Power cables - how to measure

[Stereophilus]

25 minutes ago, aussievintage said:

Raising potential doesn't create an impedance, that's an inherent property of the materials and contacts etc.

 

and as for it raising the potential, it can only do that if you already have a bad earth, i.e. one with a significant impedance to ground (in the same way as noise does it, or not).  These neutrals only connect to ground back at your switchboard, (and other places it the electricity companies distribution network), so that's your local reference for everything in your household.  Unless you happen to have had your own dedicated clean earth created. It can be a real can of worms...

Sorry, I mean power supplies that connect ground to neutral using a method that creates impedance, such as a resistor or inductor, thereby also lifting their ground potential.

[bob_m_54]

8 hours ago, Stereophilus said:

These high frequency EMI products from, let’s say “dirty” SMPS, propagate or radiate along AC circuits and diminish in amplitude according to the inverse square of distance from their origin?

No, they travel along mains circuits quite well, but filtering at the AC mains input of the device is used to remove it there. My point was about them being radiated from the power cable, which would be more noticeable than any mains 50Hz EMI

[MLXXX]

8 hours ago, Stereophilus said:

However, as we know, a lot of power supplies connect ground to neutral,

I am curious to know what kind of "power supply" you're referring to here, e.g. the consumer's mains switchboard broadly considered as a "power supply", or the power supply circuitry inside an item of audiophile equipment?

 

I note it is illegal (non-compliant) in Australia for a household appliance to effectively make a direct connection between the neutral pin of a power point and the earth pin of that power point. (In certain fault situations this could result in a lethal voltage arising on the metal exterior of an appliance.)

Edited October 7, 2020 by MLXXX

[aussievintage]

8 hours ago, Stereophilus said:

Sorry, I mean power supplies that connect ground to neutral using a method that creates impedance, such as a resistor or inductor, thereby also lifting their ground potential.

Equipment should not connect neutral to ground.  

 

1 hour ago, MLXXX said:

I am curious to know what kind of "power supply" you're referring to here, e.g. the consumer's mains switchboard broadly considered as a "power supply", or the power supply circuitry inside an item of audiophile equipment?

 

I note it is illegal (non-compliant) in Australia for a household appliance to effectively make a direct connection between the neutral pin of a power point and the earth pin of that power point. (In certain fault situations this could result in a lethal voltage arising on the metal exterior of an appliance.)

Exactly.

[Stereophilus]

1 hour ago, aussievintage said:

Equipment should not connect neutral to ground.  

2 hours ago, MLXXX said:

I am curious to know what kind of "power supply" you're referring to here, e.g. the consumer's mains switchboard broadly considered as a "power supply", or the power supply circuitry inside an item of audiophile equipment?

 

I note it is illegal (non-compliant) in Australia for a household appliance to effectively make a direct connection between the neutral pin of a power point and the earth pin of that power point. (In certain fault situations this could result in a lethal voltage arising on the metal exterior of an appliance.)

I’m not saying it should. Just that some appliances do.  I’m not talking about hifi gear generally, more things like TVs and fridges.  A lot are mass produced with this electrical setup.  There is effort involved in making something compliant to Australian standards and some don’t bother.

 

Anyway, I think we have digressed far enough.  We should try to stay on topic.  Apologies.

[MLXXX]

(Thanks, @Stereophilus. As you suggest, let's get back on topic...)

 

Looking back at all the remarks about possible adverse effects from spurious signals, I wonder whether a more direct way to address the problem of "Power cables - how to measure" would be to set up a suite of "torture tests", and then measure the extent of vulnerability/susceptibility of different pieces of audiophile equipment; and finally test to see whether substituting a different power cord could in part or in full mitigate the adverse effects found to arise when operating the equipment under non-ideal conditions..

 

I envisage that we would start with the assumption that under ideal conditions of perfect mains power at the power point, and no extraneous electromagnetic interference coming from any source, that we would expect an item of audiophile equipment (such as a phono preamp, a DAC, a CD transport, or a power amplifier) to operate properly in accordance with its specifications, simply using the power cord it came with.

 

We could then postulate that as the quality of the mains power were made to depart more and more from perfection, or extraneous electromagnetic interference were introduced at higher and higher intensities, that a point would be reached of a measurable degradation in performance of an item of audiophile equipment. 

 

We could then test for whether switching from the original power cord to an after-market power cord could reduce the adverse effects found to arise under the non-ideal conditions.

 

What is "perfect"  single-phase household mains power in Australia? Answer: a 230 volt RMS sinusoid at 50Hz with negligible harmonic distortion and very low source impedance.

