Testing/measuring guitar cables

[nickb]

Allright, so another thing come to my mind. If you connect a (10-10000pf) capacitor in series with the cable, you'll have less capacitance so the sound will get back like using shorter cable? Will this be a good idea for sort kind of "passive buffer"?

The problem is not the cable, it's the source impedance. Adding more is not going to help in a constructive way as the price you will pay for flattening the frequency response is (a) significant loss of signal and (b) greater sensitivity to the load ( amplifier input) impedance.

A low source impedance can be got from a very simple battery powered unity gain buffer using a transistor or JFET for under $1. That is the way to go to address cable capacitance and different load impedances.

[boroman]

Allright, so another thing come to my mind. If you connect a (10-10000pf) capacitor in series with the cable, you'll have less capacitance so the sound will get back like using shorter cable? Will this be a good idea for sort kind of "passive buffer"?

[Helmholtz]

For your info: I've ran a test to the 1st post: you can clearly hear the difference that connections in the process are losing its high end by adding capacitance. It's a night and day difference. Buffer makes it identical to the base signal, so agreed, when circuit is buffered, there is no signal lost.

https://www.dropbox.com/s/nqlzsqb74w...20bas.wav?dl=0

1st part: bass > preamp
2nd part: bass > pedalboard (bypassed) > preamp
3rd part: bass > pedalbopard with buffer on > preamp

And that's why you should always have as buffer in your pedalboard, unless you want a real muffled sound.

That's why I like to have a buffered (as opposed to true bypass) pedal like a Tube Screamer as first in my chain.
Exceptions would be a wah-wah or a Fuzz Face which preferably should be connected directly to the guitar.

[boroman]

For your info: I've ran a test to the 1st post: you can clearly hear the difference that connections in the process are losing its high end by adding capacitance. It's a night and day difference. Buffer makes it identical to the base signal, so agreed, when circuit is buffered, there is no signal lost.

https://www.dropbox.com/s/nqlzsqb74w...20bas.wav?dl=0

1st part: bass > preamp
2nd part: bass > pedalboard (effects bypassed) > preamp
3rd part: bass > pedalbopard with buffer on > preamp

And that's why you should always have as buffer in your pedalboard, unless you want a real muffled sound.

[Helmholtz]

You could try it like this. Connect the sig gen between the (guitar?) pickup and ground. Connect the cable under test to the pickup hot and ground. Terminate the other end of the cable with a 1M resistor. Measure the frequency response across the 1M. It should look like the simulation I posted. Add tone and vol pots as required between hot and ground at the pickup end of the cable. It's not buffered because it has the impedance of the pickup in series with the low impedance of the generator.

This is a good way to show the cable's effect on PU resonance and to compare PUs. Here the generator's source impedance should be low (no series resistor).
But it should be noted that the arrangement doesn't allow to show the PU's real filter frequency response. Reason is that with a field/string driven PU the "signal generator" (EMF) lies not in series with the PU's self-capacitance, but only in series with the inductance.
The different placement of the source in the test proposed separates the capacitances, so they won't directly add as with the real PU. The result is a frequency response that differs from simulation results and real PU measurements. Especially it produces an artefact: an addiional series resonance above the PUs main resonance (with external capacitance), which Merlin Blencowe in his book calls the PU's "self resonance". It's not real.

[Dave H]

I saw some test with p-bass pickup wired in series with (maybe) 100k resistor - but it's still - an active circuit. Like a buffered circuit, right?

You could try it like this. Connect the sig gen between the (guitar?) pickup and ground. Connect the cable under test to the pickup hot and ground. Terminate the other end of the cable with a 1M resistor. Measure the frequency response across the 1M. It should look like the simulation I posted. Add tone and vol pots as required between hot and ground at the pickup end of the cable. It's not buffered because it has the impedance of the pickup in series with the low impedance of the generator.

[nickb]

But this is still "active" driven input (active generated signal + resistor & inductor)...
I saw some test with p-bass pickup wired in series with (maybe) 100k resistor - but it's still - an active circuit. Like a buffered circuit, right?

Form the other side:
Q: If the cables are: pure silver, pure gold or silver plated and they have the same capacitance as regular copper cable - will they sound the same?

It's best you forget the word 'active' as that is not the property that make the difference. What matters is the impedance of the source be it a guitar pickup, a pedal buffer or a 10,000 watt amplifier. If the impedance it high the signal through the cable will be more affected by the capacitance of the cable. An impedance of suggested of 100k or more should be enough to allow you to hear, or better measure, the difference. You could use an audio source instead of a signal generator to drive the resistor for a more subjective approach.

A resistor is an especially simple kind of impedance so the result is a loss of higher frequencies. A pickup is a complex impedance as it contains resistance, inductance and capacitance with the result that you still generally get a loss of HF but there are bumps in the response due to responses.

