How is this possible? Shunyata power line filtering (with video)

[Keith_W]

I visited a fellow audiophile a few days ago, and he is a big believer in filtered power supplies and "clean" electricity. Before you roll your eyes, let me explain that I am totally with you - having read all the PS Audio and Shunyata reviews I am a total skeptic. He had a gizmo that claimed to measure noise in mains power and he showed me how various interventions reduced the noise. I honestly can not explain what he demonstrated to me which is why I ask for your help.

The system was a typical audiophile system - Esoteric CD transport / streamer, Mola-Mola DAC / preamp, and directly into a pair of active ATC speakers. What was interesting was the extent he had gone through to ensure clean power. He has triple phase power installed in his home, with the solar inverter on one phase, all home electronics on another, and the third phase exclusively dedicated to audio. For his system, he has a few sets of heavy snake-like power cables into a Shunyata Venom power distributor and then to his components.

He showed me this gadget which "measures" noise in AC power:

It is an out-of-production ISOTek noise analyzer. It claims to subtract the 240V / 50Hz AC and play whatever is left behind as noise (which usually manifests as crackles) through an internal speaker. The digital display reads 0 - 800 which indicates noise in "parts per 1000" indicating how much noise is in the mains power.

The video below shows the analyzer plugged into the "home appliance" outlet. Noise is at max, and if you listen carefully you can hear radio signal played back through the speaker:

The next video shows the analyzer plugged into the Shunyata Venom power distributor. According to Shunyata's webpage, it is a simple power distributor with no description of how it works beyond "CCI noise reduction" which stands for "Component-to-Component Isolation" which is a "proprietary multi-stage filter that reduces electrical noise" along with some mumbo-jumbo about cryogenic treatment, sonic welding, vibration isolation, and so on. Regardless, it is a finely built product that I would happily insert into my system if not for the fact that it was so expensive and I honestly could not see any benefit beyond its built quality and sexy looks. That is, until my friend did this demonstration for me:

Wow! This was plugged in to his purely audio phase of his home electric supply. Only 12 parts per thousand, and if you listen to the video there is barely any noise or crackle coming out the analyzer. This is a dramatic improvement. So I asked the obvious question - how do you know this is from the Shunyata and not from isolation from random electric noise injected into your mains by the other appliances on your other circuit?

So he did this demonstration for me. In this video, you will see the analyzer plugged into the same mains socket as the Shunyata (note that the mains sockets are wired in parallel). The analyzer is on the left, the Shunyata is plugged in on the right. The video starts with both plugged in, then he pulls out the Shunyata:

I was quite astounded to see that Shunyata's passive box of tricks was actually doing something. You can see a dramatic increase in noise when the Shunyata is unplugged. How is this even possible! Up to now, I thought that Shunyata sold overpriced snake oil to audiophiles. Of course, like you I had more questions:

- Would it be possible to measure and view the output via an oscilloscope? Answer: neither he or I own an oscilloscope.
- What is the difference in output at the DAC? Answer: he claims there is an audible difference, but he has not measured it, and the setup was such that it would be nearly impossible to run enough cable from another mains socket for me to do a DBT. I was regrettably not given a demonstration of the claimed audible difference which was said to be dramatic.
- Is there some way the ISOTek noise analyzer is faking the result? Answer: why should ISOTek design a product that fakes a result for products made by another company?

I am out of answers here. Of course, the BIG BIG question is whether it makes a difference out of the DAC, but I have no way to test this. Neither am I willing to shell out so much money to buy a Shunyata distributor to send to Amir to test. Would the community be able to help by:

1. Explaining how these observations could be possible
2. Perhaps someone owns a distributor which he could send to Amir?

 

[solderdude]

Are you plagued by weird noises in your system that can not be linked to ground loops ?
If not... you don't need any of this.

Every wire is also an antenna so it will pick up all nearby radio stations (micro-volt at best).

