marketing"

BTW, the thread is fine where it is. If it wasn't, I would have moved it from the git-go. What I was suggesting is to put a copy of the original post in References. As a tutorial, it's more likely to be found there by someone interested in the topic, and we can lock out comments there, leaving this thread for comments, with a link back to here. But this is just a suggestion, it's your thread to do with as you wish.

As admins here, sg and I try to be low key about such things, we're not like those Nashville record producers who can't keep their hands off the knobs . . .

[sumgai]

Dec 19, 2016 7:03:54 GMT -5 newey said:

wgen-

Any multimeter will do the trick. LCR meters like antigua is using cost a lot more, but a simple meter can be had at Home Depot for around $15. It's very handy to have and a virtual necessity if you're going to do any rewiring of your guitar. Since you're researching for pickup replacement, it's a small price increment on top of what you'd spend for a new set of pickups.

What the nice man said.

I've checked out the lower range of multimeters from Harbor Freight, less than 10 bucks... not quite up to par. Either get a more expensive one, if you happen to be in one of their locations, or else get a better meter from HD, Lowes, Menards, Radio Shack, etc. Don't bother ordering off the web, unless you have some extra cash, and you want more features than you're likely to find on the local store's shelf, that's like buying a pig in a poke. The grandaddy of 'em all is of course, a Fluke, any model... but you better bring two checkbooks, they ain't cheap! After that, many makes and models are all made by the same Chinese outfit, just with different names/colors/features/prices/etc., they're all sufficiently good for the job a modder is likely to encounter.

And for the record, if you're going to measure the output of a pickup as it's excited, you do want a TrueRMS meter. If it doesn't say TrueRMS right on the front, or somewhere close by, then your AC readings will be approximate at best, and misleading for any calculations you might want to make. This costs more, that's true, but the differences will make it worthwhile, trust me.

And, antigua said:

even seem intent on misinforming, if it means they can sell a few more pickups.

They have a word for that. It's called "marketing"

ChrisK had a better word for it: GelderFarb. For those that never caught on, that's a fractured German word meaning 'the color of money', which can then be further translated to American slang as "Bling!". Remember, one of Chris's favorite aphorisms was "Ooh, shiny!!" As with any ChrisK-ism, one word was all it took to describe a whole paragraph of meaning. (Damn I miss that feller!)

As admins here, sg and I try to be low key about such things, we're not like those Nashville record producers who can't keep their hands off the knobs . . .

This!! I tend to think, somewhat sinfully (pride), that by letting the ship sail itself, any individual personalities will tend to merge into the whole, instead of standing out, whether for better or (more likely) worse. If I had to condense it down to one sentence, I'd say "The NutzHouse isn't my place, it belongs to all of us."

'Way anchor!!

sumgai

[sumgai]

Dec 19, 2016 4:00:40 GMT -5 antigua said:

sumgai, if you have an electrical engineering background and you see any mistakes, feel free to point them out. I don't have such a back ground, so I'm learning on the job. The main reason I'm doing this is because I'm impatient to some degree, and I want to get the information out as quick as possible. I'm personally OK with some margin for error on the basis that a lot of good information with some mistakes is better than no information at all.

A lot of what has been published out there for laymen about guitar pickups is incredibly wrong, written mostly by guys who make and sell pickups, and have no real interest in informing, and even seem intent on misinforming, if it means they can sell a few more pickups. Meanwhile, the info that has been written by physicists and electrical engineers who have no financial stake, other than to sell a dozen books to nerds like me, is very dense, often abstract and academic oriented, and not focused on helping people make purchasing decisions, or debunking specific claims that are proliferating at the moment. There's a huge information gap, and I'm trying to close it as quickly as I can manage. If you can offer any corrections, that's even better.

To put it nicely, I'm fully on-board with your entire message.  I agree that there is far to much misinformation out there, about nearly everything, not just pickups. But I learned (just last week, I'm so stubborn (hard-headed!)) that I can't make the world conform to my understanding of it.  Best to leave the windmill-tilting to younger people.    My cardiologist thinks that's a good idea, too.

As per newey's and my assertion above, and a couple of postings before that, we tend to step in and say something only when it's apparent that someone's gonna get hurt if we don't step in.  Whether that be a matter of physical safety or only a matter of money ill-spent, that's pretty much the Rubicon for us.

