Is this snake oil? Dave Pensado says it isn't. Power cables

[dmills]

Yes but it's not a grounded shield; it's not part of the electrical circuit. Is it??

Actually in a typical UK TN-C-S (MEN, PME, whatever) install it is part of the electrical circuit, specifically the armour carries the neutral current (It is connected to the house 'ground' conductor at the supply companies termination point in each house).

but don't really know how it's set up in a balanced line. Even mic manufactures sometimes get that wrong.

Yea, annoying that, a point Jim Brown (Who I would listen to about cables, but probably not about how to mix a hit record) made well.

I have a fairly simple view of any sales guff mentioning the word Quantum, if it does not come with at least a few pages of latex math markup they are trying to snow you (And if it does come with the math they are probably still trying to pull a snow job).

So ya, science folk get cranky that this rubbish is still around. Because it's the same problem that causes anti vaccine scares. It is a problem worth being passionate about.

Amen.

[Hugh Robjohns]

Is it possible to delete posts on SOS forums (can't find a delete option)?

It is possible for moderators to delete posts, but not for members.

You can edit your post up to 45 minutes after its initial posting, but you can't remove it completely, and after that time period it becomes read-only.

The reason for this is to avoid confusion when other people start re-quoting your posts in their own replies.

If you really want a post to be deleted, just PM one of the moderators and explain what, where and why, and if it is appropriate to delete a post and it won't disrupt the rest of the thread, we'll be happy to do that for you.

H

[Tomás Mulcahy]

And if there isn't a measurable effect, we can still explain it with psychology...

Being wrong is human!

Actually in a typical UK TN-C-S (MEN, PME, whatever) install it is part of the electrical circuit, specifically the armour carries the neutral current (It is connected to the house 'ground' conductor at the supply companies termination point in each house).

Ah, OK. Thanks for setting me right

Jim Brown (Who I would listen to about cables, but probably not about how to mix a hit record) made well.

Is this him:
http://www.audiosystemsgroup.com/publish.htm
?

I have a fairly simple view of any sales guff mentioning the word Quantum, if it does not come with at least a few pages of latex math markup they are trying to snow you (And if it does come with the math they are probably still trying to pull a snow job).

Your version of Godwin's law

[dmills]

Is this him:
http://www.audiosystemsgroup.com/publish.htm?

Yea, that's him, see also Whitlock, Tony and Muncy.

Your version of Godwin's law

Pretty much yea, also anyone wittering about Tesla unless discussing Three phase machines or quarter wave transformers can probably be safely tuned out, nothing of value will be lost.

[The Red Bladder]

I donno Mike, I got one for my kettle and the brews now have a subtle high to them that wasn't there before. It's like the caffine stage has this wide separation that delivers a more balanced yet defined clarity to the taste buds.

That's settled it - my next computer will be from Scan!

[Martin Walker]

Quite right too - that most definitely deserves a sale!

Martin

[One Horse Town]

I can assure everyone that it is NOT snake-oil. It's bullshit.

However a friend of mine lives in very desirable part of Kensington and apparently the mains current, rather like the people, exhibits a much better class of characteristics.

[Folderol]

I can assure everyone that it is NOT snake-oil. It's bullshit.

However a friend of mine lives in very desirable part of Kensington and apparently the mains current, rather like the people, exhibits a much better class of characteristics.

Indeed so. It's not just a common sine wave but a proper sinusoidal waveform.

[IvanSC]

The two following charts are comparing 18-guage cable with the special one. 18-guage is equivalent to a 1mm^2 cable which is rated at 10 amps. Standard IEC mains cables use 1.5mm cable or 14 guage... which is what the special cable equates to. Funny that...

The Pro version uses even thicker crosssectional wires, so it's not even slightly surprising that the measured impedance is lower... but when the IEC cable is thicker than the cable in the wall there can be no further improvement in supply impedance because it is not forming the bottleneck.

The bottom graphs are illustrating the power factor effect, all of which is well understood and completely irrelevant if the power supply is engineered correctly. Of course the current from the power supply is modulated by the audio signal... how else can the signal pass through the equipment? The only relevant issue is whether the power supply can sustain the current demands... and if it can't it is inadequate to the task and therefore shouldn't be employed for the tests!

H

This was exactly MY initial reaction after sitting through the whole video.
I routinely use mains connection cables with 2.5mm sq cable "just in case" on stuff that draws significant current but they don't cost me thousands of dollars.

[Kevin Nolan]

You've misquoted in your post - I didn't say what's in the quoted box associated with my name (I'm clueless on all of this in truth!!).

[Hugh Robjohns]

Fixed the mis-quote.