 

What is "extraneous electromagnetic interference"? That could be more difficult to pin down in terms of making its spectrum and amplitude realistic as an emulation of possible real world conditions. Perhaps a number of different types of interference would need to be generated, and the effects tested for.

 

I note as general observations:

 

1. Power points in the home tend to have a source impedance exceeding 1 ohm because of the appreciable length of cabling from the power point to the domestic switchboard, and the further length of cabling from the switchboard to the street.  You can calculate the source impedance of an installed double socket power point by connecting a large resistive load  to one socket (e.g. an electric radiator) and a multimeter set for voltage measurement to the other socket and noting the drop in measured voltage when the resistive load is switched on.

 

2. Audiophile equipment if it is to be compliant should emit only low levels of interference if any, and if it is well designed will be resistant to small levels of incoming interference.  On the other hand, placing sensitive equipment immediately adjacent to power cables or other sources of interference could well affect the sensitive equipment.

Edited October 8, 2020 by MLXXX

[Ittaku]

32 minutes ago, MLXXX said:

Looking back at all the remarks about possible adverse effects from spurious signals, I wonder whether a more direct way to address the problem of "Power cables - how to measure" would be to set up a suite of "torture tests", and then measure the extent of vulnerability/susceptibility of different pieces of audiophile equipment; and finally test to see whether substituting a different power cord could in part or in full mitigate the adverse effects found to arise when operating the equipment under non-ideal conditions..

No. Go back to the original question. Just check that changing the power cord has any effect whatsoever on the output first. If it doesn't then it really makes no difference whatsoever how effective they are at mitigating what is likely not a problem at all.

[MLXXX]

9 minutes ago, Ittaku said:

No. Go back to the original question. Just check that changing the power cord has any effect whatsoever on the output first. If it doesn't then it really makes no difference whatsoever how effective they are at mitigating what is likely not a problem at all.

A problem I see with that is that the OP envisages a series of tests over time as the mains voltage and/or harmonic content and/or other spurious content vary.

 

That could take a very long time to do, would not be repeatable (as mains power fluctuates randomly), and would tend to be inconclusive I fear. It puts together system performance changes under stress (e.g. wrong voltage, excessive interference) with system performance using different power cables.

Edited October 8, 2020 by MLXXX

[Ittaku]

Just now, MLXXX said:

A problem I see with that is that the OP envisages a series of tests over time as the mains voltage and/or harmonic content and/or other spurious content vary.

 

That could a very long time to do, unfortunately be inconclusive,  and puts together system performance changes under stress (e.g. wrong voltage, excessive interference) with system performance using different power cables.

What? The point of this is to find whether power cables make a difference or not. Who cares if they filter random crap or have different capacitance or resistance or gauge and fix artificial measurements if ultimately it makes no difference to audio? That's like saying power cables are different - of course we already know that they're different, but does it matter?

[sir sanders zingmore]

9 minutes ago, Ittaku said:

What? The point of this is to find whether power cables make a difference or not. Who cares if they filter random crap or have different capacitance or resistance or gauge and fix artificial measurements if ultimately it makes no difference to audio? That's like saying power cables are different - of course we already know that they're different, but does it matter?

 

Shouldn't we establish whether or not power cables make a difference to audio by listening (under properly controlled conditions). If we establish that there is an audible difference, only then should we try measuring electrical changes etc.

 

Otherwise it's like science trying to come up with a theory for a physical phenomenon that no one has actually observed.

[Ittaku]

1 minute ago, sir sanders zingmore said:

 

Shouldn't we establish whether or not power cables make a difference to audio by listening (under properly controlled conditions). If we establish that there is an audible difference, only then should we try measuring electrical changes etc.

 

Otherwise it's like science trying to come up with a theory for a physical phenomenon that no one has actually observed.

Not to me. I'm quite certain that if there's no measurement known to mankind that shows a difference, then there won't be an audible difference (only a psychological one.)

[MLXXX]

30 minutes ago, Ittaku said:

What? The point of this is to find whether power cables make a difference or not. Who cares if they filter random crap or have different capacitance or resistance or gauge and fix artificial measurements if ultimately it makes no difference to audio? That's like saying power cables are different - of course we already know that they're different, but does it matter?

Don't shoot the messenger!

 

I am responding to the technique the OP suggested of using real world mains power over extended periods of time and trying to compare power cords that way; and to the numerous comments in this thread about interference.

