AudioQuest PowerQuest 3 Power Conditioner & Surge Protector Review

[Nango]

I have not found any of them improving the sound out of a DAC.

The Topping and the SMSL they also feature some Premium PSU, why dont they send you their models for testing pls @SMSL-Mandy @topping ???

 

[amirm]

Surely the point is whether the null (or closest to that) is audible?

For which group? Us or the believers in these products? For us, we already know the answer. And the others won't listen to a complicated answer.

 

[amirm]

The Topping and the SMSL they also feature some Premium PSU, why dont they send you their models for testing pls @SMSL-Mandy @topping ???

I have a very premium power supply I have used for testing (SBooster). I also have lab supplies I have tested with. Nothing is gained from testing with more power supplies.

 

[Nango]

I have a very premium power supply I have used for testing (SBooster). I also have lab supplies I have tested with. Nothing is gained from testing with more power supplies.

Yes, my point was they both mfr maybe know why they don't send-in their PSU as they also measure with the Analyzer.

But @amirm as far as I remember you did not test any USB 5V conditioner like the Shanti from Allo or similar until now???

 

[bobbooo]

Because capture drifts in each iteration so simple null test doesn't work. You can compensate for this in software but then (added to randomness of noise) you don't get infinite null. All you can say then is that they are the same to say, -85 dB. This kind of "gray" answer is not going to dissuade the believers. So you have done a lot of work for little value.

In the goal of ranking audio devices though, I'm not sure an exact infinite(simal) null is necessary, as the absolute difference between signal and output isn't really needed, just the null difference for each device to compare to each other in order to rank their transparency. Random noise would just be a systematic error that would average out over all the DUTs and so not affect the ranking order I would think, and could be reduced to a minimum by using a high-precision analyzer, like yours

As for capture drift, it seems @Serge Smirnoff of SoundExpert has improved on his algorithm since his original AES paper, and this is not really an issue now:

The metric has only one measurable parameter, the Difference Level (Df, dB), which measures differences between waveforms of two signals, the input and the output of some device under test (DUT) [1]. A specially-designed time-warping algorithm removes linear pitch and phase shifts from the output signal with any predefined accuracy. The algorithm finds linear pitch and phase transformation in the output signal, such that the latter has maximum correlation with the input signal. This maximum correlation Rmax determines the value of Df:

Df [dB] = 10lg(1-|Rmax|)

The time-warping operation is deterministic and produces repetitive results for any signals. Hence, Df values are objective measurements (for time-accurate DUTs, time warping is not required). Df levels can be computed with arbitrary time windows: 50ms, 400ms, 3000ms, etc.

Maybe you've already seen, but he has a thread on here about all of this:

Df levels are the lowest possible values for any two given waveforms (in digital domain). Thanks to iterative search for global minimum, which is always one and can be found with any required accuracy (currently 1e-4 dB).

He's also made his algorithm open source and freely available here. Could be an interesting project

 

[bobbooo]

Surely the point is whether the null (or closest to that) is audible?

SoundExpert are attempting to quantify the correlation between null difference measurements and listening tests here.

 

[Haskil]

My power was already "dirty" so didn't need to do that.

I once used my power generator to create bad waveforms and tested it on a headphone amp. It made no difference at all.

Thank you once again Amir! Meanwhile, influencers on the net and in magazines persuade audiophiles to invest in expensive and ineffective power conditioners.
The only device that works is a direct current suppressor present in the mains current - which audibly purrs toroidal transformers and high-power R-Core transformers ...
A diode bridge and 2 capacitors on the phase and instantly transformers that vibrate like old cats become silent ... it sells cheap all assembled. Perhaps you could test one or two?
And thank you again for everything you do for audiophiles around the world.

 

[amirm]

As for capture drift, it seems @Serge Smirnoff of SoundExpert has improved on his algorithm since his original AES paper, and this is not really an issue now:

Our own @pkane has written the mother of all null comparison software: https://www.audiosciencereview.com/...test-deltawave-null-comparison-software.6633/

Serge waited too long to update the thing.

 

[scooter]

The first spike is at 60 Hz and that is all we want. Alas, we also get a boatload of other harmonics and noise products continuing on to the end of the 90 kHz bandwidth of this test

Wow, so usually all of us have such a noisy AC line at home.
If it was clean, there could be only multiples of 60 right?

 

[solderdude]

Wow, so usually all of us have such a noisy AC line at home.
If it was clean, there could be only multiples of 60 right?

Almost correct... absolutely no one, connected to normal mains, has a clean sine wave. Nor is this needed.
There would be no multiples of 50 (or 60)Hz when the sine was clean.

One can 'clamp' peaks and/or filter when needed but if a device needs filtering it should already be in there. When it is not the device is either cheap (skimped on needed filter components) or poorly designed.
Not all audio equipment designers are fully aware of proper designing methods.

 

[ernestcarl]

Almost correct... absolutely no one, connected to normal mains, has a clean sine wave. Nor is this needed.
There would be no multiples of 50 (or 60)Hz when the sine was clean.

One can 'clamp' peaks and/or filter when needed but if a device needs filtering it should already be in there. When it is not the device is either cheap (skimped on needed filter components) or poorly designed.
Not all audio equipment designers are fully aware of proper designing methods.

Shared equipment and circuits is also something to think about. Whenever my laser printer runs, my LED lights start to flicker -- they're quite sensitive compared to the older halogens it replaced. And when someone vacuums upstairs (using one specific outlet), my door-bell freaking rings repeatedly! For that reason, I put a switch to manually turn the door bell off when it gets 'triggered'.

