PS Audio P12 Review Part 2: Power Testing

[voodooless]

@restorer-john @amirm So are you saying that the 10% drop in power of the amp could be due to a voltage regulation DOWN by the P12 in the face of a high voltage MAINS ?

That is most likely the case. Add some additional voltage sag for the transformer saturating at high power, and your there. Drop to 117 VAC would have done it.

If that is the case, then what's the point of the test ?

The point is that the claim is that the unit will give the amp more power, not less as this test proves.

 

[MattHooper]

One only has to read the responses to Amirm's review over at the PS Audio forum to see why Paul will always have a customer base.
It's an anthropologist's delight: one head plunging in to the sand after the next. To paraphrase the gist: "who cares about the measurements" and "I trust Paul's engineering" and "I can hear the difference so Amirm must be measuring the wrong stuff"...basically every subjective shibboleth you can think of. Basically, no evidence to the contrary will penetrate this belief system. (And in a perfect case of projection, some refer to people here as the "ASR Faithful." Yeesh. )

 

[LEFASR160]

That is most likely the case.

The point is that the claim is that the unit will give the amp more power, not less as this test proves.

@voodooless @restorer-john Suppose instead the voltage was LOWER in the mains and the P12 regulated it up some % using stored charge in the caps?. Would it provide more power instead of less ? It's the same logic applied to what restorer-john said, but in reverse, in which case I ask again, what's the point of this testing if what restorer-john says is correct ? If he's correct, the test is meaningless, unless I didn't understand what restorer-john said, which is a distinct possibility. Let's wait to hear from him again....it's his hypothesis, not mine.

 

[restorer-john]

As a data point, a random power outlet in my house has a voltage differential between loaded and unloaded of 9V (238V to 229V @ 9.234A) The source impedance of that Active-Neutral return is therefore 0.975R. That is quite a distance from the breaker panel on a daisy chained circuit (bedroom/kitchen/laundry) with standard cabling in a 20+ year old house (nothing else pulling current at the time on that circuit).

That makes the active/hot only circa 0.487R.

I see no reason to doubt @amirm device for the initial measurement of 0.26R for his active/hot only.

 

[CtheArgie]

One only has to read the responses to Amirm's review over at the PS Audio forum to see why Paul will always have a customer base.
It's an anthropologist's delight: one head plunging in to the sand after the next. To paraphrase the gist: "who cares about the measurements" and "I trust Paul's engineering" and "I can hear the difference so Amirm must be measuring the wrong stuff"...basically every subjective shibboleth you can think of. Basically, no evidence to the contrary will penetrate this belief system. (And in a perfect case of projection, some refer to people here as the "ASR Faithful." Yeesh. )

It wasn't a very long time ago that I started in the audiophool area. But the moment I started questioning HOW I heard, HOW I listened and WHAT I could listen, then things started to fall apart. I stopped on the mythology and went back to what I do for a living, which is on the scientific method. Once I discovered ASR, even if I can't understand everything posted, it became easy to change. Change is not hard, you just have to want to learn. Now, I have a sunk cost of crap that I am slowly selling. And my audio is much better than ever for much less money than ever. Plus, it is less complicated and smaller. You have to confront your internal monsters. Once you do, the process starts.

 

[audio_tony]

That is most likely the case. Add some additional voltage sag for the transformer saturating at high power, and your there. Drop to 117 VAC would have done it.

The point is that the claim is that the unit will give the amp more power, not less as this test proves.

I suspect the claim of more power is based on 'like for like' line voltage - e.g. the raw AC and AC output of the P12 being identical.

However this will depend on the ability of the P12 to supply enough current in the first place, which as you say, could be diminished by the transformer within the P12.

@restorer-john even taking into account the P12 regulated output - at a steady current draw (assuming sufficient raw AC line current) - the voltage drop across the device should remain constant. Any fluctuation there suggests that either the raw line input current is insufficient, or the regulation is poor. This is easy to test, by driving the amplifier at a constant high power.

 

[solderdude]

The mains impedance Amir measured with his Hot/Active line at 3R is clearly erroneous.

When Amir reviewed this amp it reached 507W peak (assuming from mains)
This time it reached 561W peak (from mains)
From 120V (PP) it reached 518W peak (10W more than directly from mains the first time)
Let's assume 90% efficiency on this amp (for peak power) which would mean it would draw 4.75A.
When the output R would be 3 ohm the generated mains voltage would sag 14V (so 106V on the amp) = -13% while regulation should be 0.5%
Safe to say the meter gave erroneous values (which Amir already questioned).

In any case the 507W and 561W are too far apart (both measured from mains.

This PP is a tough cookie to break... I say let an EMC test lab measure the darn thing.

One thing is sure though... the 'claims' about sonic improvements by the manufacturer are hard to swallow and so is the price.

 

[voodooless]

@voodooless @restorer-john Suppose instead the voltage was LOWER in the mains and the P12 regulated it up some % using stored charge in the caps?.