 

I think it is likely to be inconclusive and time consuming to do long term testing involving random fluctuations in the mains.  I think the power cords would measure the same within the level of experimental error, and at the end of the day one would not have achieved a lot.

As for making direct measurements of difference in performance in A B testing in one session, with mains voltage constant or near constant, and with typical relatively low levels of interference with typically resistant good quality equipment, and avoiding running signal cables right next to power cables, then I myself would expect no measurable difference. If there were a measurable difference in such a straightforward setting there would be published results already indicating that!  The posts in audiophile forums about power cords making a difference come from those reporting hearing a difference. I don't  recall ever seeing a measured difference reported other than in a video on YouTube but it turned out in that case that a reflector near one of the speakers had been inadvertently moved when changing the power cord, thus invalidating the experimental finding. 

 

One scenario where I could confidently predict that changing a power cord  could make a measurable difference would be a very powerful power amplifier delivering a sinewave at the onset of clipping where substitution of a slighter lower resistance cord could allow a few more millivolts for the DC supply to the output transistors. This effect would tend to be diluted considerably by the power point source impedance (much higher than the power cord resistance), the transformer winding resistance (typically higher again) and the rectifier forward conduction resistance.  And the few additional millivolts of rectified DC  would tend to be swamped to a vary large degree by random variations in the mains voltage, if testing over a longer period of time!

Edited October 8, 2020 by MLXXX

[Ittaku]

2 minutes ago, MLXXX said:

As for making direct measurements of difference in performance in A B testing in one session, with mains voltage constant or near constant, and with typical relatively low levels of interference with typically resistant good quality equipment, and avoiding running signal cables right next to power cables, then I myself would expect no measurable difference. If there were a measurable difference in such a straightforward setting there would be published results already indicating that!

Okay fine, but that is what the OP is setting out to measure.

[ray4410]

                                                

[Stereophilus]

2 hours ago, Ittaku said:

No. Go back to the original question. Just check that changing the power cord has any effect whatsoever on the output first. If it doesn't then it really makes no difference whatsoever how effective they are at mitigating what is likely not a problem at all.

Well you can test that, but the conclusions drawn are more limited than you suggest. There are more potential issues involved and part of the remit of this thread is how to explore those variables using a scientific approach.

[bob_m_54]

1 hour ago, Ittaku said:

What? The point of this is to find whether power cables make a difference or not. Who cares if they filter random crap or have different capacitance or resistance or gauge and fix artificial measurements if ultimately it makes no difference to audio? That's like saying power cables are different - of course we already know that they're different, but does it matter?

Totally agree. The scope of the proposed "tests" has gone way beyond what the whole point of the original exercise was anyway. I'm guilty of prolonging some of the off topic digression as well. But it bugs me when people don't understand the basic principles but still profess greater knowledge. Anyway, I've lost interest and don't see anything useful coming from it now. Too many diversions from the aim of the exercise.

[Guest rmpfyf]

2 hours ago, MLXXX said:

A problem I see with that is that the OP envisages a series of tests over time as the mains voltage and/or harmonic content and/or other spurious content vary.

 

That could take a very long time to do, would not be repeatable (as mains power fluctuates randomly), and would tend to be inconclusive I fear. It puts together system performance changes under stress (e.g. wrong voltage, excessive interference) with system performance using different power cables.

 

Well it gets interesting, I thnk is more the point.

 

Voltage variances tend ot be seasonal with excess PV on the grid, frequency variances though tend towards contingency events on the grid - they are, in short, the equivalent of car accidents. I could sample hundreds of time in one day and not get the same mains variances I'd get another. And Australia has had of late some very wide mains frequency tolerances (though recent chances to rules may change that again).

 

Running these tests with some characterisation of the waveform (either statistical or direct capture) is important not just to give some relative relevane to tests but also to reject conditions that are spurious within a test set.

 

Ideally we would have a grid simulator at one end though a nice Regatron machine is a six-figure sum of money. Could be done with a standby generator (e.g. diesel genset) - have done that for some tests at various sites where some facets of mains variance are to be studied coarsely - though the waveform isn't quite as nice as mains, and load sensitivity is greater.

[MLXXX]

1 hour ago, rmpfyf said:

Ideally we would have a grid simulator at one end though a nice Regatron machine is a six-figure sum of money. Could be done with a standby generator (e.g. diesel genset) - have done that for some tests at various sites where some facets of mains variance are to be studied coarsely - though the waveform isn't quite as nice as mains, and load sensitivity is greater.