As to getting correct information out to modders as quickly as possible, might I suggest that you save yourself a heart attack or two, and stop trying to outpace the rest of the world.  We've been here a long time, we've been making mistakes large and small for nearly all of that time, and we're still here, amiright?  The odds are really favorable that we'll still be here when the final bell rings, mysteaks and all, and further that no one is gonna be able to stand up and shout "I win, 'cause I'm the only one that's right!".  I don't foresee that scenario playing out for anyone currently alive, do you?

So, per your first paragraph, I'm good with what I consider a bit of less-than-accurate info, it is indeed better to have a few cooties in the strawpile than to have  no strawpile at all.  Cooties can be better sifted out over time, instead of trying to prevent them from showing up in the first place.  Witness every piece of code ever released!!!*

tl;dr

You, sir, have my full respect - keep up the good work!

sumgai

* Want to know a secret? Microsoft once did publish, for sale, a program with zero bugs - really.  It was called Learning DOS, and it was made to accompany DOS 5.0, so far back in time that Christ was still a Corporal.  How do I know?  Because a certain 'other half' was one-fifth of the team that built the product. Seriously! I still have two copies of this thing, still in the unopened shrink wrap.  Might make a good inheritance for the great-grandkids, who knows....

[antigua]

I would slow it down, but I have a rather big agenda. I still have a variety of bass pickups to test, some single coil sized humbuckers which I might just skip over, and then I need to go back and put all the data into a spread sheet so that I can compare them all more efficiently than eye balling it like I do now. I think the inductance values for all these pickups is the most valuable information to be obtained, since it's the dominant factor, so this is a great improvement over the status quo. At the same time, I have to somehow convince guitarists that the inductance is important. If you go here thedougandpatshow.com/jason-lollar-comes-for-a-visit/ and fast forward to 22:55, you see this guy named Pat ask Jason Lollar what inductance is. Keep in mind these are guys who review pickups. It's the most important metric of a pickup - and he isn't even familiar with the concept itself. The disparity between where guitarists are, and where they should be, is astonishing. It's like we're in the 1750's being sold snake oil out the back of a wagon.

The other point I'd like to express is that a lot of guitarists see this data and thins it takes the magic out of pickups. They say that they hear a lot more of a difference than a slight shift in the resonant peak when they switch out one pickup set for another. The harmonic structure that comes out of an electric guitar is so intricate that changing one parameter can be perceived as ten changes. People should know that they magic happens in our heads, not in the pickup winder's hands.

Where as in most areas of commerce, it's assumed that someone who is selling something will exaggerate what they have to offer, guitarists trust pickup vendors whole heartedly, and adopt their wild claims as though it were truth. It will require a strong offensive on the consumer side to right these wrongs. 

[sumgai]

antigua,

All of your above points are salient, to say the least. One thing of which I am sure is that there are very, very few pickup designers/makers who are wholly educated in Electrical Engineering. IOW, most of these people started out tinkering, thought they'd found the Fountain Of Youth or The Holy Grail, and like your proverbial snake-oil salesman, headed off to fleece the masses. It no longer upsets me, but I still feel sad whenever my memory spouts this knowledge back to the top of my awareness.

But my idea of "slowing down" was aimed not so much as how many reviews/analysis you can post, it was pointed more in the direction of how much 'background  theory' and real-world observed test results you post. Recall that in college, all of the information presented in 4 years could have been spewed out in less than a month. But experience shows that people just don't learn in that manner - they need some time to absorb the info, preferably in small, even tiny, increments. IOW, information overload tends to negate any and all benefits that listener might otherwise have derived... trust me on that one!

You've kind of already hinted that this is your direction, pointing out that inductance is the foundation of all things electrically magnetic in nature. It doesn't dwarf all else, but it is the basis from which nearly all else springs, you and I agree on that.  So I'm inclined to let you refine your efforts as you see fit, and certainly, to use The NutzHouse as a pulpit from which you can preach the gospel according to nearly all of mankind's collective learning over the past 150 years or so. But....