The metals used in the cable make no difference to the capacitance and at audio frequencies will not have any effect on the sound either. It's the geometry of the conductors and the type is insulating materials ( called the dielectric) used that affects the capacitance per unit length.

Give that you have found that capacitances don't have that big a range from cable to cable, your focus should be on the other factors that I mentioned in post #2. Screening and handling noise are the most important features.

[boroman]

Measuring frequency response would require a high impedance source like a signal generator with a series resistor of at least 100k feeding the cable. The signal at the other end of the cable should be measured with equipment having an input impedance of 1M (and low input capacitance).
This measurement will give a low pass filter response with a -6dB/octave slope above the corner frequency. Any anomalities/differences should show as deviation from the ideal slope.
When performing such measurements I never found any differences that could not be explained by the cap value.

To show the interaction of the cable with a PU, the PU would need to modelled using discrete inductor and resistor components as shown by Lemme.

But this is still "active" driven input (active generated signal + resistor & inductor)...
I saw some test with p-bass pickup wired in series with (maybe) 100k resistor - but it's still - an active circuit. Like a buffered circuit, right?

Form the other side:
Q: If the cables are: pure silver, pure gold or silver plated and they have the same capacitance as regular copper cable - will they sound the same?

[Helmholtz]

My question was how to test the "passive" quality of the cable.

I think the main difference isn't really between active and passive operation. Rather it's about total circuit impedance. When one of the cable ends is terminated by a low impedance/resistance, the influence of the cable on circuit performance could hardly show in the audio range.

As a cable mainly acts as a several hundred pF capacitor with small losses, it can be tested like a cap.

Measuring frequency response would require a high impedance source like a signal generator with a series resistor of at least 100k feeding the cable. The signal at the other end of the cable should be measured with equipment having an input impedance of 1M (and low input capacitance).
This measurement will give a low pass filter response with a -6dB/octave slope above the corner frequency. Any anomalities/differences should show as deviation from the ideal slope.
When performing such measurements I never found any differences that could not be explained by the cap value.

To show the interaction of the cable with a PU, the PU would need to modelled using discrete inductor and resistor components as shown by Lemme.

[Dave H]

My question was how to test the "passive" quality of the cable. Since you're not able to generate 20hz-20khz noise spectrum just above your pickup....

For a straight guitar to amp connection it should be enough to just measure the cable capacitance as a measure of "quality" as all cables with the same capacitance will sound the same (ignoring handling noise etc) If you want to have an idea of frequency response it could be modeled in LTSpice as below. The plot is for cables of 300p, 600p, 1200p. I disconnected the guitar's vol and tone controls to show the resonant peaks better. With tone and volume connected the peaks are flatter.

[boroman]

I believe my point was to compare a short cable and a long one, not a cheap one and an expensive one.

Yes, you're right, but when you have active circuit (e.g. after the buffer) the long/short cable also does not matter here (no loss in the signal).
We talk about lengths up to 20 meters or so (I don't know when another buffer should be put it, but I'm pretty sure it might be huge length for the need of another one)
The passive circuit gets benefits from shorter/longer cables and for cheap/expensive...

[Enzo]

I believe my point was to compare a short cable and a long one, not a cheap one and an expensive one.

[boroman]

The longer the cable, the more accumulated capacitance. If you use a six foot cable down to your pedal board, it will matter a lot less than if you use a 20 foot, or even a 30 foot cord. COmpare two 20 foot cords and you might hear the difference in capacitance.

it also matters how you test a cord. if you just plug into the amp with it and dial up a tone, then you never notice that cord A and cord B differ in that one has a treble knob at 4 and the other at 5. I never dial in tone by number, I do it by ear. So if I can dial in a tone, then the cord is OK.

Try this experiment. COmpare two cords of the exact same materials, one 6 foot and one 20 foot.

This is only valid when you have straight guitar>amp setup. If you have any "active" pedal on, like buffer for example, there is no "capacitance" effect and no difference in the sound between using $10 and $300 cable.
My question was how to test the "passive" quality of the cable. Since you're not able to generate 20hz-20khz noise spectrum just above your pickup....

[olddawg]

I’ve done blind tests with “golden ears” musicians and no one could tell the difference with a range of cables in good repair for the most part. The exception might be “curly” cords. I’ve even wired up pedal boards with cut up “vcr” type RCA cables side by side with Mogami and no one can tell the difference. The biggest problem I see with cables is the cheap ones develop more mechanical noise over time. I have no idea how you can test for that.

[nevetslab]

My favorite silent plug is the big Switchcraft #172:

https://www2.mouser.com/ProductDetai...BdlAVtMyIxaKI=

Nice. I haven't seen that one. I'll pick some up once California opens back up, as Pacific Radio is right down the street from my shop....major cable and connector supplier for the Entertainment industry.