[DC-Choppah]

I am just wondering, what is the repeatability of the analog audio gear over the measurement interval - starting from a cold start.

I would like to see the same test done where the gear is left overnight. Then you repeat the Dave P. test where they hear a difference. I believe Dave. he has ears.

Next day, do the test again, but this time, don't change the power cord. Just fire up the gear from cold, play the tune once, take the time it would have taken to swap power cords, then listen again. It will be hot for the second time. Please post the results. Can Dave hear the difference without actually changing the power cord? Or is the thermal state the issue here?

You could also control for the thermal effect by sometimes doing it backwards. Play the song with the good power cable first, then play with the stock power cable. Does order matter?

Analog gear is temperature dependent and is designed to work at steady state thermal equilibrium. That means it heats up and then the temp stays constant after a while. Just like your car's engine. Electronics are designed around that equilibrium point.

During that warm up period everything is in flux.

That Dave (and his other producers) can hear that stuff is awesome.

Personally, I really liked the song and felt inspired by it in both cases. The difference between the two audio files seems completely irrelevant to me regardless of the interesting physics and engineering involved.

[Hugh Robjohns]

You are quite right about the variances that can manifest in analogue (and sometimes digital) equipment with temperature changes. Indeed, I mentioned this myself earlier on in the thread in relation to some misleading cable-comparison tests with which I was involved some years ago.

Also, while I have as much respect for the acuity of Pensado's ears and his aural talents as you clearly do, it would be foolish to think his listening skills can't be derailed just as easily as anyone else's by confirmation bias and other psychological flaws in the right circumstances. We are all frail humans and our sense of hearing is neither absolute not unconditionally reliable in anyone.

Clearly this mains cable has a very different construction from standard cables: it doesn't employ inter-twined cores for the Earth, Live and Neutral wires, and it does have an overall grounded shield. These attributes -- especially the grounded shield -- will have a substantial effect on both radiated emissions generated by the equipment power supply or being carried on the mains supply, and also on the received interference which may be injected into the equipment power supply and/or local mains distribution.

This is all perfectly well-known science and engineering, and none of it involves or invokes the mysteries of theoretical quantum physics.

I, personally, don't doubt that Pensado heard a difference, and while I haven't tested the files myself, I don't doubt they show a difference either. I've heard and experienced similar things myself often before.

What I do doubt -- and despise, actually -- are the claims made of the reasons for these audible differences. Although presented as a scientific test it was quite obviously nothing of the sort. There are so many blatantly obvious unknowns and ignored variables involved that the whole thing is utterly farcical.

The single most likely explanation for the reported audible differences is radiated and/or RF interference -- something that is incredibly easy to measure and assess given the right tools -- but which wasn't done. And it's no wonder why not, since such tests would almost certainly have revealed that under-performing equipment was the real issue here, not a magical cable.

Having said that, we all live in the real world, and 'under-performing equipment' is a fact of life, especially at the budget end of the scale. So if a shielded mains cable helps some under-performing system to perform better by excluding interference, that's probably a good thing. Some people will undoubtedly be prepared to pay well over the odds for a 'special mains cable' in the satisfaction that it makes their own specific system perform better.

For me, though, it would be far better to understand the issues and encourage manufacturers not to cut corners by investing in properly engineered equipment that does not create, and is not susceptible to, external interference, such that over-priced shielded mains cables are not required in the first place!

It's exactly the same thing story with expensive master clocks: converters with seriously dodgy clocking circuitry being made to sound less bad through the more pleasant distortions introduced by external clocking!

It's utterly bonkers... but many just don't get (and don't want to know about) the real-world engineering and much prefer the psychology of 'golden-ears' magic....

H

[The Red Bladder]

it would be far better to understand the issues and encourage manufacturers not to cut corners by investing in properly engineered equipment that does not create, and is not susceptible to, external interference, such that over-priced shielded mains cables are not required in the first place!

It's exactly the same thing story with expensive master clocks: converters with seriously dodgy clocking circuitry being made to sound less bad through the more pleasant distortions introduced by external clocking!

It's utterly bonkers... but many just don't get (and don't want to know about) the real-world engineering and much prefer the psychology of 'golden-ears' magic....

This!

A customer came to us with a whole rack full of 'magic' nonsense, such as power conditioners, magic cable and external clocks. He told me that without all that gubbins, his converter box (from a famous UK manufacturer of such things at a premium price!) sounded edgy at the top end (i.e. HF distortion) and it occasionally lost sync and crashed his system.