For a very low current draw device such as a DAC not as demanding [also the DAC current draw could be expected to be relatively steady].  For a high power monoblock amp delivering bursts of full power audio waveforms, more horsepower needed to begin with, and because of the varying load the source impedance would become a more critical issue.   A lot of effort potentially involved, and for mainstream audiophile equipment being fed close to the nominal mains voltage, and reasonably resistant to incoming hash on the mains supply, the prospects of finding a measurable performance difference strike me as bleak.

 

What will people at the extreme ends of the "power cord difference sceptical" - "power cord difference believer" mindset continuum tend to say, if the result is "no statistically significant difference"?  The sceptics will tend to say, "of course, what else would you expect?". The believers might say, "So what? It's what we hear that counts.  Once again measurement has failed.".

 

And if a statistically significant measured difference were found?  The believers would tend to say, "Of course, finally you've been able to measure what we've been able to hear all along!".   And the sceptics might say, "Can we have more detail on your experimental technique and analysis - something appears to be amiss.".  Or they might say, "That equipment appears to be unduly sensitive to interference and does not reflect good design.".

 

In my opinion, it's for these sorts of reasons that power cord discussion on audiophile forums, regrettably, tends to lead nowhere, no pun intended.*

 

_________

 

* Will investigation of a power lead, lead somewhere?  Perhaps if a pun needs to be explained, it's not a good pun!

Edited October 8, 2020 by MLXXX

[Guest Benjet]

2 minutes ago, MLXXX said:

For very low current draw device such as a DAC not as demanding [also the DAC current draw could be expected to be relatively steady].  For a high power monoblock amp delivering bursts of full power audio waveforms, more horsepower needed to begin with, and because of the varying load the source impedance would become a more critical issue.   A lot of effort potentially involved, and for mainstream audiophile equipment being fed close to the nominal mains voltage, and reasonably resistant to incoming hash on the mains supply, the prospects of finding a measurable performance difference strike me as bleak.

 

What will people at the extreme ends of the "power cord difference sceptical" - "power cord difference believer" mindset continuum tend to say, if the result is "no statistically signfificant difference"?  The sceptics will tend to say, "of course, what else would you expect?". The believers might say, "So what? It's what we hear that counts.  Once again measurement has failed.".

 

And if a statistically significant measured difference were found?  The believers would tend to say, "Of course, finally you've been able to measure what we've been able to hear all along!".   And the sceptics might say, "Can we have more detail on your experimental technique and analysis - something appears to be amiss.".  Or they might say, "That equipment appears to be unduly sensitive to interference and does not reflect good design.".

 

In my opinion, it's for these sorts of reasons that power cord discussion on audiophile forums, regrettably, tends to lead nowhere, no pun intended.*

 

_________

 

* Will investigation of a power lead, lead somewhere?  Perhaps if a pun needs to be explained, it's not a good pun!

It's for these sorts of reasons that INSERT NAME OF TOPIC discussion(s) on audiophile forums, regrettably, tends to lead nowhere,

[Guest rmpfyf]

27 minutes ago, MLXXX said:

For very low current draw device such as a DAC not as demanding [also the DAC current draw could be expected to be relatively steady].  For a high power monoblock amp delivering bursts of full power audio waveforms, more horsepower needed to begin with, and because of the varying load the source impedance would become a more critical issue.   A lot of effort potentially involved, and for mainstream audiophile equipment being fed close to the nominal mains voltage, and reasonably resistant to incoming hash on the mains supply, the prospects of finding a measurable performance difference strike me as bleak.

 

What will people at the extreme ends of the "power cord difference sceptical" - "power cord difference believer" mindset continuum tend to say, if the result is "no statistically signfificant difference"?  The sceptics will tend to say, "of course, what else would you expect?". The believers might say, "So what? It's what we hear that counts.  Once again measurement has failed.".

 

And if a statistically significant measured difference were found?  The believers would tend to say, "Of course, finally you've been able to measure what we've been able to hear all along!".   And the sceptics might say, "Can we have more detail on your experimental technique and analysis - something appears to be amiss.".  Or they might say, "That equipment appears to be unduly sensitive to interference and does not reflect good design.".

 

In my opinion, it's for these sorts of reasons that power cord discussion on audiophile forums, regrettably, tends to lead nowhere, no pun intended.*

 

_________

 

* Will investigation of a power lead, lead somewhere?  Perhaps if a pun needs to be explained, it's not a good pun!

 

I have no doubt that varying components and topologies will elicit differing sensitivities. 

