Impact of AC Distortion & Noise on Audio Equipment

[H-713]

Why you say that? You think people buy these boxes to only improve the performance of vintage products?

On what basis do you say other devices will act different?

I never said that people will only buy regenerators to improve the performance of vintage equipment, though that is certainly one potential use case. What I'm saying is that very big conclusions have been drawn from a very, very limited sample size.

 

[solderdude]

Vpeak of the clear sine is higher.

It's the Vpeak of the sine which determines the Vp of the unregulated DC.
In fact the differences may even be smaller than I calculated when one takes in consideration that the flattened sine has more time to replenish the DC (which sags due to the load) so the ripple may even be smaller at max output power of the amp.

Less than 1dB less max output power isn't going to change the sound signature anyway given the fact that not many people will be running their music clipping their amps. That extra dB would not help. You need 6dB more at least when amps reach clipping levels.

 

[IVX]

An (auto)transformer or an online UPS are the only devices that can return the sagging AC voltage line to its original voltage.

I agree with that 100.0% And the difference would be measurable even with a cheap oscilloscope, right?

 

[JohnYang1997]

I never said that people will only buy regenerators to improve the performance of vintage equipment, though that is certainly one potential use case. What I'm saying is that very big conclusions have been drawn from a very, very limited sample size.

I don't think anyone is saying there's absolutely no impact from AC power characteristics. The point is that you can just get almost any device you ever need without such issue and without spending that much money on something you do not need. If you watch the video from Amir for this topic, it's very upfront in the beginning.

 

[pma]

What I'm saying is that very big conclusions have been drawn from a very, very limited sample size.

Worse, the method used in post #1 does not tell much about possible effect of mains irregularities to audio component noise. It is the method improper to the estimated goal of the thread.

 

[amirm]

I never said that people will only buy regenerators to improve the performance of vintage equipment, though that is certainly one potential use case. What I'm saying is that very big conclusions have been drawn from a very, very limited sample size.

The big conclusions are there because three very different devices showed zero difference in the face of massive distortion. And we know why. Because they convert AC to DC and filter the heck out of it. Since just about any audio device does the same thing, then we have every right to predict that the same applies to just about any electronic device out there.

What are you suggesting? That people keep buying these devices until we test a few thousand ones and come up with the same outcome???

 

[JohnYang1997]

It's the Vpeak of the sine which determines the Vp of the unregulated DC.
In fact the differences may even be smaller than I calculated when one takes in consideration that the flattened sine has more time to replenish the DC (which sags due to the load) so the ripple may even be smaller at max output power of the amp.

Less than 1dB less max output power isn't going to change the sound signature anyway given the fact that not many people will be running their music clipping their amps. That extra dB would not help. You need 6dB more at least when amps reach clipping levels.

Yep. An amp with more power will solve all these issues.

 

[solderdude]

the difference would be measurable even with a cheap oscilloscope, right?

Well the waveform would still be crappy when using an (auto)transformer that increases the voltage 7%.
When it has a +10% tap you even end up with a very small increase in max. output power !!
A 220V autoformer with a 230V tap (+4.5%) or 240V tap (+9%) would increase the output to 215V and 225V respectively.

Still... less than 1dB more headroom is not going to do much for the sound.
When one is clipping their amps regularly (to audible levels) it is time to buy an amp with at least 4x higher power rating anyway.

 

[amirm]

Worse, the method used in post #1 does not tell much about possible effect of mains irregularities to audio component noise. It is the method improper to the estimated goal of the thread.

No. Noise is shown in the FFT and measured in THD+N and SINAD. No difference was seen.

I have measured incredibly low amounts of noise without the aid of any power conditioning in wide range of devices. By your notion, that should not be possible.

 

[DualTriode]

Hello All,

Some play with tubes and discrete transistors for the fun of it. The power supplies and voltage regulators can be and often are bigger heavier, more complex and more expensive than the attached tube or simple JFET circuit. With the power supplies and shunt regulators in shielded steel junction boxes resting on the damp (grounded) concrete floor I could do pretty well with getting rid of the power supply harmonics and noise. Even with the shielded junction boxes grounded on the damp concrete floor and a separate ground rod driven in the soil outside the 60Hz came in through the air.

See the sound floor in the attached FFT 12B4A tube HPA FFT from the early days of the APx555 on my bench.