People in general, and guitarists in particular (bassists included) are a very close-minded lot. The fact that you're trying to educate the general player's market just substiantiates this as a self-provable fact. For the vast majority of players, seeing Steve Morse or John Jorgensen cleanly execute 400 notes per second is enough of an endorsement for them to go out and buy the same guitar, or at least the same pickups. After all, it works for the big boys, right? So why not for me (meaning, me the little guy). And this conditioning is only reinforced by the fact that neither Steve nor John are themselves immersed in technology, they only know that if they can buy it off the shelf (or have it custom made), then all is right with the world, and the eggheads can go suck asselschwanse.

Case in point: A few years ago, Guitar Center bought Eric Clapton's "Blackie" for just pennies short of One - Million - Dollars (US). That's enough to feed lunch to everyone, not just the poor kids, in a city the size of Los Angeles for a week, perhaps longer. And why did the price rise so high, at least a thousand times over its intrinsic value? The old "It was good enough for E.C., so it must be worth more. Just look at all that tone he got out of it!" mindset.*   That's not so much greed as it is gullibility.  And marketing people know that.... hell, they make their living by using that knowledge.  How they sleep at night, I'm not gonna speculate.

.......

I really had intended not to carry this out so far, and to let this derailment float off into the sunset.  Sorry about that.

sumgai

* Some member here once had a signature that read "There ain't no tone in my fingers.  One night I shook them all over the place, and not one tone fell out of any of them!"

[antigua]

I'm mostly in a data gathering stage right now. I'm trying to get these pickup values out there while I've got a momentum going, so this is like just one phase of a longer term plan. It's also fun to talk about them as well, since I discover lots of interesting things in the process, such as the correlation between coil thickness and capacitance, or inductance and AlNiCo grade, etc. One big thing I found is that the way coil tapping is done with single coils causes an outrageously high capacitance to result. The data has been useful at enhancing the GuitarFreak pickup modelling, and on and on. 

Overall, everything is going according to plan. There are other players like wgen who appreciate the extra data points being made available. I also post much of this data to the forums where they are relevant, like Strat pickup data is posted to Strat Talk, etc. The only ones that I haven't posted elsewhere are most of the humbucker tests, because, for whatever reason, the humbucker set tends be a more hostile, so I have to approach them with a lot more care. Once I have all the data in a spreadsheet, I think it will be a lot easier to promote, because then side by side contrasting will be easy and convenient, and I think players will see that the data is not only useful, but should be demanded of pickup makers. They can all afford a $200 Extech. Most likely have one, and just choose not to publish their recorded values. 

[blademaster2]

I have been scanning this thread with great interest, as I myself performed some spice modeling of my own guitar circuitry many years ago during university (electrical engineering).

While it impresses me, I like to bear in mind that like any simulation the results are only as good as the model. The characteristics of pickups and other guitar electronic components are simplified to linear mathematical models of inductance, capacitance and resistance by using inductors, capacitors and resistors. As Antigua indicated, these are often distributed things and not really fully correct when lumped (it would be terribly complex to even begin to model them as distributed components, so I am not suggesting we ever try).

Still, getting a sense of the overall frequency response and the resonant peaks has value to us all in understanding why we hear what we hear and it gets us pretty far down the road.

The notoriously difficult part of the electric guitar when viewed as a system, often completely ignored, is the effect of the volume of the amplified sound in the room. My wife would complain if I employed this to get the tone that I love (I wait until she is out), but for recordings and live performances it is probably not realistic to ignore it as an important part of the resulting tonal quality. Consider Jimmy Page's solo on "Ramble On" where the guitar feeds back the notes to more resemble a keyboard than a guitar (he does it using a few overdubs in this song and I love that sound). That was made possible by the volume feeding back the acoustic signal through the guitar body - not via the strings or electronics. To model this would involve including another signal source, with another gain factor, and coupling this signal additively into the overall guitar signal through a resonant and frequency/phase-response altering body material (yes, I say again that the body material influences this especially when the volume goes way up).

Of course, much smaller would be the filtering effect of the body material properties on the initial generated signal itself, as we have all debated countless time before, which is independent of amplifier volume.

Thanks so much - I love to see this simulation stuff, as it helps me better understand that portion of the beautiful system formed by the electric guitar and amplifier. I would like to hear how someone can find a way to add the acoustic aspects of the system to make it even more complete.

[One final note: When I met Steve Morse I was impressed by his working knowledge of electricity, signals and EMI. He was not schooled as an electrical engineer but he made a point of recognizing directionality of EMI noise sources, non-idealities of buffer amplifiers, and gave me an interesting and practical interpretation of them from a musician's perspective.]

