• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

PS Audio Noise Harvester AC Cleaner Review

Where'd you get it from?
My father, a veteran electric engineer gave it to me yesterday, just so, to try. He told me it filters out washing machine and other kind of noises pretty effectively, but it's a real simple circuit - as an engineer he doesn't believe in different sounding cables either, if within an acceptable thickness and quality copper.. so he's absolutely on the anti snake-oil side. He saved some (not much) good old NOS stuff from scrapped mainframes and other kind of old computers, measuring devices etc. while he was working (these have already landed at the garbage collector at that time so all such things were sent for destroying/recycling, .. back then about dunno, 10 years ago, 15 ?)..

Anyway, he has just stumbled upon this device in his back garage and thought he gives it to me, knowing I'm making myself familiar with electricity and Arduinos nowadays. (Basically I'm an IT guy). I think I'll use it for my PC's Genius 2.1 hobby hifi first, I always hear a pop when somebody opens the warm water tap (the circulating boiler starts to heat) and then again one when it stops. It's a rental flat (yet, but moving soon into own house) so we can't completely redesign the in-house electricity here, although pretty new (10yrs).

Anyway, just thought I'm bringing it into this thread for fun and to laugh :) but if it turns out it's even useful, woo-hoo. :D It's really looking and smelling oldschool, built like a tank. Otherwise not big, about 10cm on the long side.

Edit: called him right now, it's from an old Russian industrial computer.
 
Last edited:
"They deploy state-of- the-art EMI filtering technology to “short out” (or shunt) erratic surges and spikes of electrical energy (i.e., dirty electricity), "

Isn't that what these do?
pr3i44046a.jpg


I've seen the videos with the tester and wondered if it is tuned to measure a specific frequency that their 'filter' actually does cause to drop
No, theses filters, like virtually any industrial filter on the market, dont' have devices inside to explicitly dissipate noise energy into heat except for some losses in the coils (which are normally designed to have losses, in this application). Lossless filters would rely mostly on reflecting the noise "back to where it came from" rather than truly eliminate it.

It is up to the user to install extra "high-frequency dissipaters", in the end. The most basic solution is parallel R-C series cells ("snubbers"), the problem is though that this is hardly a one-size-fits-all theme. Another is usage of dampened RF-lossy mains cabling with a ferrite compound around each wire, from e.g. Belgian company Eupen. There also have been approaches using turmaline-filled capacitors but we are leaving a lot of common grounds here. Turmaline does have dissipative properties, though. The question is if these devices using turmaline are effective and really measure to improve things.

Technically, I can't imagine of any negative sideffects when trying to dissipate unwanted noise energy as much as possible provided it is implemented to fit the actual situation.
 
My father, a veteran electric engineer gave it to me yesterday, just so, to try. He told me it filters out washing machine and other kind of noises pretty effectively, but it's a real simple circuit - as an engineer he doesn't believe in different sounding cables either, if within an acceptable thickness and quality copper.. so he's absolutely on the anti snake-oil side. He saved some (not much) good old NOS stuff from scrapped mainframes and other kind of old computers, measuring devices etc. while he was working (these have already landed at the garbage collector at that time so all such things were sent for destroying/recycling, .. back then about dunno, 10 years ago, 15 ?)..

Anyway, he has just stumbled upon this device in his back garage and thought he gives it to me, knowing I'm making myself familiar with electricity and Arduinos nowadays. (Basically I'm an IT guy). I think I'll use it for my PC's Genius 2.1 hobby hifi first, I always hear a pop when somebody opens the warm water tap (the circulating boiler starts to heat) and then again one when it stops. It's a rental flat (yet, but moving soon into own house) so we can't completely redesign the in-house electricity here, although pretty new (10yrs).

Anyway, just thought I'm bringing it into this thread for fun and to laugh :) but if it turns out it's even useful, woo-hoo. :D It's really looking and smelling oldschool, built like a tank. Otherwise not big, about 10cm on the long side.

Edit: called him right now, it's from an old Russian industrial computer.
It's definitely cool.:cool:

One of these days I'll research Russian/Soviet audio and post about it. There was serious science put into electronics and acoustics.
 
Here's mine, almost as beautiful as the AC iPurifier. Real 100-years-lasting sputnik design.
View attachment 69379View attachment 69380

Ok, just got it yesterday as a gift to do something with it if I want to.. I'll probably play around a bit with that to see what kind of effect it has on my system at all. Previously it was part of a big 3-phase computer power supply. I assume I can connect the phases on both sides with eachother and use this then on my single phase AC mains (230V/50Hz).

:cool::D

Connect phase (L) to one of the inputs A1 and connect the load to A2.
Connect Neutral to N1 and the load Neutral to N2.
Connect safety ground to the ground connection. ONLY operate this device from a mains socket with safety ground (PE).
 