I took the offending box into the workshop and opened it up. It relied on the shielding on the audio cables for its chassis/earth! I put in a decent sized cable from the chassis buss on the main board to the housing and all those problems of distortion and crashing vanished.

He thought I was some kind of electronics genius! All I did was indulge in some basic common sense. But why the manufacturer sought to save on one short bit of cable baffles me!

[DC-Choppah]

But also, looking at the data on the web site, they are describing current spikes on the mains that are being drawn by the switching power supply. So if the inductance (which acts as a low pass filter) of the power line is increased, the switching power supply could become starved for current if there is too much inductance in the power cord.

While you have to live with the inductance of the wires in the walls, the power cord can add inductance (in series) to the line, especially if it is flexible and cheap and goes in a loop, or a curve. It could be the limiting factor with regards to the inductance that the power supply sees.

Minimizing inductance in a switching power supply seems to be known to be helpful even for high end gear. As I understand it, the switching power supply can get starved for refill current due to this inductance and you get voltage sag at that moment which modulates the sound. So the effect will depend on the program material. Hence Dave can hear it. And it would cause distortion, voltage droop, everything they described.

So I will assume that even the A&H mixer might benefit from a lower inductance power cord even if it has a 'perfect' power supply in all other respects.

What is interesting then, is this additional inductance can be minimized by simply adding a piece of the same stiff and straight Romex type wire in the wall from the plug to the equipment in place of the flexible power cord. Basically, make the power cord no longer the weak link in the series inductance chain.

It looks like the whole battle ground for these cable is how to make them equal to a stiff piece of Romex, but also be flexible.

Once you minimize the series inductance beyond what the wire in the wall has, then the power cord improvement has no more effect.

BTW the rules for reducing the inductance of a wire are:
Reduce its length (make it short)
Increase the bend radius (make it straight)
Increase the number of a parallel wires (make it into a bunch of wires)

These describe the design of the cable.

I suspect that the stiffness of the expensive cable is such that it prevents you from bending it into a tight corner. Meanwhile the stock cable can be bent into a tight radius.

[Folderol]

But also, looking at the data on the web site, they are describing current spikes on the mains that are being drawn by the switching power supply. So if the inductance (which acts as a low pass filter) of the power line is increased, the switching power supply could become starved for current if there is too much inductance in the power cord.

This is complete and utter bullshit - sorry, there is no other possible description
It shows total lack of understanding of how a switchmode PSU actually works.

[DC-Choppah]

There are regulations in effect regarding the harmonic content of the mains current draw from switched power supplies: https://en.wikipedia.org/wiki/IEC_61000-3-2

Note that these are given as harmonics of the (60 Hz) fundamental. The current draw from a 'chopper' supply is just as shown on the website, namely happening in little pulses - the current has been 'chopped'.

As described on the wiki site regarding switched power supplies (https://en.wikipedia.org/wiki/Switched- ... wer_supply)

Simple off-line switched mode power supplies incorporate a simple full-wave rectifier connected to a large energy storing capacitor. Such SMPSs draw current from the AC line in short pulses when the mains instantaneous voltage exceeds the voltage across this capacitor. During the remaining portion of the AC cycle the capacitor provides energy to the power supply.

As a result, the input current of such basic switched mode power supplies has high harmonic content and relatively low power factor.

Hence the whole reason for the regulations of various classes of equipment. This is a well-known phenomenon in switching power supply design.

Since the current is being draw in bursts with harmonic content at the higher frequencies given in the regulation table, the power line impedance (mainly series inductance) is going to be an issue if these harmonics can be supported or not. If the waveform of the switching power supply changes, it is going to modulate the signal.

One just has to look at the plots on their web site and see the pulses in current. Assume you would measure the same thing if you hooked up a voltage and current measurement O-scope to that mixing console. I believe the plot. Why not?!? It's an easy measure to make yourself if you don't believe it. But don't try to do it with an AC meter that reports RMS power of a sine wave. You have to look at the waveform.

And the calculation of the impedance of a cable is simple and well known. So it is easy to see that if the power cord can not support (due to its series inductance) the current draw waveform, then the PSU will have some effect on the signal output. The PSU designer must assume a certain amount of impedance on the input at these higher harmonics. If we exceed that by adding a series inductance, then it will have an effect.

I think the only controversial thing here is that folks never realized that their gear was drawing current in these little bursts on the main power line. The power line folks surely knew so they made regulations since it costs them money to have PSUs with low power factors (highly chopped current pulses).

Of course there are PSUs and amplifiers that have a power factor of 1 by design. But apparently the AH mixer board that they measured does not.

Seems that most BIG equipment (big consoles, big amps) are more prone to this effect since they draw more power and to give them a 'clean' supply with power factor of 1 would make the supplies big and heavy.

