time domain reflectometry

[fieldwrangler]

Folks -

Anyone think that unintended variations (in tension/stretching/wire diameter, etc.) of coils will show up with this test technique?

Bob Palmieri

[Jazz P Bass]

Not too sure what wire you want to test but a true TDR test requires two conductors.

[Mike Sulzer]

No, you have too much capacitive and inductive coupling in the coil.

[Mike Sulzer]

Not too sure what wire you want to test but a true TDR test requires two conductors.

One wire over ground can be a transmission line.

[bbsailor]

One wire over ground can be a transmission line.

Mike is correct!

You also need one more thing; the velocity factor of the transmission medium you are using. Velocity factor - Wikipedia, the free encyclopedia

When technicians look for faults in long transmission lines such as coax or fiber optics cable, a pulse is sent down the line from the TDR tester. Then, when the pulse reaches the fault a reflected pulse is echoed back to the TDR tester. The time it takes to make the round trip is derated by the velocity factor of the medium to accurately calculate the distance from the TDI tester to the fault. Most common transmission media have known velocity factors to make locating the fault fairly accurate, usually down to a few feet or at worse a few yards for very long distances. As TDI testers evolved from the very early ones, the accuracy improved greatly.

Assuming you could get a reading from a pickup (which I doubt) with a broken or shorted wire, the velocity factor would need to be known and would vary by the wire size, winding compactness and dielectric constant of the insulation on the wire.

Joseph J. Rogowski

[David King]

This thread is ringing a bell for me. I'm sure we've discussed this before or something like it.

[Joe Gwinn]

This thread is ringing a bell for me. I'm sure we've discussed this before or something like it.

We have discussed the impulse response of a pickup (which I measured a few years ago), but I don't recall any discussion on TDR.

TDR will give you the impulse response function as well, but messed up. TDR is for wideband systems (meaning many megahertz), not audio-range resonances.

[fieldwrangler]

Thanks all!

This all makes sense to me. It also opens the door to the larger question, which I'm sure has been substantially addressed in numerous previous posts, but may have some new elements affecting the general point of view.

What's worth measuring these days, in both conventional multi-turn hi-Z and also these pioneering single-turn ultra-low-Z (with xformer) rigs?

And what's the best affordable computer-based setup for such investigations? Me & my Extech, sine wave generator & scope (with resistive impedance measuring hookup between the last two devices) are starting to feel our age.

Bob Palmieri

[Joe Gwinn]

What's worth measuring these days, in both conventional multi-turn hi-Z and also these pioneering single-turn ultra-low-Z (with xformer) rigs?

And what's the best affordable computer-based setup for such investigations? Me & my Extech, sine wave generator & scope (with resistive impedance measuring hookup between the last two devices) are starting to feel our age.

Well, with a scope and a signal generator, one can measure impulse response functions directly. I used 10 microsecond pulses of at least 10 volts peak (into 50 ohms), the max my generator could provide. Failing 10 volts, one can cobble a one-transistor amplifier to generate a big enough pulse. In prior universes I used a 2N4404, if memory serves. The generator drives a small drive coil that is placed where the strings would be, so the pulse has to get through the cover, if there is one. The per-shot induced signals were quite clean looking, and I would average 512 responses together, to largely cancel out noise in the scope front end. The scope was triggered directly from the generator via a dedicated sync path, so things didn't dance around.

If the impulse response is a single cleanish peak, the pickup will sound clear, if the peak is narrow enough If too wide, sound will be dull. If there are audio-frequency wiggles after the peak, the pickup will sound chimey. And so on.

[bbsailor]

Well, with a scope and a signal generator, one can measure impulse response functions directly. I used 10 microsecond pulses of at least 10 volts peak (into 50 ohms), the max my generator could provide. Failing 10 volts, one can cobble a one-transistor amplifier to generate a big enough pulse. In prior universes I used a 2N4404, if memory serves. The generator drives a small drive coil that is placed where the strings would be, so the pulse has to get through the cover, if there is one. The per-shot induced signals were quite clean looking, and I would average 512 responses together, to largely cancel out noise in the scope front end. The scope was triggered directly from the generator via a dedicated sync path, so things didn't dance around.

