PS Audio Noise Harvester

[blokey]

Peer review, with people he thinks don't know anything because they read books, and really know things, no way.

Paul

I don't know, his posts exhibit quite a few symptoms of Groupthink. Leads me to think there might be a fair few others with the same, uh, inclinations.

Of course, we know what they're going to say already.

 

[Paulhoff]

I think you're getting old, Paul. With advanced science, you can peer-review your own work.

Yep, all you have to do is say it and it is true, yea works for me. Where and when does the book burning start, and where do I get those arm bands.

Paul



Paul goose steps out of the room.

 

[Paulhoff]

1. Witches Burn. This one is fair enough - though the villagers suggest trying to actually burn her as way of testing this.
2. Wood Burns. Hence witches are made of wood. How do you check that she is made of wood? Try building a bridge out of her, one suggests - but Bedevere points out that you can also make bridges from stone.
3. Wood Floats. Bedevere gently leads them to this point, and asks them if they know anything else that floats.
4. Ducks Float. They actually have a lot of trouble thinking of something else that floats - and it is Arthur, who has just arrived on the scene, who says: 'A Duck!' (stunned amazement and dramatic music.)
5. Therefore... The logic goes: that if she weighs the same as a duck, she's a witch and they can burn her. So they put her on a set of scales with a duck, and of course she does weigh the same ('it's a fair cop').

Paul

 

[ExtremeSkeptic]

I have to say, I'm a little curious about these double-blind tests. You would have required at least two assistants: one to set up the environment by either installing the Harvesters or not for each trial, and another to be the experimenter, who must not be aware of the set up at any time.

In order to score 99.99% on any test, you would have to correctly identify the correct option 9,999 times out of 10,000.

This seems very rigorous. Perhaps you should document it and submit it for peer review.

It was multiple choice, anything else isn't worth doing because it isn't a challenge.
http://www.head-fi.org/forums/showpost.php?p=2209663&postcount=15


With my 8 power cord blind test I left the room and went into another room wearing earplugs and covered my eyes, the assistant changed the cable and went to a different room afterwards, after a certain time I came back to my room and tried to hear what cable was plugged in. No contact was made with the assistant. Two assistants aren't needed.
http://www.youtube.com/watch?v=S_bMgSYpZac

 

[MRC_Hans]

You think everything goes through and nothing gets affected in the path?

Where did I say that? However, in a linear setup (and cables are very linear), different signals don't interact (music reproduction would be impossible if they did).

Magically the noise from the AC wiring enters the speakers because they are connected?

What are you talking about?

Electrical signals from the cables entering the speakers don't output electrical signals from the diaphragms.

They don't?

Energy is never destroyed, it is just converted. Just because it is noise at first doesn't mean it always stays that way.

Of course not. Your point?

With your way of thinking if you use the breaks on your car the asphalt doesn't heat up.

You obviously don't understand what I say.

AC noise entering the power supply gets converted. The noise doesn't magically disappear.

Correct.

The conversion process of the noise reduces the performance of the power supply which leads to worse sound.

Explain how converting noise (for which the power supply is constructed) reduces the performance of the power supply.

It's the same with cables, the longer the wiring is the weaker the signal gets. Because energy is converted to heat from the dielectric.

No that is not the mechanism, in the audio range. You need to be in the HF range before dielectric loss plays a role.

In the audio range (and up to about 10MHz) conductivity of the wires is all that matters.

Hans

 

[blokey]

It was multiple choice, anything else isn't worth doing because it isn't a challenge.

With 10,000 questions?

With my 8 power cord blind test I left the room and went into another room wearing earplugs and covered my eyes, the assistant changed the cable and went to a different room afterwards, after a certain time I came back to my room and tried to hear what cable was plugged in. No contact was made with the assistant. Two assistants aren't needed.

Unfortunately your results will have been skewed by your expectations. You knew there would be one of 8 power cables and were purposefully listening for a difference. Not very objective. A better method would have been for your assistant to swap power cables at random without your knowledge and later quiz you on any differences heard.

 

[Paulhoff]

In double blind tests done with people who think they can heard the different between wire (and other woo-woos) and finding out that they only score as well as guessing, will then make excuses like, “the double blind testing method is flawed and that is why they failed”, “there is too much pressure on them perform”, “their ears got tired with all the testing” etc. (they sound so much like dowers when they have failed their testing). But before the test they all say it is easy for them to tell the difference, and the test will be a breeze, they know they well pass with flying colors because they have tested themselves a thousand times already. After the failure it is very rare for one of them to admit they were only fooling themselves without knowing it all along.

Paul

 

[EternalSceptic]

In double blind tests done with people who think they can heard the different between wire (and other woo-woos) and finding out that they only score as well as guessing, will then make excuses like, “the double blind testing method is flawed and that is why they failed”, “there is too much pressure on them perform”, “their ears got tired with all the testing” etc. (they sound so much like dowers when they have failed their testing). But before the test they all say it is easy for them to tell the difference, and the test will be a breeze, they know they well pass with flying colors because they have tested themselves a thousand times already. After the failure it is very rare for one of them to admit they were only fooling themselves without knowing it all along.

