Impact of AC Distortion & Noise on Audio Equipment

[Promit]

@amirm I'm almost afraid to ask but ... would you be willing to run these tests on a guitar tube amp? I know it's a wacky scenario but there is no shortage of people hawking power conditioning equipment for those either, and a lot of them still rely on extremely primitive power filtering circuits.

 

[Promit]

Is it only a matter of time until an audio company starts marketing an off-the-grid system including "audiophile grade" solar panels, battery storage and inverter?

Well now I'm curious what the output of my solar inverter looks like.

 

[Angsty]

OK, so I'm going to open myself up to some ridicule here. I bought an APC H-10 power conditioner and surge protector for my new system; paid about $200. I'm aware that it's highly unlikely to have any benefit to the sound quality from the speakers, and its overvoltage protection circuitry is maybe no better than I could get from something I could get off the shelf at Home Depot. But doggone it, those fancy blue lights are so pretty!

$200 for audio jewelry is not so bad.

Here’s the gist of what I’ve read so far:

• For most AC power mains, line noise is not a problem for audio equipment. Even intentionally “dirty” AC sources may not create audible problems.
• For mains with specifically diagnosed problems, purchase a specifically designed solution; a generically designed “power conditioner” is not likely to make your system sound better.
• Surge protection is different from voltage regulation is different from noise filtration. Know your problem (and it’s real impact) before seeking a solution.
• Conversely, don’t buy a “solution” to a problem you don’t have.

 

[tmtomh]

Nope. I'm not arguing that at all. I'm arguing that the tube based devices have so much intrinsic distortion (due to the use of tubes we assume) that they totally mask anything their power supplies may or may not do to the sound. They are dreadful no matter what AC power they are fed with. Thus they don't tell us much more.

I certainly have no problem with the idea that it would be interesting and possibly useful if @amirm were to add another device to this test - one that had poor SINAD, but dominated by noise rather than distortion.

But I still don't see how your objection makes any sense: those tube DACs' poor SINAD is dominated by distortion as you say - but that also means the converse is true, which is that they have relatively low noise (even the 60Hz spikes are much lower than the 2nd and 3rd order harmonics of the test signal). And that - once again - means that if these power "cleaners" had any effect, it would show up as reduced noise below 1kHz in the graphs, and possibly also as reduced random, non-harmonic spikes in the midrange and beyond - just as the measurement of the PS Audio AC cleaner indeed showed some reduction of that random noise in the spectrum.

But that's not what the graphs show.

So your argument might be valid when it comes to those tube DACs perceptually masking any improvement produced by the AC cleaners. But I don't think your argument holds water when it comes to what we can see/read from the measurements - which show no improvement.

What you are basically looking for is a DAC with poor SINAD that's dominated by noise rather than distortion, on the hypothesis that the AC purifiers could reduce that noise. But that assumes that this hypothetical DAC's noise is a result of defective/insufficient AC filtering - and that's precisely the point of Amir's test here: with the possible exception of some as-yet unknown rare outlier, consumer audio devices already have sufficient built-in AC filtering.

 

[tmtomh]

There's a logical fallacy here--that what makes quality is the distortion of the audio amplifier. We don't know if those poor devices had good or bad power supplies, but we are pretty sure they had good power supplies, with respect to their ability to address dirty power, because dirty power didn't affect them.

So, the question isn't "do these devices help?", the question is "are there devices that depend on good power supplies to work well and don't have them?" That question was not explored by Amir's choices.

I agree with Amir as far as busting the myth that all devices benefit from cleaner power (as if they have no ability to do their own cleaning), and that is indeed the myth that is behind the sale of many such devices. But I also agree with John that the test does not support the categorical conclusion that external power conditioners are always a waste of money, because they certainly do solve specific problems in specific cases for reasons that can most definitely be described and measured.

Rick "who has actually looked at the waveform of a Music Hall Cruise Control on a scope, and found it to be quite smooth--a visibly better waveform than Amir's wall power--but who uses it only to provide adjustable turntable speed control" Denney

There's no logical fallacy - there is only a disagreement about how likely it is that a significant number of audio devices on the market have such poor AC filtering that they can be helped by these AC cleaning devices.

It's the same argument as folks have about USB purifiers/galvanic isolators. Amir has tested those too. and - except for the Schiit Modi 2 - neither he nor anyone else could find a single DAC that showed any measurable difference with vs without a USB purifier in the chain. And if memory serves, even with the Modi 2 I believe only one of the USB purifiers he tested made any improvement in the measurements.

So what are the chances that there's a DAC or other audio device out there that (a) has good noise and distortion performance overall, but (b) has such poor AC filtering that it can measurably benefit from one of these AC cleaners, and (c) is uniquely good-sounding enough and expensive enough that it makes one whit of sense to add an AC cleaner to it instead of just getting rid of it and buying a better DAC? (Because remember, the USB purifier that could improve the Modi 2 cost more than the Modi 2 itself - for the price of the purifier you could just buy a better DAC.)

I'd say the chances are slim to none, and the burden is on you to provide some evidence otherwise.