I trust Dave's ears. So I looked into this.

Makes sense to me.

Cool!

[Folderol]

Power factor correction is used because current and voltage go out of phase if a load is essentially either inductive of capacitive. A SMPSU is principally capacitive. That's why Electricity companies don't like them. They don't like dodgy waveforms much, but the substation distribution transformers tend to smooth them out rather well.

I'm very familiar with the pulse nature of these PSUs - I've worked with them for years, and while I agree they can throw out a lot of nasty crap if poorly designed. They can easily feed back into sensitive parts of the kit they are powering, but the idea that they are being starved of current is still rubbish. Apart from which the very inductance that is supposed to be so devastating in the mains cable would (if it was actually a significant amount) actually tend to average out the pulses (inductors store energy too, you know). When you consider that a typical PSU is rated at 90-250V AC, and you're running it at 230, you'd need some serious inductance to drop 140V!

In the days of valve radios they used inductors to smooth the ripple after rectification. Typically these were 20 Henrys, and consisted of very many turns of wire on an iron core. A couple of metres of mains cable in free air will have an inductance in micro-Henrys.

P.S.
Large industrial power supplies and inverters have extensive filtering. People think it's to protect them for mains problems. It's not. It's to protect everything else from them.

[DC-Choppah]

But his measurements are showing that the current pulses are there when plugged in and operating. That means that the power company's transformers have NOT smoothed them out. And the small amount of inductance that is in the wiring and the power cord is also not enough to smooth them out. So the demand on the current supply is a signal that contains all those higher harmonics. We assume that the mixer meets the regulations for the 40th order harmonic series.

If we believe his impedance measurement for the power cord where he measures 60 mOhms for a 2m length at 120 Hz, then if I want to see how much of that is inductance, I can use this calculator: https://ampbooks.com/home/amplifier-cal ... nductance/.

I get about 2 mOhms of reactance from 3 microHenries of inductance. And 42 mOhms of resistance.

Bend the wire into a loop though, and that increases the inductance to 10 microHenries.
Now the reactance is about 4x larger. That means that the impedance curve is going to be more frequency dependent now that more of the load is reactive. (https://www.eeweb.com/toolbox/loop-inductance/)

OK so far. But what is the input impedance of the switching power supply? This guy measured one. https://www.youtube.com/watch?v=RDPWtNJX3uM&app=desktop

Since the PS is capacitive, the impedance goes down to zero as the frequency increases. He measures almost 0 Ohms at 10kHz.

So that means for the higher order harmonics of the switching PS waveform, they have a small but INCREASING impedance as plotted on the power cord web site. Meanwhile, the PSU input has a much larger (relative to the power cord) input impedance at 60 Hz, but eventually, there will be a harmonic where the two impedances become closer together.

When the impedance start to get closer together, they form a voltage divider where the impedance of the cord must start to matter.

I agree that for the fundamental (60 Hz) the impedances are so different (say 4Ohms for the PSU and 60 mOhms for the cable) that it can't matter.

But for the higher order harmonics, things are conspiring so that at some point that harmonic will have a dependence on the cable impedance. That means that wrapping the cable into a circle WILL change the magnitude of the higher order harmonics in the current waveform. But only for the highest order harmonics as we have calculated.

Dave P can hear this stuff !!?!!??

[Folderol]

Bend the wire into a loop though, and that increases the inductance to 10 microHenries.
Now the reactance is about 4x larger. That means that the impedance curve is going to be more frequency dependent now that more of the load is reactive. (https://www.eeweb.com/toolbox/loop-inductance/)

That would be true if it was a single core, but it's already a squashed loop with live and neutral in antiphase, so effectively a bifilar winding - a well known technique used to get very low inductance wire-wound resistors used at high frequencies.

I'm not arguing the guy can't hear something, just not what he thinks.

[ef37a]

Yes Will, had this conversation with a guy on a guitar forum a year of so ago. The thread was about the need (or not) to unroll mains extensions.

I said do it of course because of the dissipation but the matter of inductance was raised and it took me a time to convince him that a mains cable on a reel (or a speaker or mic cable for that matter) was essentially non-inductive.

The matter of third party protection was familiar to me in a very small, low power way with the one of the first all solid state colour TVs by RBM. This used a single 1/2 wave thyristor regulator,( basically a light dimmer with feedback!) to regulate the 180V for the line stages.

It chucked the negative half cycle away but it found its way back up the mains supply and caused no end of trouble. The Electricity Board (as they then were, Bless 'em) had a moan and wee had to retrofit a toroid doughnut in the live supply as a mod, 'bout the size of a small tin of Heinz beans and as heavy as a big one!