 

I'm an engineer and possibly by virtue of this or as a consequence thereof I've never minded too much what people have tended to say. If there is no measurable difference then either the measurement method is not sufficiently accurate, resolute or both, or the phenomena does not exist. I would offer that in audio the former is a very real possibility. A measurement at speed exceeding 16 real bits resolution is a tricky thing to do, and most lab experiements would prefer to be 3 or more bits above the measurement of interest - let alone that what we're trying to evaluate is essentially continuous even if recreated via digital process. 18-bit DAQ kits are pricey. If anything significant is found then all it means is that something significant is found; further insight requires further analysis. It really doesn't may to get too wrapped up in internet forum warrior discussions. We have all been there.

 

I am neither too swayed by what people hear, as it takes a significant amount of temerity to drop a four-figure sum of money on a power cord for belief in performance only to admit that you cannot hear a difference. I'd prefer to measure a lot in addition to qualitative feedback particularly because it requires a very particular skill set to use our ears/brains in a manner devoid of bias. I am not saying it is not possible; a person may trust their own ears or someone else's, and ultimately any person that is happy with what they have is in a good space. 

 

My main mission is to understand a little more of what works enough to put together something that is robust for my use, because I have zero intention of spending significant money. That is not a slight to people that do, I'm simply not in a position to do as much. 

[Guest rmpfyf]

1 hour ago, Benjet said:

It's for these sorts of reasons that INSERT NAME OF TOPIC discussion(s) on audiophile forums, regrettably, tends to lead nowhere,

Don't know. We're doing pretty good here. Have hope.

[Ittaku]

I have a focusrite scarlet 2i2 ADC which is good for 192kHz with true >18 bit resolution measured to the AES17 standard if you wish to borrow it.

[kukynas]

1 hour ago, MLXXX said:

For a very low current draw device such as a DAC not as demanding [also the DAC current draw could be expected to be relatively steady].  For a high power monoblock amp delivering bursts of full power audio waveforms, more horsepower needed to begin with, and because of the varying load the source impedance would become a more critical issue.   A lot of effort potentially involved, and for mainstream audiophile equipment being fed close to the nominal mains voltage, and reasonably resistant to incoming hash on the mains supply, the prospects of finding a measurable performance difference strike me as bleak.

 

What will people at the extreme ends of the "power cord difference sceptical" - "power cord difference believer" mindset continuum tend to say, if the result is "no statistically significant difference"?  The sceptics will tend to say, "of course, what else would you expect?". The believers might say, "So what? It's what we hear that counts.  Once again measurement has failed.".

 

And if a statistically significant measured difference were found?  The believers would tend to say, "Of course, finally you've been able to measure what we've been able to hear all along!".   And the sceptics might say, "Can we have more detail on your experimental technique and analysis - something appears to be amiss.".  Or they might say, "That equipment appears to be unduly sensitive to interference and does not reflect good design.".

 

In my opinion, it's for these sorts of reasons that power cord discussion on audiophile forums, regrettably, tends to lead nowhere, no pun intended.*

 

_________

 

* Will investigation of a power lead, lead somewhere?  Perhaps if a pun needs to be explained, it's not a good pun!

 

precisely.... but for sure there's always benefit of at least trying something even if the result isn't what we expected...

 

 

 

[Warren Jones]

11 hours ago, MLXXX said:

I am curious to know what kind of "power supply" you're referring to here, e.g. the consumer's mains switchboard broadly considered as a "power supply", or the power supply circuitry inside an item of audiophile equipment?

 

I note it is illegal (non-compliant) in Australia for a household appliance to effectively make a direct connection between the neutral pin of a power point and the earth pin of that power point. (In certain fault situations this could result in a lethal voltage arising on the metal exterior of an appliance.)

I think the confusion here is that in the switchboard the neutral and earth bars are connected with an MEN (Multiple Earth Neutral) link. In a fault situation this causes an imbalance in the current flowing in the active and neutral and trips the breaker.

 

Australia uses the MEN system other countries use different typologies

Edited October 8, 2020 by Warren Jones

[aussievintage]

1 minute ago, Warren Jones said:

I think the confusion here is that in the switchboard the neutral and earth bars are connected with an MEN (Multiple Earth Neutral) link. In a fault situation this causes an imbalance in the current flowing in the active and neutral and trips the breaker.

I don't know why this causes so much confusion.  But yes, and that's the only place it is ever connected in the home.  It just means the active voltage is referenced to ground.    Probably would surprise some to know there are SWER (single wire earth return) systems used to deliver power to remote places using only  one wire.  Imagine hifi running from that (actually, other than poor regulation of the power, it would work fine).