Measuring pickups

[mozz]

Plan to log everything i make with a LCR meter. From the manual:

"Any non-ideal component can be represented as a resistive component either in series or in parallel with a
reactive component. Depending upon the characteristics of the component and the measurement frequency,
either the series or parallel model will be the more accurate model.The more ideal a component is, the less
the two models differ. However, as Q for inductors decreases towards 10 and Q for resistors or D for
capacitors increases towards 0.1, the series and parallel values begin to diverge. In general, most parts are
best approximated by the series model. Manufacturers often specify which representation should be used
when testing their devices. MIL or industry standards also specify test conditions."

Default is series model.
Also default is 1khz, but the meter will also let you choose 100hz,120hz, 1000hz,10000hz
Anyone have some average values?
I see at one time there was a pickup-wikipedia webpage, appears to be gone?

[John Kolbeck]

This regards the Extech 380193 specifically, but the settings you're supposed to specify should apply to your meter also http://music-electronics-forum.com/t28229/

It appears that you have an SR715 or SR720. They say that an LCR meter has to be able to accurately test a "low Q inductor" in order to get reliable figures when testing a guitar pickup, I'm not sure if your LCR meter qualifies. I assume that's something you'll want to determine before you proceed, though. Apparently, most inductors have a much higher Q factor than a guitar pickup, presumably because most inductors tend to be smaller and/or use larger wire than a pickup, resulting in smaller resistances and higher Q factors.

[Mike Sulzer]

A complicating factor in measuring pickup impedance is that some pickups have both series and parallel loss. The series loss is due to the coil resistance, of course, and the parallel loss is from eddy currents in the cores. However, as a practical matter, the effects of eddy currents are very small (but not completely absent) when the measurement is made at 100 Hz. Attached here is a measurement of an extreme case, the Seymour Duncan hot rails. By measuring at many frequencies and employing some numerical analysis, you get the impedance with the effect of the coil capacitance removed. The result for the imaginary part of the impedance should be increasing linearly if an inductor were all that is involved (the dashed gray line). But the yellow line is the actual result, and, it is significantly curved. Some people like to say that the inductance is a function of frequency (and you do measure lower inductance at higher frequencies with the meter), but I think it is more accurate to say that the model of the pickup impedance contains more components than just an R, L, and C.

SDhrb.pdf

Plan to log everything i make with a LCR meter. From the manual:

"Any non-ideal component can be represented as a resistive component either in series or in parallel with a
reactive component. Depending upon the characteristics of the component and the measurement frequency,
either the series or parallel model will be the more accurate model.The more ideal a component is, the less
the two models differ. However, as Q for inductors decreases towards 10 and Q for resistors or D for
capacitors increases towards 0.1, the series and parallel values begin to diverge. In general, most parts are
best approximated by the series model. Manufacturers often specify which representation should be used
when testing their devices. MIL or industry standards also specify test conditions."

Default is series model.
Also default is 1khz, but the meter will also let you choose 100hz,120hz, 1000hz,10000hz
Anyone have some average values?
I see at one time there was a pickup-wikipedia webpage, appears to be gone?

[bbsailor]

Plan to log everything i make with a LCR meter. From the manual:

"Any non-ideal component can be represented as a resistive component either in series or in parallel with a
reactive component. Depending upon the characteristics of the component and the measurement frequency,
either the series or parallel model will be the more accurate model.The more ideal a component is, the less
the two models differ. However, as Q for inductors decreases towards 10 and Q for resistors or D for
capacitors increases towards 0.1, the series and parallel values begin to diverge. In general, most parts are
best approximated by the series model. Manufacturers often specify which representation should be used
when testing their devices. MIL or industry standards also specify test conditions."

Default is series model.
Also default is 1khz, but the meter will also let you choose 100hz,120hz, 1000hz,10000hz
Anyone have some average values?
I see at one time there was a pickup-wikipedia webpage, appears to be gone?

Another factor that might affect subjective pickup sound quality is the initial first 30 milli-seconds of the initial attack. See this article, that I just located on the web, to provide more details about this very subjective issue. https://www.researchgate.net/publica...sient_Response

I have always been aware how sensitive the human ear is to initial sound transient quality but this research provides some additional details to help explain why some pickups may be perceived to sound better and may indicate some other measurement techniques that may be needed to replicate desired pickup features other than traditional measurements.

Joseph J. Rogowski

[John Kolbeck]

Another factor that might affect subjective pickup sound quality is the initial first 30 milli-seconds of the initial attack. See this article, that I just located on the web, to provide more details about this very subjective issue. https://www.researchgate.net/publica...sient_Response

I have always been aware how sensitive the human ear is to initial sound transient quality but this research provides some additional details to help explain why some pickups may be perceived to sound better

That research paper is a bit ponderous. Can you cite the portion of that research that you felt was particularly insightful?

One thing caught me eye, though:

This is because an accuratemodel of pickup would involve, for instance, the eddy-currents on the pole pieces of the pickup. Such eddy currents lead to“wolf sound”. For a properly adjusted guitar, string heights above pole pieces are adjusted to make it imperceptible and so we do not need to model the eddy-currents.

That seems very incorrect.

[bbsailor]

That research paper is a bit ponderous. Can you cite the portion of that research that you felt was particularly insightful?

One thing caught me eye, though:



That seems very incorrect.