A later chassis from Philips, the G11 solve the issue by using a pair of thysistors in a full wave circuit. All these designs of course had the chassis at neutral (you hoped!) or at 1/2 mains V in and TWO pin mains plugs were still quite common!

Dave.

[DC-Choppah]

Bend the wire into a loop though, and that increases the inductance to 10 microHenries.
Now the reactance is about 4x larger. That means that the impedance curve is going to be more frequency dependent now that more of the load is reactive. (https://www.eeweb.com/toolbox/loop-inductance/)

That would be true if it was a single core, but it's already a squashed loop with live and neutral in antiphase, so effectively a bifilar winding - a well known technique used to get very low inductance wire-wound resistors used at high frequencies.

I'm not arguing the guy can't hear something, just not what he thinks.

I agree that for the normal mode of current through the mains, the power cord is a bifilar winding. Therefore bending the cord into a loop would create no additional inductance. Unfortunately, for these high frequency harmonics that we are concerned with, they can be common mode instead of differential and in that case they would a) be susceptible to inductance due to bending, b) radiate from the cord.

This appears to be well known if one searches for 'common mode noise on power lines due to switching power supplies'. There are a few nice tutorials with lots of measured data. If these signals were differential, then they would not radiate. So the fact that folks are measuring their radiation means that they must not be differential.

This is a very niche subject as we are discussing very high frequency signals on power lines that it appears folks think should be filtered out. But clearly they are not filtered out. The reason, is that they are common mode signals.

[Folderol]

I shall make one last attempt...
This is goalpost moving. Common mode output from the PSU will have no affect at all on it's own performance. The only thing that matters to it is the stability of the voltage across the reservoir capacitor, which will be quite unaffected. Added to which, the fact that it is a capacitor means that there will be no measurable HF across it's terminals anyway.

[dmills]

I shall make one last attempt...
This is goalpost moving. Common mode output from the PSU will have no affect at all on it's own performance.

Clearly true as far as it goes, but real audio toys however tend to have other connections to grounded things.

Now consider a common switchmode line lump power supply (Often Class II), it has a small class Y cap between one of the mains lines and the output (A class I will likely have one leg of the output tied to earth, a different can of wyrms).

The output will probably have a common mode voltage present relative to earth due to interwinding capacitance (which is what this cap is intended to reduce by shunting that HF to neutral), but when you connect some external earth into the box powered by that supply, you create a current loop N->HF Cap->Output cable->Audio box->audio cable->class I device -> earth->neutral, this loop area can be significant.
This current is NOT A PROBLEM if you design everything right and use balanced interfaces with limited input bandwidths, but that cost money.

Basically the designers trade common mode voltage for the potential of a large loop at HF when the rest of the system is considered. For EMC testing the common mode voltage is the bigger problem as it radiates better from a shortish cable.

None of this stuff is easy, and there are ALWAYS tradeoffs, that's why they call it engineering.

It does not much help clarity when we have people who we might trust to mix records but not to design electronics coming up with gibberish explanations.
One would hope that none of this stuff was magic!

It is possible to design entirely satisfactory audio gear using switched mode supplies, I do it routinely (You just need to pay attention to the detail), but combining an external switcher with single ended analogue audio connections leaving the box is just asking for it, I would be very reluctant to go there (You either get shafted by a ground loop with a class I supply, or by HF loop area with a class II).

[DC-Choppah]

I shall make one last attempt...
This is goalpost moving. Common mode output from the PSU will have no affect at all on it's own performance. The only thing that matters to it is the stability of the voltage across the reservoir capacitor, which will be quite unaffected. Added to which, the fact that it is a capacitor means that there will be no measurable HF across it's terminals anyway.

Just to be clear for others who might be trying to follow this, as you say, the power supply looks like a capacitor across its terminals, so no HF is present. But this capacitor is across the hot/neutral wires from the mains.

The common mode current signal is the signal across the neutral/ground instead.

If the ONLY ground is back through the main power cable, then this signal would be differential again on the power cable and would NOT be affected by inductance or radiate from the cable.

If however the current coming in on the neutral wire on the power cord goes to ground through the audio output, then this mode will be single ended on the power cord AND will modulate the signal as heard on the output. It modulates the signal since it's strength is a function of the signal.

All that is needed is to filter the common mode of the mains input and this goes away. I suppose for the AH mixer that was used, those filters would have been too big and expensive so where not included?

I wonder what the extra cost would have been for the extra filters in the mixer, vs the cost of the better power cord which solves the problem a different way?