The Greenwavefilters have an output socket, so they seem to be normal filters like the ones TA240 showed. I am still astonished that these can filter within the audioband to make an audible difference with that EMI meter. Maybe the EMI meter takes ultrasonic noise and shifts it down to audible frequencies...

The video shows it is a parallel filter. They also have versions with an output. I have no idea if these also have common mode filters on board.

The meter is a bit suspect. It shows 250mV without and 40mV with plug. The meter says 97% reduction and becomes quiet. from 250mV to 40mV is just 16dB attenuation though.
Indeed, maybe they do down-convert higher frequencies to audible ones.
With 16dB reduction it is about equally effective as the PS audio device.
 
I wondered how that could have much effect when I first saw it. The AC line has a low source impedance, and this gadget is supposed to be able to knock down relatively small unwanted signal (noise) by just 'shorting it out'?
 
It's definitely cool.:cool:

One of these days I'll research Russian/Soviet audio and post about it. There was serious science put into electronics and acoustics.


We had pretty much "junk" from Russia in my childhood (Central Europe), I mean audio... however when I grew up I figured out it was mainly caused by weak, today absolutely inferior speaker drivers and less because of electronics themselves. I think they made some cool electronics, the main concept was durability and oversizing, everywhere where they could and due to larger parts resulting in more deviation from factory nominal values they strived for forgiving circuits. On the other hand, they didn't make a big deal about distinguishing between military and non-military components, the one-size-fits-all logic ruled manufacturing too, plenty of parts and devices were made with military use purpose (not all of course but plenty), the overproduction was sooo overwhelming that almost everybody could grab a piece of these toys and build with them in the commercial sector too.

This was especially valid for tubes, where replacing an original model with a different one without any kind of adjustments and the thing still worked was common, so tolerances were higher as well. But my very first amp, driven by 2 small 5$ Class D modules from Aliexpress are also driven by 2 mil-grade transformers with tons of primaries and secondaries connected in the proper way (all used practically) to a conventional Graetz bridge, big electrolytics with film cap on top and as such feeding the Class D modules. Works like a charm, no hum and absolutely cool, although on top power reaching factory specs.

A simple example: besides my 4x 6S33S-V (6C33C-B) tubes (building a pushpull just for fun) I bought NOS 6N1P and 6N6P tubes too, for chips, to drive my headphone amp. Also just for fun, not really wanting to overachieve a THX AAA Drop whatever feed-forward based thing of today. ;-)


I asked the seemingly veteran guy where he got these. He said, oh, well, there're still 3-4 military fairs per year in the country where old military stuff, including disarmed weapons, sometimes 1-2 tanks, but tons of smaller things are to be found, like unused NOS military clothing, boots, military kitchen tools, fire distinguishers, audio and RF amplifiers, tyres, bearings, reel tapes, agricultural tools, transformers, aggregators, and whatever else you can imagine.. I was like omg.. He told me last year he witnessed the destroying of half tons of NOS never used RF and audio tubes (!!!) just because of the palladium inside which will be then recovered in some kind of German huge factory for reuse. I was shocked. On one hand people sometimes go and fight for rare NOS tubes on eBay and elsewhere, on the other hand the great Soviet Union has manufactured soooooooo much of these we simply can't imagine, supplies still last ... this is an absolute nonsense and people not knowing their value get their hands on a big pile of such and have them smashed for some one-time money whereas if they would be patient they would definitely earn 100x more on adiophiles and collectors, selling these devices online one by one. But time is often a factor so .. apparently there're still plenty of tubes out there in the dark depth of some warehouses and those who have political (or other kind of) connections to these people get the approval to scrape some hundred kilos of these occassionally. They measure tubes in hundred kilos, not in pieces. Hundred kilos. I was like nah, can't be... go f* you :D (Ok, kidding, I'm a nice guy but you know what I mean).. this is insane.

This is where Russian audio begins. A whole different crazy world of standards, manufacturer goals and resulting devices, quality spreads. Pretty much worth to study it deeper. For me I'm nowhere yet, still learning the basics and have a strong helping hand when needed so at the moment it's not my way but those who dig deeper and devote time & effort and some googling/reading will find treasure for sure.


Cheers. :)
 
Connect phase (L) to one of the inputs A1 and connect the load to A2.
Connect Neutral to N1 and the load Neutral to N2.
Connect safety ground to the ground connection. ONLY operate this device from a mains socket with safety ground (PE).

Thanks for the comment, this is the way I wanted to do first. What about option 2 for 1 phase, like connecting (shorting) A1+B1+C1 and A2+B2+C2 ? Line input on "1" side, on this common stack, line output on "2" side, N1-N2 and ground remains the same like you said. Not sure what kind of effect such a parallel connecting would have with respect to combined coil inductances, combined capacitances and combined resistors inside (and then filtering point/slope).