Sorry for being so long winded, I love this topic  ......

[sumgai]

I didn't mean to disrespect Mr. Morse, he's one of my idols, make no mistake. But like most other players, he's knowledgable about what his technicians bring to his attention, whenever he observes (suffers) a problem. The fact that he is interested, and not just a "fix it or else" kind of guy is all the more reason to respect him, IMO. But I did select him and John J. as examples of extraordinary players who literally ooze tone whenever they play, and yet they don't need to know what reluctance or permeability are in order to do so.

As to changing tonality via acoustic feedback, that's true all the way.  I personally have held the idea, for decades, that any guitar's tone can be improved by playing loudly, and preferably directly in front of the speaker(s).  One might be surprised at just how good a $50 Teisco Del Rey can sound, given the right stage and amp/speaker arrangement.

My personal favorite story is: Years ago, before I entered these hallowed halls, I installed a Ghost system onto my Strat (now gone to greener pastures). As explained by GraphTech's technician at the time (Les?), the graphite material of the bridge saddle lowered the resonance peak level (not the frequency) by about 3dB, from 4.5 to next to 1.5dB... seriously.  Suddenly the guitar's "bite" was less edgy.  But even with that reduced peak, the axe still had nannies, and when cranked, it still commanded respect from all and sundry.

Which comes to point out the following.....

As Runewalker has pointed out many a time, no one part or piece can be singled out as the leading cause of Tone.  IOW, we must take all of the pieces together in order to figure out what's what.  But that means that we do need to understand, at better than a rudimentary level, what each piece does, and how it contributes to that overall Mojo Tone.  Hence, these discussions are of high value.  All we need do, and indeed, all we can do, is arrive at a standardized base point, i.e. a methodolgy that will repeatedly come up with the same results every time, no matter who's doing the testing.  And after that, we do the work, on not one or two samples, but hundreds of them.  This will be the only way to eliminate manufacturing anomolies, whether they first appear to beneficial or otherwise.  That means that a lot of Nutz are gonna get press-ganged into doing something "for The Gipper", eh?

sumgai

[stratotarts]

Dec 19, 2016 18:24:34 GMT -5 blademaster2 said:

...

While it impresses me, I like to bear in mind that like any simulation the results are only as good as the model. The characteristics of pickups and other guitar electronic components are simplified to linear mathematical models of inductance, capacitance and resistance by using inductors, capacitors and resistors. As Antigua indicated, these are often distributed things and not really fully correct when lumped (it would be terribly complex to even begin to model them as distributed components, so I am not suggesting we ever try).

 ......

John's 6-component pickup circuit model does manage to track measured results to fractions of a decibel. There are also some interesting models in Zoeller's work, yet untranslated to English. So there is a lot more hope for it than you suggest. But it bears repeating that it is possible to capture a plot of complex impedance instead of only a composite of the real and imaginary impedance as everyone seems to be doing now (including the V5 integrator). That is what Mike Sulzer's rig can do. Such a data set allows perfectly accurate modeling without having any composite ideal component model at all, since it represents exactly the impedance that is presented at the pickup terminals, no matter what component configuration exists inside it. I am not sure that it can work well with external excitation, so I think there is still a place for the V5 and similar approaches, as a complimentary method. For one thing, I have a hard time believing that eddy losses can be measured accurately by exciting the pickup coil itself, since the influence of the string field is asymmetrical with respect to the coil, and materials such as covers that stand between the two may prove to have a larger influence than would be indicated by a coil based (conducted) measurement.

Since the pickup is a two terminal device, if the complex impedance at a certain frequency is known, what you measure is the superposition of arbitrarily complicated linear circuits inside that do not need to be known. The vector can be plugged directly into a simulation of a load network, and the response plotted. It is not necessary to know anything at all about the internal circuits in order to do this. There is no limit to this principle, as far as I know, as long as the components remain linear. It is also easy to derive basic parameters such as inductance and capacitance this way, Mike has apparently already done that.

By refining the analytical methods, it might be possible to also automatically translate the data set into a matching 6-component model, or other models that are more complex than the basic LCR one. The main usefulness of such reductionism is to aid in classifying and comparing different designs.