Filtering of mains is not voodoo or a black art. Some well placed inductors and capacitors can already bring improvements.
Measured improvements does NOT mean this will result in audible improvements.
These are separate things and all electronic equipment should be immune to crap on mains... simply because mains is NEVER 'clean'

 

[DSJR]

NO *domestic* stereo system needs humongous power leads like that, surely?

Back in Jurassic times when I had bigger ATC actives, I did find that cable ferrites on their mains cables did appear to make the bass 'sound' a bit clearer. if I removed them, I 'felt/imagined (you choose)' the bass slightly less good. these things were dirt cheap, so I used them (my AVI preamp was totally immune to this stuff I remember, but its designer was an rf engineer prior to his audio business so I believe knew a thing or two about 'rf clean' amp circuit designs)

 

[Keith_W]

Are you plagued by weird noises in your system that can not be linked to ground loops ?
If not... you don't need any of this.

Actually, that is a good (but separate) point. How is it that ground loops can produce hum in an amplifier with a well designed power supply? Why should weird noises that aren't linked to ground loops make it to the output of an amplifier? Shouldn't the PSU be able to remove any noise from the mains?

 

[DVDdoug]

Amir has tested some (or at least one) power line filter and it didn't have any affect on the audio.

I am out of answers here. Of course, the BIG BIG question is whether it makes a difference out of the DAC.

Every (active) audio device has a power supply, either internal or external, sometimes USB power from a computer,* or sometimes a battery. The power supply converts the power line AC to DC, usually at a lower voltage and sometimes to multiple voltages and/or negative voltages.

The DC voltage is ALWAYS filtered and often regulated so it should be (nearly) noise free and rock-solid stable. (There is always some noise in everything.) A good power supply will be fairly tolerant of noise on the power line or reasonable variations in power line voltage.

Sometimes noise from something like a light dimmer or an electric drill will "leak through" or sometimes when an air conditioner or something switches-on the voltage may temporarily drop (and you might see your lights flicker) and noise may get-through from that. But the main filter in a power supply is capacitors and capacitors store energy so the DC doesn't drop instantly and a brief loss of AC might not affect the audio.



* USB power is notoriously noisy. It has no effect on the digital audio data but sometimes it gets into the analog circuitry. Usually the DAC (or audio interface, etc.) filters it out but sometimes the filtering is imperfect. If you hear the noise it's usually a high-pitch noise from the computer's switching power supply or from the digital switching inside the computer. The computer's power supply is pretty-well isolated from the AC power line.

 

[solderdude]

Actually, that is a good (but separate) point. How is it that ground loops can produce hum in an amplifier with a well designed power supply?

This has very little to do with the power supply but more with poor grounding and layout/wire routing/groundplane management in devices than power supply quality.

Why should weird noises that aren't linked to ground loops make it to the output of an amplifier?

Because these noises are common mode and can travel through groundplanes, via's, incorrectly done PCB layout and thus become audible.
Break the common mode loop and the noises are gone.

Shouldn't the PSU be able to remove any noise from the mains?

Most do. But only differential mode and not common mode.
Common mode filtering and signal routing is key.
Something not all designers are familiar with unless they know a thing or 2 about EMC/ESD/RF etc.

 

[Palladium]

IIRC Amir fed a DAC with AC power that was incredibly dirty with -13dB harmonics and the audio output still measured every bit as good as much cleaner power.

 

[Keith_W]

This has very little to do with the power supply but more with poor grounding and layout/wire routing/groundplane management in devices than power supply quality.


Because these noises are common mode and can travel through groundplanes, via's, incorrectly done PCB layout and thus become audible.
Break the common mode loop and the noises are gone.


Most do. But only differential mode and not common mode.
Common mode filtering and signal routing is key.
Something not all designers are familiar with unless they know a thing or 2 about EMC/ESD/RF etc.

All right, is there a way to apply commode mode filtering besides removing the ground on my equipment?

 

[solderdude]

One should preferably not remove the safety ground.
A common mode filter can lower RF common mode signals.
For audio bandwidth common mode attenuation galvanic separation is the eaiest way if fixing is needed.
line level transformers for analog or digital isolators (chips, optical or transformer)

 

[GXAlan]

Measured improvements does NOT mean this will result in audible improvements.