If the impulse response is a single cleanish peak, the pickup will sound clear, if the peak is narrow enough If too wide, sound will be dull. If there are audio-frequency wiggles after the peak, the pickup will sound chimey. And so on.

Joe,

I have done some work on Pulse Induction metal detectors and the term "time constant" (TC) is used frequently to define the search coil and target electrical characteristics. However, in guitar pickups, the term "time constant" is rarely mentioned. Assume a SC pickup is 2H at 6000 ohms, then the SC pickup TC would be 2/6000 or .0003333 seconds for the pickup current to rise to about 63 percent of maximum. How does the pickup coil's TC affect the impulse response test results?

Thanks

Joseph J. Rogowski

[Joe Gwinn]

I have done some work on Pulse Induction metal detectors and the term "time constant" (TC) is used frequently to define the search coil and target electrical characteristics. However, in guitar pickups, the term "time constant" is rarely mentioned. Assume a SC pickup is 2H at 6000 ohms, then the SC pickup TC would be 2/6000 or .0003333 seconds for the pickup current to rise to about 63 percent of maximum. How does the pickup coil's TC affect the impulse response test results?

I've played with Pulse Induction as well. Their coils are typically critically damped, so they won't ring at all, and will shut off quickly. The time constant being discussed in the PI world describes the ringing in the metal object to be detected. One wants the drive pulse (or collapse of the magnetic field) to be far faster than the decay rate of the induced ringing, so the ringing can be detected.

Transposing to the pickup world, one wants the drive coil to be critically damped as well, but the pickup coil is what it is. The ringing will come from any nearby metal, including covers, slugs, magnets, baseplates, metallic pickguards, etc. Anything with eddy currents.

The LR time constant of the pickup isn't usually of much interest, because the amplifier is high impedance, so little current flows. More important is self-resonant frequency, perticularly when loaded with some cable capacitance, pulling the resonance down near the frequencies of guitar notes.

The impulse response function contains everything electrical known about the pickup, unlike the power spectrum (which ignores phase). Which is better? Depends on what one is trying to do. It's pretty easy to measure both.

[Mike Sulzer]

The impulse response, or equivalently, amplitude and frequency response, is possibly much more important for guitar pickups than in most audio applications. Why? The ear does not respond to phase, at least at relevant speeds, but the electric guitar is generally played through non-linear electronics. Thus, waveforms with the same power spectrum but different impulse responses can produce different harmonics, converting waveforms possibly indistinguishable if processed linearly into audibly distinct waveforms after the non-linearity. That said, however, most pickups have nothing unique about their impulse responses: in practice, amplifier and effect non-linearities tend to decrease differences rather than increase them.

I've played with Pulse Induction as well. Their coils are typically critically damped, so they won't ring at all, and will shut off quickly. The time constant being discussed in the PI world describes the ringing in the metal object to be detected. One wants the drive pulse (or collapse of the magnetic field) to be far faster than the decay rate of the induced ringing, so the ringing can be detected.

Transposing to the pickup world, one wants the drive coil to be critically damped as well, but the pickup coil is what it is. The ringing will come from any nearby metal, including covers, slugs, magnets, baseplates, metallic pickguards, etc. Anything with eddy currents.

The LR time constant of the pickup isn't usually of much interest, because the amplifier is high impedance, so little current flows. More important is self-resonant frequency, perticularly when loaded with some cable capacitance, pulling the resonance down near the frequencies of guitar notes.

The impulse response function contains everything electrical known about the pickup, unlike the power spectrum (which ignores phase). Which is better? Depends on what one is trying to do. It's pretty easy to measure both.