This really is an unprecedented "stone soup" gathering of interest. I think it is really great how many of us have specialized and attacked different aspects of the problem. It's been about half a year since I started the integrator project, and if the next half year brings the same intensity of research, it will contribute to the largest effort to understand pickups that has happened since they were invented. The free and open exchange of opinions and information has been one reason that it has motivated so many knowledgeable and hardworking people to contribute. We have had mostly positive feedback and many thanks from others who were curious about our findings. A few people didn't get it, but that always happens.

[JohnH]

Our threads here have definately taken the science of guitar and pickup electronics further than has been presented anywhere before on a public site.

On the modelling, the derivation and use of impedance curves is promising so long as it also comes with a voltage vs frequency curve. You need real and complex impedance plus voltage vs frequency to capture it. An impedance curve that varies with frequency will still give a flat-line response into a high impedance input unless voltage change is also modelled.

The method of making a 6-part model, although it works quite well, is tricky to form an algorythm for, since there are 6 variables and the result is a close fit rather than an exact match. Ive been using 3 resistors, 2 inductors and 1 cap. To find 6 values you need 6 pieces of info, which are usually highly interactive. The easiest one to get right is to match the measured dcr. So i make the main coil resistance a function of tbe other two and tbe dcr. This then leaves 5x variables. These varaibles cant be fully seperated but each has a more primary effect on certain aspects of the response. It takes many interations to home in.

In our loaded and unloaded models, the single cap most directly affects the frequency ratio of the two peaks, in combination with the cable capacitance.

Tbe load resistor (in parallel with the whole model) is most effective in changing the height of the unloaded peak.

The main inductance gets the peak frequencies where they should be. The second inductor with its series resistor shapes the curves, getting the width of tbe peaks and any mid dips modelled. These are the trickiest ones to select and equally reasonable modells can have a range of values here and still be a good (though not perfect) match.

[antigua]

Dec 20, 2016 5:11:08 GMT -5 JohnH said:

Our threads here have definately taken the science of guitar and pickup electronics further than has been presented anywhere before on a public site.

On the modelling, the derivation and use of impedance curves is promising so long as it also comes with a voltage vs frequency curve. You need real and complex impedance plus voltage vs frequency to capture it. An impedance curve that varies with frequency will still give a flat-line response into a high impedance input unless voltage change is also modelled.

The method of making a 6-part model, although it works quite well, is tricky to form an algorythm for, since there are 6 variables and the result is a close fit rather than an exact match. Ive been using 3 resistors, 2 inductors and 1 cap. To find 6 values you need 6 pieces of info, which are usually highly interactive. The easiest one to get right is to match the measured dcr. So i make the main coil resistance a function of tbe other two and tbe dcr. This then leaves 5x variables. These varaibles cant be fully seperated but each has a more primary effect on certain aspects of the response. It takes many interations to home in.

In our loaded and unloaded models, the single cap most directly affects the frequency ratio of the two peaks, in combination with the cable capacitance.

Tbe load resistor (in parallel with the whole model) is most effective in changing the height of the unloaded peak.

The main inductance gets the peak frequencies where they should be. The second inductor with its series resistor shapes the curves, getting the width of tbe peaks and any mid dips modelled. These are the trickiest ones to select and equally reasonable modells can have a range of values here and still be a good (though not perfect) match.

I'm still not sure what the "eddy current" resistor and inductor actually do. I understand inductive/capacitance resonance, but I'm not sure how the parallel resistor and inductor manage to create a second low pass filter that somehow leaves the original inductive/capacitance resonance mostly untouched, as opposed to, say, a resistor by itself, which purely dampens the original resonance, or just an inductor by itself, which acts as a high pass filter. Put the two together in series, and it does neither of those things that they do by themselves. 

AFAIK, actual eddy current attenuation is external to the pickup circuit; it's within the  "AC source" black box. In LTSpice I have to live with the fixed voltage for the AC sweep, but can you just model the eddy currents as an attenuation of the "AC source" with GuitarFreak?

Lastly, I don't quite understand how the impedance curve differs from the voltage by frequency bode plots. The impedence is reactance and resistance, we know the resistance is whatever it is, so if you drive the pickups with a constant voltage in, the voltage curve that comes back out must owe to the variable impedance with frequency. So, how is an impedance curve different from what we're doing now?