Pickup amplitude trend with respect to pickup height; any theories?

[John Kolbeck]

I recorded some samples of plucking the G and D strings of a Strat, with the pickups at two heights, 3mm and 6mm, with the intention of doing a frequency analysis to determine which frequency ranges were more or less emphasized.

The results of the frequency analysis were rather ambiguous, but I did notice something else when looking at the raw amplitudes in the DAW, and that is that when the pickups were set to 3mm distance (in green), the amplitude over time showed a greater tendency to favor either the positive or the negative at different times, in an oscillating fashion, so that the end result is that it looks somewhat like a snake with a waving tail, where as the 6mm distant waveform (in purple) looks a lot more uniform in the positive in negative side, with a more linear downward slope as the energy in the string decays.





Any idea why this would be? I realize that the string's pattern of movement is not perfectly uniform in general, that the ratio of vertical to horizontal movement changes as the string vibrates, which probably explains why the amplitude is not a perfect slope, but what I don't understand is why having the pickup mounted closer to the string exaggerates that non-linearity in the positive to negative ratio (such that it looks more snake like).

Here's my theory: we know voltage is created when the flux density changes in the coil, so my thinking is that when the pickup is closer to the string, there is sharper magnetic focus upon the string, so that when the string is further "off center", to the left or the right of the pole piece, it experiences a proportionately sharper drop in flux, where as when the pickup is further away, and the flux is more spread out, so there is not as proportionately large of drop in flux as the string moves "off axis". The greater disproportion in flux density imparted upon the string as the string moves in and out of a tighter area of flux would cause greater degrees of amplitude variation as the string's vibration evolves from being more horizontal to more vertical, and back again. I'm not super confident this is what is going on, though.

I'd be interested to know if anyone has a theory as to how that snake-like amplitude pattern comes about when the pickup is closer to the strings.

[rjb]

Stratitus.
The waveforms in green are "riding" on a low frequency oscillation, caused by the magnets "grabbing" the string.
Or something like that.

[Richard]

Perhaps the closer pickup is picking up more of the vibration parallel to the guitar top as the "axis" of vibration moves but this seems like it would show less variation. I'm imagining that a string moving parallel with the top of a guitar has more movement in relation to the pickup polepiece the closer they are to each other. Possibly the magnetic field is changing the vibrations or causing the different harmonics to move in and out of phase. I doubt it is the cause of this phenomenon but could you describe how you performed the tests?

[rjb]

This might be a clue: The "wiggle" appears to be sinusoidal in nature.
What, approximately, is the period of the wiggle- that is, the time between peaks?
The frequency in Hz is one divided by the period. f = 1/T
Does that frequency happen to have any relationship to line frequency, or anything else you can think of?

[John Kolbeck]

Perhaps the closer pickup is picking up more of the vibration parallel to the guitar top as the "axis" of vibration moves but this seems like it would show less variation. I'm imagining that a string moving parallel with the top of a guitar has more movement in relation to the pickup polepiece the closer they are to each other. Possibly the magnetic field is changing the vibrations or causing the different harmonics to move in and out of phase. I doubt it is the cause of this phenomenon but could you describe how you performed the tests?

I'll have to give what you're saying some deeper thought.

I just plucked the open G and D strings, one at a time in a consistent location over the middle pickup, making sure to use about the same pressure each time. You can see in the peaks about how consistent the pick attack is.Then I trimmed the audio so that there would be three plucks in a row, each lasting exactly two seconds.

[bbsailor]

I recorded some samples of plucking the G and D strings of a Strat, with the pickups at two heights, 3mm and 6mm, with the intention of doing a frequency analysis to determine which frequency ranges were more or less emphasized.

The results of the frequency analysis were rather ambiguous, but I did notice something else when looking at the raw amplitudes in the DAW, and that is that when the pickups were set to 3mm distance (in green), the amplitude over time showed a greater tendency to favor either the positive or the negative at different times, in an oscillating fashion, so that the end result is that it looks somewhat like a snake with a waving tail, where as the 6mm distant waveform (in purple) looks a lot more uniform in the positive in negative side, with a more linear downward slope as the energy in the string decays.





Any idea why this would be? I realize that the string's pattern of movement is not perfectly uniform in general, that the ratio of vertical to horizontal movement changes as the string vibrates, which probably explains why the amplitude is not a perfect slope, but what I don't understand is why having the pickup mounted closer to the string exaggerates that non-linearity in the positive to negative ratio (such that it looks more snake like).