 

[Francis Vaughan]

If so, it should be trivial to get the problem fixed at the source.

How? What are the technical and regulatory steps that can fix this?

 

[rdenney]

I do have a 500-watt isolation transformer that I bought used for about a C-note more for fun than anything. It’s a medical device, made by Tripplite and sold to hospitals to isolate medical equipment from mains noise. No audible effect with any of my stereo stuff, of course, but it gives me more outlets and looks cooler than a power strip.

I have lost a leg of supplied power at the transformer that left half the house with a brownout and the other half with about 170 volts. That cost some money. Surge protectors didn’t help (or died in the attempt, including the expensive ones in the panels), and those aren’t exactly free, either. Fortunately, the audio stuff was all turned off. This isolation transformer would have taken the hit, I think, and it has tight (130-135 VAC RMS as I recall) over-voltage protection.

I have specified power conditioners for computer equipment rooms, and even inertial regenerators for critical UPS transfer switching.

I use UPS units for most of my systems to provide surge protection and out-of-range voltage protection as much as anything. I have about six of them.

But as far as filtering noise goes, most audio equipment already has it in abundance. Buying audio equipment with well-filtered power supplies is, of course, a strategy. But there are other reasons for managing power quality, too.

Rick “who’d like to see Amir’s test with a phono cartridge and preamp” Denney

 

[Francis Vaughan]

There is a clear logical fallacy of false syllogism occurring here.

The tests so far have been:

There is a claim that poor audio equipment can benefit,
Here is a poor bit of equipment,
It didn't benefit,
Therefore, there do not exist any examples of equipment that can benefit.

This is junk, and the sort of crap arguments I would expect from the audiophool community.

Suddenly from this being a scientific exercise in measurement it has become one where the burden of proof is now on others to disprove the above fallacy of logic. In a science based forum this is depressing. Science has some serious problems in recent times, the reproducibility crisis being uppermost. Lets try not to go down the same path, and hold ourselves to a higher standard.

 

[John B]

Filtering I don’t get but regenerating a nice stable voltage within your device’s expected voltage range must have a benefit for some folks. I’d rather expose the regenerator to my voltage swings than the power supplies on my out of production/warranty gear. Maybe a bit outside discussion scope here through.

 

[rdenney]

There's no logical fallacy - there is only a disagreement about how likely it is that a significant number of audio devices on the market have such poor AC filtering that they can be helped by these AC cleaning devices.

It's the same argument as folks have about USB purifiers/galvanic isolators. Amir has tested those too - except for the Schiit Modi 2, neither he nor anyone else could find a single DAC that showed any measurable difference with vs without a USB purifier in the chain. And if memory serves, even there I believe only one of the USB purifiers he tested made any improvement in the measurements.

So what are the chances that there's a DAC or other audio device out there that (a) has good noise and distortion performance overall, but (b) has such poor AC filtering that it can measurably benefit from one of these AC cleaners, and (c) is uniquely good-sounding enough and expensive enough that it makes one whit of sense to add an AC cleaner to it instead of just getting rid of it and buying a better DAC? (Because remember, the USB purifier that could improve the Modi 2 cost more than the Modi 2 itself - for the price of the purifier you could just buy a better DAC.)

I'd say the chances are slim to none, and the burden is on you to provide some evidence otherwise.

My turntable is a Thorens. Not the best ever made by any means but a quality product. The cartridge is an Audio Technica AT-440mla—not a budget model at all among moving-magnet cartridges.

I get a 60-cycle hum from that source. Is it the cartridge picking up the field of the AC synchronous motor? The cartridge wiring doing the same? The wall wart spraying EMF all over the place?

That stuff is not junk, and performs well overall (by phono standards). I can notch out the hum with EQ, of course. Would it or the phono preamp (which has a mountain of gain) hear AC line noise? Something is. I don’t think it’s dirty power, though. But it could be—dirt produces a lot of spurs. Unlike with the orthogonal relationship between the electrical signaling and the data it carries on a USB cable, analog power supplies or high-gain devices are plausibly subject to noise caused by power line spurs.

A regenerator didn’t help, suggesting to me that the phono pre-amp is hearing line noise from a power supply, not the cartridge hearing it from the mains. So it didn’t solve that problem. (But it certainly did have other benefits). Amir already suggested he could extend the test to a phono preamp. I’m encouraging it.

But in regards to the burden of proof, I’m not the one who published an article, nor the one who implied in that article—challenged by experts—that amplifier distortion is an indicator of power supply filtering quality. So, with all due respect, I feel somewhat unburdened.

Rick “believing authors carry the burden of evidencing their conclusions” Denney

 

[Bartl007]

Studio engineers and professionals are not new to power and the issues that can arise in complex systems. I would not consider a DAC plugged into a "power conditioner" a complex system. A few good articles here if you have the time:
https://drive.google.com/file/d/1sgAReu7CtYiSKXhiLjaF0UewMR4WTEzL/view?usp=drivesdk

https://drive.google.com/file/d/1aEof-jlw7B3ULwF54QF4NtjWm9a3oz2e/view?usp=drivesdk

 

[amirm]

Is it only a matter of time until an audio company starts marketing an off-the-grid system including "audiophile grade" solar panels, battery storage and inverter?