Paul

Why should they? They have not the slightest idea of logic and scientific systematic explorations and they are therefore unable to understand what's going on when they are tested scientifically vs their own "scientific" tests.

Simply totally uninformed and unable to understand information provided to them. Sad.

 

[Paulhoff]

Posted by me on another thread, but it works here too:

The way I see it is that schools and parents too, do not teach one thing that is more important than probably anything thing else that needs to be taught, and that is how to break down ideas (and or) concepts and test them to see if they hold up to critical thinking.

There are a few people on this forum and on other forums that I go to that have shown that they do not know how to do this. And when given other points of view and new information, they refuse them outright by sighting themselves as an example because only they see this new idea (and or) thought. They we give examples that don’t hold up and will defend with more poor examples (and or) just ramble on.

But mainly they will refute science without knowing how science truly works. They have no understanding that science is always in flux and they see that as a weakness, but what they don’t understand is that, that is what makes it stronger. They state that science doesn’t know everything, well science has never said that, and they greatly underestimate what it does know.

Paul

 

[Fnord]

This is what I understand so far...

​

Here we have a combined schematic / block diagram of the RS device. It's based upon the photos posted in this thread and the description given by its designer. The components shown are the ones that I'm certain of, and the blocks ... well ... I don't know what's in them, but I can guess their functions.

C1 is the main filter element. It is a capacitor placed directly across the power lines. Its value is either 0.47 micro-farads or 0.047 micro-farads.

C2, C3, and C4 are coupling capacitors. Their values seem to be the same as C1. Their function seems to be to resonate with T1, thus providing a "tuned" input for the bridge rectifier RECT1.

LED1 is an off-the-shelf blue Light Emitting Diode, or LED.

R1 is what's called a "bleeder" resistor, in that its sole purpose is to bleed off the charge on C1 whenever the device is unplugged. Its value seems to be 10 mega-ohms.

RECT1 is a bridge rectifier. I don't know if it's a four-diode type used with a two-pole secondary, or a two-diode type used with a split secondary.

T1 is a step-down transformer. I'm assuming a voltage ratio of 10 to 1, thus making the output 12 volts with an input of 120 volts.

The Voltage Controlled Oscillator (VCO) could be designed with a four-layer thyristor, or with a unijunction transistor as the active element. This should be followed by a bipolar or field-effect transistor to drive the LED.

In operation, C1 provides most of the power-line filtering. C2 through C4 and T1 select noise from a band of frequencies and use the energy therein to drive the VCO, which in turn drives the LED.

[OPINION]

In conclusion, I would say that the device works as advertised, although not exactly by the means stated by the manufacturer. C1 does the actual work, while the rest of the components are there to run a pretty flashing light (i.e., a "Gimmick").

For about $20.00 worth of parts, this device will provide some actual noise reduction for the consumer, and a 400% profit for the manufacturer.

[/OPINION]

-Fnord of Dyscordia-

(PS: If anyone can provide clearer pictures of the device, I can provide a better description of how it operates, and a more relevant opinion, as well. -F-)

 

[marting]

Fnord,

I believe you are incorrect about T1. It is a step-up, turning 1 vac into 10 vac but at freq above 10khz. It, together with the series caps, represents no significant load at 120 vac / 60 hz.

 

[Fnord]

Fnord,

I believe you are incorrect about T1. It is a step-up, turning 1 vac into 10 vac but at freq above 10khz. It, together with the series caps, represents no significant load at 120 vac / 60 hz.

Would you please provide:

Size of the core and its material.
Gauge of the primary winding and number of turns.
Gauge of the secondary winding, number of turns, and taps, if any.

You may be on to something.

-Fnord-

 

[marting]

Would you please provide:

Size of the core and its material.
Gauge of the primary winding and number of turns.
Gauge of the secondary winding, number of turns, and taps, if any.

You may be on to something.

-Fnord-

I have the same info you have. It's fairly obvious that the toroid is operating in step up mode, given the winding shown in the photo and realistic limits on the permeability.

By rectifying the output into a 4v load (or so) the reflected impedance in bridge conduction would be quite low at freqs >100khz. After the amplified "noise" has dropped below the diode bridge clamping level (determined by the cap/zener or whatever), the toroid is operating no load so you just have the winding inductance in a tank with the series caps. This would provide a slightly dampened response that appears to be about 8-10khz and is consistent with their waveforms. It looks like the scope picture was from the secondary after the high freq 10 v/v gain of the toroid.

This also lets them make the claim, legitimately though highly hyped, that the LED is dissipating harvested "noise."