Here's my theory: we know voltage is created when the flux density changes in the coil, so my thinking is that when the pickup is closer to the string, there is sharper magnetic focus upon the string, so that when the string is further "off center", to the left or the right of the pole piece, it experiences a proportionately sharper drop in flux, where as when the pickup is further away, and the flux is more spread out, so there is not as proportionately large of drop in flux as the string moves "off axis". The greater disproportion in flux density imparted upon the string as the string moves in and out of a tighter area of flux would cause greater degrees of amplitude variation as the string's vibration evolves from being more horizontal to more vertical, and back again. I'm not super confident this is what is going on, though.

I'd be interested to know if anyone has a theory as to how that snake-like amplitude pattern comes about when the pickup is closer to the strings.

John,

See this thread for an interesting discussion related to the asymmetry that you observed. http://music-electronics-forum.com/t21125/

Try this test using your same DAW setup. Pinch the test string right above the neck pickup. Pull the string up about .125 inches (or a little less if your action is low). Release the string and observe the initial attack and note the asymmetry of the differences in the signal level above and below the zero line. Do this with the pickup at close and farther distances from the string and note how the amount of asymmetry varies with pickup distance from the string. Try this again over the bridge pickup and note the amount of asymmetry. It should be less because the string moves less over this pickup. The vertical motion of the string generates a greater initial peak voltage than the horizontal motion. This pinch test is what I discovered to try to isolate the vertical and horizontal string motion effects of what you hear but only for the first few cycles before the string begins to do its normal elliptical rotation.

Now pinch the string again but this time pull it sideways parallel to the surface of the guitar top and observe the first few cycles. Since one complete horizontal cycle passes over the magnet pole twice, this direction of motion creates a stronger second harmonic at twice the normal string frequency. However, it does not last too long as the string rotates but you will definitely see a difference if you do this experiment as I recommend.

You can even hear a difference between a vertical pinched and released string versus a horizontal pinched and released string done over the pickup. It is more apparent in the initial attack. This has implications for the initial direction of your pick striking the string(s) and ultimately your picking technique.

Try it and let us know what you discover.

Joseph J. Rogowski

[John Kolbeck]

This might be a clue: The "wiggle" appears to be sinusoidal in nature.
What, approximately, is the period of the wiggle- that is, the time between peaks?
The frequency in Hz is one divided by the period. f = 1/T
Does that frequency happen to have any relationship to line frequency, or anything else you can think of?

The top two sounds files are of an open D, the bottom two are an open G, blue is far, green is near.

The "humps" in the D at 3mm appear to last about 1.33 seconds peak to peak, or 0.75 Hz. The "humps" in the G a 3mm appear to be 0.852 secs peak to peak, or 1.174 Hz. It's real easy to identify where the humps are with the "G" string wave, and a little more murky with the "D".

Here are a some zoomed in versions:

Here's the first .25 seconds:



Here is the first two whole seconds, spread across two screen shots:

[John_H]

The results of the frequency analysis were rather ambiguous, but I did notice something else when looking at the raw amplitudes in the DAW, and that is that when the pickups were set to 3mm distance (in green), the amplitude over time showed a greater tendency to favor either the positive or the negative at different times, in an oscillating fashion

This is probably due to the string vibrating randomly. The direction of the vibration is changing, and exciting the coil with a constantly changing signal. This could be the cause of the fluctuation, and the effects could be intensified by the greater magnetic pull of being closer.

[rjb]

Isn't 3mm awfully close?
I'm more of an acoustic guy- but if I were to set the neck pickup of my Tele 3mm from the open strings, then played at the top frets, the strings would hit the pickup.
It looks like you may have sub-sonic sympathetic vibrations going on. I wouldn't be surprised if the height adjustment screws and springs are so loose that the pickups are bouncing. How's that for a theory?

[John Kolbeck]

Isn't 3mm awfully close?
I'm more of an acoustic guy- but if I were to set the neck pickup of my Tele 3mm from the open strings, then played at the top frets, the strings would hit the pickup.
It looks like you may have sub-sonic sympathetic vibrations going on. I wouldn't be surprised if the height adjustment screws and springs are so loose that the pickups are bouncing. How's that for a theory?

The strings didn't hit the pickup's pole pieces. I got them close in order to compare the extremes of near-ness and far-ness. A 6mm the pickups are about flush with the pick guard.

3mm distance can cause "wolf tones" on the low E and A, but I didn't hear that occur with the D and G strings, which I was testing with in this case.

The pickups were not loose in either case. They are tightest when they are the closest, since there is standard rubber tubing around the mounting screw that gets progressively tighter the higher the pickup is positioned.

[rjb]

The pickups were not loose in either case....

Oh, well. It was just a "theory".