 

[Labjr]

I've yet to see an AudioQuest product that's not a gimmick! Must be great marketing. I'd like to see testing of their directional cables with the battery packs.

 

[GXAlan]

Our own @pkane has written the mother of all null comparison software: https://www.audiosciencereview.com/...test-deltawave-null-comparison-software.6633/

Serge waited too long to update the thing.

Interesting!

Shared equipment and circuits is also something to think about. Whenever my laser printer runs, my LED lights start to flicker -- they're quite sensitive compared to the older halogens it replaced. And when someone vacuums upstairs (using one specific outlet), my door-bell freaking rings repeatedly! For that reason, I put a switch to manually turn the door bell off when it gets 'triggered'.

It would be interesting if you put a vacuum cleaner on an isolation transformer or your led lights on an active line conditioner how it would work. Do you have problems when running the microwave? Both the vacuum and laser printer are high current devices.

 

[bobbooo]

Our own @pkane has written the mother of all null comparison software: https://www.audiosciencereview.com/...test-deltawave-null-comparison-software.6633/

Serge waited too long to update the thing.

Yes that looks like great software. I see in that thread I linked to he incorporated Serge's Df metric into DeltaWave after a discussion on how to calibrate their algorithms to match. I believe Serge's is still more accurate as it finds a single global minimum when iterating over time shifts to eliminate drift, but it seems DW's computation is faster, if less accurate.

Serge's software also provides 'similarity space' plots, plotting devices with a similar distortion profile when playing real music closer together in a 3D graph, which is very interesting. For example, it plots the ~$1000(!) iBasso DX220 quite close to the $9 Apple dongle in similarity space:

As well as giving them a very similar Df value of -29.6 and -29.1 dB respectively when playing real music (the similar distortion profile can also be seen from the very similar Df histogram shape):

This is despite the fact the DX220 measures significantly better in the sine tests, which is supported by your measurements of the Apple dongle and the DX200 (assuming the latter is similarly performing to the DX220), demonstrating a SINAD difference of 9 dB, as well as showing they have different THD profiles:

Serge talks about such differences in standard synthetic test measurements yet similarity when playing real music (what he calls the 'm-signal') here:

Comparison of df-measurements with m-signal (histogram) to df-measurements with white noise confirms this - some DUTs showing low (good) df levels with m-signal have awful df measurements with white noise (Apple devices in particular). So, there is no necessity to "cope" with this. In df-metric perceived audio quality is not measured with tech signals. They are for developers/manufacturers of audio equipment as they help to understand what can be improved in circuitry design.

All this misunderstanding originates from the popular false assumption - if Sine signal has lower distortion then all other signals will also have lower distortion. Df measurements clearly show that this is not true. This misconception is the main reason why traditional audio measurements correlate badly to perceived quality estimations.

Obviously order of harmonic distortion does matter in terms of audibility/objectionability (odd>even, higher>lower) as research shows, so individual THD tests are still useful for a finer grained judgement of sound quality, but it seems the Df metric (calculated using either Serge's or the DeltaWave software) would be a better measure of overall transparency and so sound quality than SINAD, seeing as it reveals the actual level and profile of distortion when playing the real music we use the DUTs to listen to, as opposed to just artificial test tones.

 

[KxDx]

But the Microdynamics!

Sa(i)d John Darko https://darko.audio/2019/05/fight-the-power-with-audioquests-powerquest2-3/

I can't even read anything on his site anymore. He is the absolute epitome of all this industry's problems right now.

 

[Veri]

I've yet to see an AudioQuest product that's not a gimmick! Must be great marketing. I'd like to see testing of their directional cables with the battery packs.

Cable battery packs. The horror.

 

[Labjr]

I can't even read anything on his site anymore. He is the absolute epitome of all this industry's problems right now.

I left his site years ago, when he became an MQA fanboy. All those reviewers are the same. They take money from manufacturers of the equipment that they review.

 

[AudioSceptic]

For which group? Us or the believers in these products? For us, we already know the answer. And the others won't listen to a complicated answer.

I meant for the believers. If they can't hear anything in the (near) null, how can they then claim that there is a difference that doesn't show in measurements but is nevertheless audible?

 

[AudioSceptic]

I have a very premium power supply I have used for testing (SBooster). I also have lab supplies I have tested with. Nothing is gained from testing with more power supplies.

I'd really like to see you test one of the Paul Hynes PSes <https://audiophilestyle.com/forums/topic/37373-the-paul-hynes-sr7/?tab=comments#comment-733206> <https://www.paulhynesdesign.com/products>, but doubt that anyone will send you one.

 

[amira]

This is almost certainly too small and too cheap to be a series mode device.

Correct. Audioquest's Garth Powell was the guy at Furman behind their SMP "Series Multi-Stage Protection" surge protection. They were sued and prevented from calling it "Series Mode" protection.

He brought it over to Audioquest. In reality, it's 2 small MOVs shunting to neutral before the surge gets to a small inductor. The over-voltage relay only protects 1 of the 2 MOVs, and the first MOV also lacks a thermal fuse, so you'll never know whether or not it's still functional. Once the 2 small MOVs inevitably take enough hits, the until has to be replaced. How on earth they call it non-sacrificial surge protection is beyond me.

All in all, it's a jack of all trades master of none surge protection. The over-voltage relay is it's best feature by far. It has barely any series mode protection, and only a couple MOVs. You're better off getting a proper MOV-based surge protector with lots more of them and proper thermal fuses. Same goes for series mode protection, you need a much larger air core inductor that won't get saturated as quickly.

SurgeX (This model also features inrush current elimination)