Thanks DT

 

[wwenze]

It's the Vpeak of the sine which determines the Vp of the unregulated DC.
In fact the differences may even be smaller than I calculated when one takes in consideration that the flattened sine has more time to replenish the DC (which sags due to the load) so the ripple may even be smaller at max output power of the amp.

Less than 1dB less max output power isn't going to change the sound signature anyway given the fact that not many people will be running their music clipping their amps. That extra dB would not help. You need 6dB more at least when amps reach clipping levels.

Aye, this sentence here gave me a eureka moment

Since we know how the current waveform of such power supplies looks like, and how it does not conduct for 90% of the time, since the diodes would be reverse-biased

Which would mean that no matter how horrible the rest of the AC looks, it does not matter, we just need the small duration when the voltage is high to be looking good.

 

[solderdude]

Exactly.
And to add to that.. it often is not the mains that is limiting the very high peak currents in this short moment it conducts but the transformer itself.
I wrote a bit about this subject for those interested.

Below an actual measurement, the primary winding is the one doing most of the current limiting in this example.
The DC output voltage is 24.9V and the load current is 3.3A, the drawn power by the 7.5Ω load is thus 82W.
Notice how a 3.3A current draw from the DC rails draws short 24A current peaks in just a few ms every 20mS (or 16ms)

 

[H-713]

The big conclusions are there because three very different devices showed zero difference in the face of massive distortion. And we know why. Because they convert AC to DC and filter the heck out of it. Since just about any audio device does the same thing, then we have every right to predict that the same applies to just about any electronic device out there.

What are you suggesting? That people keep buying these devices until we test a few thousand ones and come up with the same outcome???

Again, I'm going to make the argument that you're using three devices to represent the entire consumer audio industry. None of those devices are ones that I'd be particularly concerned about having issues. You're an engineer. You should know that three examples of not exhibiting an issue does NOT prove that the issue doesn't exist.

In fact, there's a damn good chance all three devices you used have REGULATED power supplies. That is far from being the case with all equipment.

-Lots of tube equipment still uses AC for the filaments, and there is (in many cases) good reason for that.
-Lots of solid-state power amplifiers have unregulated linear power supplies
-Lots of tube equipment, especially older equipment, uses unregulated power supplies due to the cost and general PITA of designing a high voltage regulated supply.
-There are plenty of class A amplifier stages that don't use any feedback and have a lousy PSRR.
-A really bad situation might be a piece of tube equipment that uses an unregulated power supply and uses an AC filament supply.
-Lots of people still use legacy equipment, whether you like it or not.
-Lots of people use boutique equipment that isn't up to ASR standards, whether you like it or not.

Again, I am ABSOLUTELY NOT saying that people should be buying $3000 power conditioners. In fact, I don't think anyone is saying that. What some of us are saying, however, is that the tests performed are not sufficient to back up the conclusions that were drawn. I don't care if you're trying to do a service (stopping people from wasting money), insufficient data to support conclusions is a problem that casts doubt on the credibility of the author.

 

[IVX]

solderdude, maybe you are right, and 1-2db doesn't matter but I know a lot of DAC users hate 16/44.1 format, prefer 24/96 at least, let's say them about 1db doesn't matter too?

 

[solderdude]

This is Amir's conclusion: (my emphasis)

Conclusions
Up until now we have tested a number of devices that reduce AC noise/distortion only to find them do nothing for the output of audio products. In this little research project, we went the other way, producing very dirty AC feed. Yet three devices from very different origins show zero, and I mean zero, dependency on AC quality. You could argue that we got unlucky with three devices not being sensitive but hard to make a case that they simply did not care.

Of course the explanation is clear: all of these devices first convert mains power to DC and then use it. DC by definition means no variation so filtering is used to remove noise and distortion. Sure, some remains but the rest of the circuit also has immunity to power supply vagaries. By the time we look at at the output of the audio product, we are so, so far away removed from AC that its "fidelity" makes no difference.

With both engineering knowledge and objective measurements backing each other, the conclusion that you don't need to worry about quality of your AC mains is exceptionally strong. This is in the context of fidelity of course where vast majority of these audio tweaks are sold.

Perhaps Amir should have written:

the conclusion that you don't need to worry about quality of your AC mains is exceptionally strong with the 3 tested devices.