 

[jimlintott]

But isn't this device no different than just plugging C1 into an outlet? R1 keeps me from getting a nice zap if I unplug and touch the leads but the rest of the components just drive the LED.

So C1 will eat up some high frequency transients but so will any amplifier's power supply. So the only real gain is that you have blinky LED. Likely no other difference at all?

(Take it easy on me I took electronics in high school about thrity years ago.)

 

[Fnord]

But isn't this device no different than just plugging C1 into an outlet? R1 keeps me from getting a nice zap if I unplug and touch the leads but the rest of the components just drive the LED.

So C1 will eat up some high frequency transients but so will any amplifier's power supply. So the only real gain is that you have blinky LED. Likely no other difference at all?

(Take it easy on me I took electronics in high school about thrity years ago.)

That's pretty much what I think. In my humble opinion, the device is like any other $10.00 line conditioner, but the blinky blue light is what will cost you another $89.99 and applicable sales tax.

Audiophiles are always seeking a newer, better way to enhance their listening pleasure.

Unfortunately, there's a seeker born every minute.

 

[Fnord]

I have the same info you have..."

Bummer.

I'd like to get my hands on one just to put it on my analyzer and see what the resonant bandwidth really is, and also to see what kind of VCO (or whatever) they're using to trigger the LED.

From the pics, it looks like it's all discrete components -- no ICs or SMTs anywhere. But thos pics don't show everything.

 

[Paulhoff]

It reminds me of those digital clocks that where suppose to clear up the sound on stereo systems 10 or more years ago, more woo-woo.

Paul



A good RF coke is all you need on the power line cord just before it enters the equipment.

 

[bjb]

Fnord, thanks for the schematic! I'm suprised that the device is simply in *parallel* with the hot and neutral, which means it doesn't really filter the 120V power! I had expected it to be in *series* with the hot line, but as implemented, it absorbs no more 'high frequency noise' than a lightbulb plugged into the same socket as your amplifier. The cheapest surge supressor from Wal-Mart will do a better job at suppressing spikes than this thing.

I should clear up one misconception in this thread. Line filters that use a choke in series with the power line do not send noise power to ground, as claimed by PS Audio. Instead, high frequency noise current will cause the choke to develop a voltage that opposes the high frequency noise voltage. Since the choke is in series with the power line, the voltage it generates is *subtracted* from the 120V/60Hz line. The noise is not absorbed (destroyed), but reflected from the audio equipment.

It is true that the reflection is never going to be 100% complete so a small amount of noise is going to get through (whatever remains needs to be dealt with in the power supply). But hanging a filter in parallel with the 120V source only affects the biggest spikes that come along. Amplifiers have been using a small cap on the power line since the vacuum tube days (my 60's Fender guitar tube amps have them) but like the choke, the don't absorb 100% of the spike. In the Noise Harvester, the only active element is capacitor C1, and while I'm sure it helps a small amount, it certainly isn't doing $100 worth of good.

 

[Fnord]

Fnord, thanks for the schematic! I'm suprised that the device is simply in *parallel* with the hot and neutral, which means it doesn't really filter the 120V power! I had expected it to be in *series* with the hot line, but as implemented, it absorbs no more 'high frequency noise' than a lightbulb plugged into the same socket as your amplifier.

Actually, an incandescent light bulb presents a resistive (non-reactive) load to the power line.

A parallel capacitor produces the same result (by different means) as a series inductor; they both increasingly attenuate increasingly higher frequencies -- the cap by shunting, and the choke by blocking.

For example, if C1 (from my drawing) were a 0.047 microfarad capacitor, its reactance at 60 Hz would be about 56 kilo-ohms, but its reactance at 6000 hz would be about 560 ohms -- 1% of its former value!

(And at 10 kHz, C1 would have a reactance of 338 ohms.)

Thus, the reactance of C1 decreases with a proportional increase in frequency. This is the basis of its high-frequency filtering action.

 

[ktesibios]

So as far as filtering action goes, it's essentially an across-the-line capacitor, using the cap as the shunt leg and the source impedance of the AC line as the series legs of a simple U-section low-pass filter.

That could be useful in reducing differential-mode noise on the line, but it would be more effective and way cheaper to buy a capacitor rated for class X service (with the appropriate safety agency approvals) and install it where the AC line cord enters the piece of equipment. Typical unit prices for film caps rated for class X are in the 1-2 dollar range.

Of course, when you open up your amp you might well find that it already has AC line interference suppression filtering installed- and then there's the question of how much of a problem differential-mode line noise really is, given the low-pass filtering action of the main filter caps in the equipment's power supply, the transient-suppression caps which are often connected across rectifiers to protect the diodes from spikes, the small caps which are sometimes connected across power transformer secondaries and, in the case of low-level gear using regulated DC rails, the additional filtering effect provided by the voltage regulator circuitry.

Two dollars worth of filtering. I guess that means the other 97 bucks is the value of the blinky-light circuit.