To answer your question directly, the initial attack in the first 30 milli-secnds affects our sound perception

I always encourage critical thinking about scientific principals. My posting of this web reference was to encourage MEF members to evaluate the possibility that distributed capacitance in a pickup coil winding was any different from the bulk capacitance of a coax guitar cable at about 35 pf per foot? In other words; does the distributed capacitance within a pickup coil differ from the bulk capacitance of the coax cable capacitance connecting the guitar to the amplifier input? I am simply sharing my pickup aural observations and Web discoveries, not making any brash statements about pickup technology.

Critical thinking advances most technology!!!

Joseph J. Rogowski

[John Kolbeck]

Others may disagree, but I've found that the attack characteristic of a pickup tends to follow the overall output and treble response. Lots of treble = glassy attack, medium treble = punchy attack, low treble = softer attack, so I tend to no longer think of a pickup's transient response as anything specific unto itself. I can't think of any physical reason for why a pickup would behave one way during the transient period, and another during the the decay period.

[bbsailor]

Others may disagree, but I've found that the attack characteristic of a pickup tends to follow the overall output and treble response. Lots of treble = glassy attack, medium treble = punchy attack, low treble = softer attack, so I tend to no longer think of a pickup's transient response as anything specific unto itself. I can't think of any physical reason for why a pickup would behave one way during the transient period, and another during the the decay period.

John,

See this MEF link http://music-electronics-forum.com/t36537/ for a good technical discussion about pickup capacitance measurement.

Remember that the human ear is most sensitive in the sound spectrum near the resonant peak of most pickups (2KHz to about 4KHz) known as the Fletcher Munson Curve. The low E string on the guitar is 82 Hz and takes 12.195 milli seconds to make one cycle and in the 30 milli seconds will go through 2.46 cycles, the high E string 3.0487 milli seconds and goes through 9.84 cycles and the high E on the 12th fret 1.524 milli seconds and goes through 19.68 cycles in 30 milliseconds. The amount of energy in the treble response of the sound spectrum region that you characterize as glassy attack, punchy attack or softer attack mostly contains the upper harmonics of the string and how well aligned these harmonics are to the fundamental frequency goes into how we perceive the first 30 milli seconds of the initial attack and is that subjective characteristic about the pickup sound. Pickup geometry translates horizontal string motion into a 2nd harmonic output component as the string passes over the magnet twice in one cycle while the vertical motion of the string contributes to the peak fundamental frequency output. However, since the magnet and coil are below the string, the output from vertical string motion has a higher asymmetry when the string is closer or farther from the coil and magnet but has a more symmetrical output from sensing the horizontal string motion.

During the decay period, string motion is much less than during the initial string displacement from the picking or strum. How long this period lasts relates to how much energy from the strum or string pick is maintained within the vibrating string(s) by reducing energy being absorbed from the string(s) by the nature of the guitar construction.

Rick Turner reported (here on MEF) hearing sound differences by using Litz wire to make the pickup coil and I would believe that most of the sound difference is within the first 30 milliseconds of the string attack.

Joseph J. Rogowski

[Mike Sulzer]

Others may disagree, but I've found that the attack characteristic of a pickup tends to follow the overall output and treble response. Lots of treble = glassy attack, medium treble = punchy attack, low treble = softer attack, so I tend to no longer think of a pickup's transient response as anything specific unto itself. I can't think of any physical reason for why a pickup would behave one way during the transient period, and another during the the decay period.

Exactly right.

[John Kolbeck]

John,

See this MEF link http://music-electronics-forum.com/t36537/ for a good technical discussion about pickup capacitance measurement.

Remember that the human ear is most sensitive in the sound spectrum near the resonant peak of most pickups (2KHz to about 4KHz) known as the Fletcher Munson Curve. The low E string on the guitar is 82 Hz and takes 12.195 milli seconds to make one cycle and in the 30 milli seconds will go through 2.46 cycles, the high E string 3.0487 milli seconds and goes through 9.84 cycles and the high E on the 12th fret 1.524 milli seconds and goes through 19.68 cycles in 30 milliseconds. The amount of energy in the treble response of the sound spectrum region that you characterize as glassy attack, punchy attack or softer attack mostly contains the upper harmonics of the string and how well aligned these harmonics are to the fundamental frequency goes into how we perceive the first 30 milli seconds of the initial attack and is that subjective characteristic about the pickup sound. Pickup geometry translates horizontal string motion into a 2nd harmonic output component as the string passes over the magnet twice in one cycle while the vertical motion of the string contributes to the peak fundamental frequency output. However, since the magnet and coil are below the string, the output from vertical string motion has a higher asymmetry when the string is closer or farther from the coil and magnet but has a more symmetrical output from sensing the horizontal string motion.

During the decay period, string motion is much less than during the initial string displacement from the picking or strum. How long this period lasts relates to how much energy from the strum or string pick is maintained within the vibrating string(s) by reducing energy being absorbed from the string(s) by the nature of the guitar construction.

Rick Turner reported (here on MEF) hearing sound differences by using Litz wire to make the pickup coil and I would believe that most of the sound difference is within the first 30 milliseconds of the string attack.

Joseph J. Rogowski

Suppose I take it at face value, that the sound quality of a pickup is mostly judged by what is heard during the period of initial attack. In order for that to matter from a measurement standpoint, it would have to also be true that the pickup's transfer characteristics are non-linear with respect to amplitude, which is to say "the pickup has these values when the string is plucked hard, but it has these values when the string is plucked softly", supposing some sort of saturation occurs at a certain amplitude, then you would have to set out to determine the "plucked hard" values so that you can say this is what matters in the first 30 milliseconds and it's this measure by which the pickup is mostly judged. But, AFAIK, pickups are mostly, if not completely, linear, with respect to amplitude.