Personally I would agree that it could not apply to all audio equipment but it, most likely, will apply to the vast majority of audio equipment.
Unfortunately, in most cases, people that experience improvements when doing some power conditioning may well have changed other aspects than just cleaning up the L-N (differential) mains voltage and mistake that for the actual L-N voltage being cleaner where it could also have been different cable routing, grounding, common mode currents having changed...

 

[solderdude]

solderdude, maybe you are right, and 1-2db doesn't matter but I know a lot of DAC users hate 16/44.1 format, prefer 24/96 at least, let's say them about 1db doesn't matter too?

That, however, has nothing to do with mains voltages nor 1-2dB but is another discussion depending on masters, filtering used, placebo or other effects. It certainly has nothing to do with 1-2dB more headroom as headroom is exactly the same in both cases.

 

[Headchef]

It is hand waiving nonsense. AC power is never at a set voltage. Are we to believe that the performance of our amplifier is constantly changing as AC voltage is??? Who would release an amplifier with such large and audible variations?

Yes, your peak power availability may be reduced due to voltage sag. What else is new? That has nothing to do with quality of the power.

Surely if, as you say, peak power availability may be reduced due to voltage sag, then by adding a device to provide a consistent power feed is going to ensure that there are no adverse effects due to fluctuations?

it seems to make logical sense that the removal of power supply fluctuations would be a sensible thing to do, as it would ensure that a system would be devoid of fluctuations, optimal operation no?

 

[JohnYang1997]

This is Amir's conclusion: (my emphasis)

Conclusions
Up until now we have tested a number of devices that reduce AC noise/distortion only to find them do nothing for the output of audio products. In this little research project, we went the other way, producing very dirty AC feed. Yet three devices from very different origins show zero, and I mean zero, dependency on AC quality. You could argue that we got unlucky with three devices not being sensitive but hard to make a case that they simply did not care.

Of course the explanation is clear: all of these devices first convert mains power to DC and then use it. DC by definition means no variation so filtering is used to remove noise and distortion. Sure, some remains but the rest of the circuit also has immunity to power supply vagaries. By the time we look at at the output of the audio product, we are so, so far away removed from AC that its "fidelity" makes no difference.

With both engineering knowledge and objective measurements backing each other, the conclusion that you don't need to worry about quality of your AC mains is exceptionally strong. This is in the context of fidelity of course where vast majority of these audio tweaks are sold.

Perhaps Amir should have written:

the conclusion that you don't need to worry about quality of your AC mains is exceptionally strong with the 3 tested devices.

Personally I would agree that it could not apply to all audio equipment but it, most likely, will apply to the vast majority of audio equipment.
Unfortunately, in most cases, people that experience improvements when doing some power conditioning may well have changed other aspects than just cleaning up the L-N (differential) mains voltage and mistake that for the actual L-N voltage being cleaner where it could also have been different cable routing, grounding, common mode currents having changed...

There's nothing wrong. Quality doesn't matter. Voltage varies across the world. So I don't think voltage is part of quality. Only the peak voltage matters. And only matters with unregulated supplies, eg unregulated linear supply.

 

[JohnYang1997]

Surely if, as you say, peak power availability may be reduced due to voltage sag, then by adding a device to provide a consistent power feed is going to ensure that there are no adverse effects due to fluctuations?

it seems to make logical sense that the removal of power supply fluctuations would be a sensible thing to do, as it would ensure that a system would be devoid of fluctuations, optimal operation no?

no
Proper solution: get an amplifier with more power.

 

[solderdude]

it seems to make logical sense that the removal of power supply fluctuations would be a sensible thing to do, as it would ensure that a system would be devoid of fluctuations, optimal operation no?

Most devices these days have wide range SMPS so they don't care what the actual mains voltage is. Most audio devices (even the ones with transformers) will all have rectifiers followed by regulators that (should) have enough headroom to take any normal power supply voltage changes into consideration.

Only some older gear and power amplifiers as well as some (poorly designed) audiophile equipment could be 'sensitive' .

The tested device does NOT do anything against power supply fluctuations. It only filters RF (common and differential mode) far above the audible range and it 'clips' very short peak transients (MOV's). This can be done equally efficient, or even better, at much less cost !

Only regenerative UPS actually create a new cleaner sine wave (under a constant load which power amps aren't)