It's a bit different from the PS type of magic, there you simply put it into any slot of the AC power outlet, this one seems to be a "serial" one at first sight (meant for 1 phase in this context).

Mains socket is safety grounded + the usual 'fi relay'. (Residual-current circuit braker if Google is right) :) This is obligatory everywhere and strictly controlled for private homes too, through the utility at the electric meter/circuit braker.
 
He told me last year he witnessed the destroying of half tons of NOS never used RF and audio tubes (!!!) just because of the palladium inside which will be then recovered in some kind of German huge factory for reuse. I was shocked. On one hand people sometimes go and fight for rare NOS tubes on eBay and elsewhere, on the other hand the great Soviet Union has manufactured soooooooo much of these we simply can't imagine, supplies still last ... this is an absolute nonsense and people not knowing their value get their hands on a big pile of such and have them smashed for some one-time money whereas if they would be patient they would definitely earn 100x more on adiophiles and collectors, selling these devices online one by one. But time is often a factor so .. apparently there're still plenty of tubes out there in the dark depth of some warehouses and those who have political (or other kind of) connections to these people get the approval to scrape some hundred kilos of these occassionally. They measure tubes in hundred kilos, not in pieces. Hundred kilos. I was like nah, can't be... go f* you :D (Ok, kidding, I'm a nice guy but you know what I mean).. this is insane.

If some tubes were used only for Soviet military uses, or used only in audio gear in Russia, it may have little to no value on the market. People want the classic power and pre-amp tubes that were used everywhere in audio, or in particular famous units. If you can't drop a tube into a Marshall, Vox, or Fender amp from the 50's or 60's, guitar players will not be buying them as an example. All designs since those amps pretty much use those tubes. I know less about the audio tube amp market, but I'm sure it is similar. The tubes that were crushed for palladium were probably what most would consider oddball tubes that few audiophiles would have a use for.
 
Thanks for the comment, this is the way I wanted to do first. What about option 2 for 1 phase, like connecting (shorting) A1+B1+C1 and A2+B2+C2 ? Line input on "1" side, on this common stack, line output on "2" side, N1-N2 and ground remains the same like you said. Not sure what kind of effect such a parallel connecting would have with respect to combined coil inductances, combined capacitances and combined resistors inside (and then filtering point/slope).
Parallel inductors work like parallel resistors, so you'd get 1/3 of the inductance that way.
 
I count 22 sockets in my 1-bedroom flat.

I have a total of 45 (38 in use) outlets just in my lab. Total consumption with everything on (all the test gear, computers and desoldering/hot air rework and soldering) is less than 2400W. Our power points are 10A continuous with a 16A breaker.

The voltage drop is 6v on 232V with a 1500W heater on that line. Recently, the nominal voltage was dropped in our area (they spent a day up poles adjusting T/XF taps) from 240V down to the now standard 230V. Not impressed.
 
Thanks for the comment, this is the way I wanted to do first. What about option 2 for 1 phase, like connecting (shorting) A1+B1+C1 and A2+B2+C2 ? Line input on "1" side, on this common stack, line output on "2" side, N1-N2 and ground remains the same like you said. Not sure what kind of effect such a parallel connecting would have with respect to combined coil inductances, combined capacitances and combined resistors inside (and then filtering point/slope).

It's a bit different from the PS type of magic, there you simply put it into any slot of the AC power outlet, this one seems to be a "serial" one at first sight (meant for 1 phase in this context).

Mains socket is safety grounded + the usual 'fi relay'. (Residual-current circuit braker if Google is right) :) This is obligatory everywhere and strictly controlled for private homes too, through the utility at the electric meter/circuit braker.

As @mansr already mentioned. The inductances would be in parallel so the inductance would be 3x smaller. It would become less effective for common mode as well because for the common mode filter to work the current in the n-coil needs to be the same as in the L coil.
For lower frequencies this will be the case as current divides over the l-inductors but the higher the frequency the more imbalance there will be because the inductance differs between L and N.
Paralleling the capacitance increases them. This will draw more current and shift the effective frequency downwards again which compensates for the dropping inductance. So in effect it will still work but won't make the filter more effective or be able to handle more power.

It is very different from the PS thing. This is a regular common mode filter for 3-phase power supplies and does very, very little when used in a similar way (plugging it in parallel to another outlet). This one has to power devices by them being connected through it instead of parallel.
Furthermore this works from a higher frequency upwards and filters common mode garbage. The PS audio works at lower frequencies, does nothing for common mode garbage and takes up a wall socket.
1 PS audio device plugged in in the kitchen won't be effective in the living room so you need to put one close to the outlet that powers the equipment.