This is the key statement. Look at the PS Audio Noise Harvester.

PS Audio Noise Harvester AC Cleaner Review

This is a review and detailed measurements of the PS Audio Noise Harvester and AC Cleaner. It is on kind loan from a member and sells for $99. The harvester is a bit bigger and uglier than typical phone charger: Claim to fame is that AC noise is captured gradually and then discharged all of...

www.audiosciencereview.com

It does what it says it does. But we have yet to find a “system” or scenario where it makes an audible difference.

Shunyata has some cardiac electrophysiology products focused on having cleaner signals on what is essentially an oscilloscope. If I was building a EP lab in a hospital that probably was built on the 1970s, there may in fact be justification for these tweaks. As it stands, the most favorable speculation is that there once were products with poor DC regulation where power conditioning mattered but with anything tested to date, we are not seeing big improvements at the post audio gear electrical measurements stage which is already one step away from audible differences.

 

[steabert]

I recommend you really go read the review of the ps audio powerplant12 on this site. It shows indeed that it makes a difference for distortion peaks close to the mains frequency. But also that none of that matters.

I did roll my eyes (actually more of a painful squint) already before you told us not to. What's going on here is another case of praying on people's OCD. They design a device that can measure and amplify that particular frequency range, so you can have something to worry about. But looking at Amir's review that covers more than that, there's no benefit and there are problems in other areas.

This particular part of the HiFi industry is drunk on anything that can make people worried, so they can sell them a solution. That's why they always stay clear of proper blind tests.

Regarding OCD, it usually is the case that the person needs to find something to worry about almost immediately after a solution was found, with only very brief periods of satisfaction. There is always a thought that there could be something missing. Companies praying on this are unethical.

This is a case where measurements are used to a amplify part of a problem that doesn't affect audio playback. Sad.

 

[Keith_W]

OK, let me sum up what has been mentioned in this thread and what I have learnt:

1. Noise in mains power lines is common. It can be injected via radio transmissions, other household appliances, and Ethernet over Power.
2. This has no effect on performance of audio electronics if there is a decently designed power supply. This has been tested by Amir in several reviews of power conditioning products.
3. The only way noise can affect audio electronic performance is via the ground.
4. Common loop ground filtering via galvanic isolation can be applied if necessary.

 

[solderdude]

1,2,4 yes.

3: Ground and ground loops are not the same thing though. Safety-ground and signal-ground are 2 different things that can sometimes bite each other.
It could also be that a lack of ground (in certain devices) or a more susceptible design in electronics (doesn't need to be the power supply but can be) can make noises audible.
Also not (properly) screened audio cables or mains cables and audio cables that are tied together over a length can induce unwanted signals.
When plagued by 'audio gremlins' first thing to do is proper wire routing and methodological problem finding before resorting to buying 'fixes' that may not be needed.

Rule of thumb. When you don't hear any weird sounds that are not part of the recording all is fine and no extra measures need to be taken nor will they be beneficial.
Filters can even make things worse.
Don't try to 'improve' or 'fix' things that aren't broken.
So... no need to use USB isolators or other 'sound improving' devices when they are no gremlins. 'Sound quality' will not improve.
It is easy to hear improvements though even when nothing has changed in reality. On this premise all 'audiophile' tweaks actually work.
A different mindset usually leads to hearing 'improvements'.

 

[Speedskater]

The power supplies in audio components are great sources of AC power line noise/interference.
Linear powers supplies mostly under 10kHz, while SMPS's (Switch Mode Power Supplies) are in the low radio range.

 

[gfinlays]

Well designed audio equipment has no issue with noise on the mains:

 

[kmidst]

Besides for power line filters not being necessary at all in audio systems, there could be a detrimental side effect. You're adding more resistance to the front end of your system, which drops some voltage that your audio gear would count on for its own power supply. That being said, a good PS will allow a fairly large tolerance in input voltage. But it could be possible for the filter to do more harm than good.