Not the solar part but the rest is already done.

This is the stromtank. They charge a ton of money for them but they have pretty low capacity. And internally it has an inverter so brings with it whatever noise and distortion it generates.

 

[John B]

I switched to a Hegel V10 and all my noisy cartridges fell silent with same power source and TT setup.

 

[IVX]

Our group did investigate the degree of affect the AC sine purity to the sound and found that it could be noticeable only with low loop gain amps contain nonregulated power supplies. Tube amps, vintage amps, and so on. One day I had my regenerator's synchronizing broken, and the device started to increase distortions(by adding high-frequency artifacts) instead of reduction, I was able to hear that very clear. We didn't try to compare any DACs with clean/dirty AC sine, I didn't see a technical prospect to find the difference(all DACs I know have well regulated PS) and had no spare time. The only way to test AC regenerator is real load i.e. old-school power amp with a tank of caps after transformer+diodes. The Power Factor of such things is .7-.8 + huge inrush current. PS-audio regenerator(at least in 2005) being loaded rated power and PF=.7 will show you a comparable THD+N level as your outlet has. The problem is too high output impedance of such regenerator, so, to power 100Wx2 class AB amp you better go with 1000W or so rated PS-audio regenerator. My regenerator wasn't DIY-like design with class B amp + step-up trafo and had 0.004ohm of the output impedance that's why the load very slightly affected the THD+N. It was class D amp, UcD modulator(I did ask B. Putzeys is that would be legal, he said - no problem) + extra loop filter to provide 80db of the loop gain at 50Hz, and stable work at crazy10uF of capacitive load(it was P. Mcgowan requirement because for the beginning we had pre-agreement to design that for PS-audio but finally the design was sold to China, that's why I'm here). The feedback was taken from the outlet connector directly, class D residual switching was reduced up to millivolts by 4ord output LC filters. The topology was a compensation type in other words, the device placed in serial with AC outlet and produces anti-harmonics to its output to compensate initial distortions of the AC line. This way internal class D amp needs only the power equal to THD of AC line: 2000VA load with the input AC line THD = 10% requires just 200W of the compensation. No surprise that the efficiency of such regenerator reaches 98%, and 2000W device was implemented in the case like today you can find "AC filter" i.e. like a power-cord extension for 5 outlets, a bit beefy one.

 

[JohnYang1997]

Studio engineers and professionals are not new to power and the issues that can arise in complex systems. I would not consider a DAC plugged into a "power conditioner" a complex system. A few good articles here if you have the time:
https://drive.google.com/file/d/1sgAReu7CtYiSKXhiLjaF0UewMR4WTEzL/view?usp=drivesdk

https://drive.google.com/file/d/1aEof-jlw7B3ULwF54QF4NtjWm9a3oz2e/view?usp=drivesdk

One core issue for recording studios is that there are often too many boutique devices that are single ended and earthed. Getting that sorted out is absolute pita. Another is that many noises that are usually at low level in the reproduction are boosted into audible range in the studio. Exciter, compressor, dynamic eq, high gain distortion amplifiers etc all do this. Also instruments like electric guitar are inherently prone to magnetic interference. And they also use single ended connections.
If everything is done in a box and digitally, there would be no issue. Although it may not happen for many studios.

 

[b4nt]

I feel now much better here. Everyone disagrees

 

[Francis Vaughan]

If everything is done in a box and digitally, there would be no issue. Although it may not happen for many studios.

Latency is the elephant in the room. A heck of a lot of music is produced "in the box" to the extent that they dispense with human players for many tracks and autotune the hell out of singers etc. But if you want actual real life humans creating real life music, there are limitations to what you can do. Ensemble playing by virtuoso musicians can't be created in the box.

 

[JohnYang1997]

Latency is the elephant in the room. A heck of a lot of music is produced "in the box" to the extent that they dispense with human players for many tracks and autotune the hell out of singers etc. But if you want actual real life humans creating real life music, there are limitations to what you can do. Ensemble playing by virtuoso musicians can't be created in the box.

Um not really. Latency has not been the issue for a long time.
In the box doesn't mean it's not real performance.

 

[amirm]

@amirm I'm almost afraid to ask but ... would you be willing to run these tests on a guitar tube amp?

I don't have a tube amp.

 

[C0B-1D3-9]

There is a clear logical fallacy of false syllogism occurring here.

The tests so far have been:

There is a claim that poor audio equipment can benefit,
Here is a poor bit of equipment,
It didn't benefit,
Therefore, there do not exist any examples of equipment that can benefit.

This is junk, and the sort of crap arguments I would expect from the audiophool community.

Suddenly from this being a scientific exercise in measurement it has become one where the burden of proof is now on others to disprove the above fallacy of logic. In a science based forum this is depressing. Science has some serious problems in recent times, the reproducibility crisis being uppermost. Lets try not to go down the same path, and hold ourselves to a higher standard.

Ahh, sadly all communities are like that unless they do it for living, such as actual scientists.