Then also all devices are used for audio have rectifiers and reservoir/smoothing caps so peaks and crap are removed very effectively so it seems pointless. Of course HF peak currents through mains cords might be lower which could be beneficial when you have unscreened or poorly screened non-balanced interlinks running in pararallel very close to the mains wires.
You should not do this anyway.
 
Thank you for this review, Amir. Clean power is one of the few audiophile concerns that "makes sense" in my mind. I really hope someone will be kind enough to send in one of the larger re-generators so we can see just how sensible it is in reality.

Also, is there any possibility for you to test for noise with a bunch of noisy stuff on the same electrical course, like phone chargers, LED dimming stuff, refrigerator etc? Myths are meant to be tested :D
 
The video shows it is a parallel filter. They also have versions with an output.

You are right, they indeed don't rely on current running through them.
 
As @mansr already mentioned. The inductances would be in parallel so the inductance would be 3x smaller. It would become less effective for common mode as well because for the common mode filter to work the current in the n-coil needs to be the same as in the L coil.
For lower frequencies this will be the case as current divides over the l-inductors but the higher the frequency the more imbalance there will be because the inductance differs between L and N.
Paralleling the capacitance increases them. This will draw more current and shift the effective frequency downwards again which compensates for the dropping inductance. So in effect it will still work but won't make the filter more effective or be able to handle more power.

It is very different from the PS thing. This is a regular common mode filter for 3-phase power supplies and does very, very little when used in a similar way (plugging it in parallel to another outlet). This one has to power devices by them being connected through it instead of parallel.
Furthermore this works from a higher frequency upwards and filters common mode garbage. The PS audio works at lower frequencies, does nothing for common mode garbage and takes up a wall socket.
1 PS audio device plugged in in the kitchen won't be effective in the living room so you need to put one close to the outlet that powers the equipment.

Then also all devices are used for audio have rectifiers and reservoir/smoothing caps so peaks and crap are removed very effectively so it seems pointless. Of course HF peak currents through mains cords might be lower which could be beneficial when you have unscreened or poorly screened non-balanced interlinks running in pararallel very close to the mains wires.
You should not do this anyway.

Great explanation, thank you. :) I appreciate.

Back to the PS thing, why do manufacturers (and users) care about the quality of AC at all, at the very last mile, if we have as part of a linear PSU the transformer itself, then all the basic AC filtering circuits (mentioned before by someone), then after rectification the big buffer caps.., making nearly all of our devices baby-proof ? Especially when I think about modern, 'audio-grade', superb quality, well designed switching supplies (which don't induce that much noise nowadays than simpler designs in the old times).. e.g. looking at an ICE, Hypex module, would the quality of the sound be affected by an extra filtering device like the PS Harvester at all ? I don't think so. I would scratch my head more thinking about smps vs linear in terms of peak reserve (current) and power factor, but I'd never come to the idea of a simple extra filter which doesn't make the top of the sine wave text-book perfect anyway. (Not sure if I'd need textbook-perfect sine wave in front of an SMPS but that's a different topic, the waveform provided by the utility itself is just okay I think). :rolleyes:

Anyway I like the channel of PSA and like 'Uncle Paul' himself, a great veteran with a great life full of experiences but I'm not sure if I would really benefit from such a device. Looking at these measurements of Amir I'd say as long we're talking about pure facts, each of us would fail the ABX blind tests. But when we talk about psychoacustics and the placebo factor, it might contribute to my listening experience just by knowing I've done something good for my gear, because my brain is part of the equation at this point and is cheating me. However, talking on this level, this little device does nothing more - or even less - than a glass of beer or just listening right after a good sex, with relaxed brain and maybe a little smile on my face.
 
You are right, they indeed don't rely on current running through them.

It doesn't make the meter any less questionable (as you mentioned) when they don't play open with what they measure and what one hears and above all why 15dB attenuation results in the difference between lots of noise coming out ofthe speaker and total silence of the speaker.
Looks like the speaker might be muted below a set noise level.
 
Thank you for this review, Amir. Clean power is one of the few audiophile concerns that "makes sense" in my mind. I really hope someone will be kind enough to send in one of the larger re-generators so we can see just how sensible it is in reality.
I actually have one of the old PS Audio regenerators. I measured it once and it is as clean as my Lab AC generator. So it does what it says but again, won't make a difference in the analog output of the audio product.

Also, is there any possibility for you to test for noise with a bunch of noisy stuff on the same electrical course, like phone chargers, LED dimming stuff, refrigerator etc? Myths are meant to be tested :D
The circuit I used *is* very noisy. It has a computer with its switching power supply, my Audio Precision analyzer, the device being tested, etc. As I reported the PS Audio harvester was blinking away. And the spectrum shows good bit of unwanted stuff.

Now, I did not plug in a drill or anything as I doubt audiophiles do that. :) I can try next time.
 
Back
Top Bottom