I doubt it can regulate up. My bet is that the output voltage is always lower than the input.

 

[NoSnakeOil2]

I recently reviewed the PS Audio PowerPlant P12 AC power regenerator. There, I measured its performance with either no load or typical source device. Questions were raised as to its performance with higher power loads. Here is that test.

View attachment 190503

So I unpacked and lugged the 50 pound P12 back to my lab , and paired it with an Outlaw 2220 monoblock. This is a linear amplifier and doesn't have a regulator power supply meaning it can produce more power if allowed.

To give the P12 full benefit of doubt, I used a Pangea AC 14SE MKII Signature Power Cable (2 meter) to feed it AC. It costs US $120.

PS Audio claims or implies that you get more dynamics and power due to P12 having lower impedance. We will test this.

PS Audio P12 Measurements
Let's test the power capability first and foremost by measuring maximum and peak power of the Outlaw 2220 power amplifier using my standard "Raw AC" wall power:
View attachment 190504
These are my standard measurements but in case you are not familiar with them, on the left is more or less continuous power if we allow the amplifier to reach 1% THD+N. I say more or less as the test lasts about a second or two in each try.

On the right the duration of the signal is much lower. As explained, we have a sine wave that stays at low level (and hence power) for 480 cycles but then peaks to full value for just 20 cycles. This standard was created for car audio manufacturers that want to throw out big numbers for watts. As we see, the outlaw does produce a lot more power, reaching more than half a killowatt. This is far more than the base measurement because the power supply capacitors can maintain that duty cycle well enough to generate that much power.

Now let's route the 2220 amp through PS audio:
View attachment 190505

As you see, we lost 7.5% power in both continuous and burst power. As should be the case as the PS Audio P12 is introducing losses in the AC line.

As another test, I have an Ideal AC line tester, the 61-164. This is a $400 instrument for quick and useful powerline measurements. One of its unique features is ability to measure the impedance of each of the three lines in your AC mains. Let's focus on the Hot wire and see how my standard power strip that I use for powering everything ("Raw AC") performs:
View attachment 190506

Impedance is just 0.26 ohm. When we remodeled our home I made sure every circuit is powered using 12 gauge as opposed to 14. And my power strip is a beefy one as well. I am also not hugely far from our breaker panel.

Now let's power the PS Audio P12 using the Pangea power cord and measure the impedance of its high current outlets I used for previous testing:
View attachment 190507

As you see, impedance has jumped to 3 ohm or more than 10 times higher! I should note that I don't know how accurate this measurement is but given the wide differential, we can likely rely on the fact that PS Audio P12 *system* does have higher impedance. I wrote "system" there because P12 introduces yet another AC cord and outlet so naturally those contribute to degrading its impedance. In other words, we are measuring the whole thing, not just some internal impedance.

Conclusions
It is clear from the above tests that using the P12 degrades available power to an amplifier, not increase it. Both continues and dynamic power are limited, as they should. You are inserting another cable and box (power regenerator) which has its own losses. Burst capability in power amps is provided by its internal power supply capacitors that have a very low impedance path to the amp being inside it. Trying to do that with AC is like pulling the end of a wet noodle and expecting the other side to move with it.

The data here is the reason a number of power amp manufacturers recommend to power their units directly from the wall and not through a AC regenerator. You want the least impedance path to the AC outlet and that is a simple power cord.

While with low current devices we could hardly find a harm caused by the PS Audio P12 (other than to your bank account), with power amps we are seeing a distinct degradation of power. In this regard, I think any marketing material without proper objective back up to the contrary is irresponsible here. Please heed the advice from me and many amp designers that you don't want to use these regenerators for power amplifiers.

-----------​

As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

​

I saw that PS Audio has a new "Mark" cable that costs $33,000. So, the above product is obviously extreme value for money. )))))

 

[MattHooper]

It wasn't a very long time ago that I started in the audiophool area. But the moment I started questioning HOW I heard, HOW I listened and WHAT I could listen, then things started to fall apart. I stopped on the mythology and went back to what I do for a living, which is on the scientific method. Once I discovered ASR, even if I can't understand everything posted, it became easy to change. Change is not hard, you just have to want to learn. Now, I have a sunk cost of crap that I am slowly selling. And my audio is much better than ever for much less money than ever. plus, it is less complicated and smaller. You have to confront your internal monsters. Once you, the process starts.

Yes, to ever find out you are wrong you first have to care about whether you are wrong and have some way of "finding out how I'd be wrong."

This isn't something the purely subjective epistemology offers.

Because you can always imagine a difference and that will trump any and all objective evidence to the contrary. And even on the purely subjective paradigm you can't be "shown wrong" by someone else's subjective experience of not hearing a difference. That just means they don't have your magic powers...er...powers of perception (or resolving system...etc).

It's a Dogmatic system of Faith that one doesn't recognize, and the inevitable psychological projection is that anyone who challenges it is being "Dogmatic."

Maddening.

 

[Larry B. Larabee]

Unfortunately I can't perform dual measurements like this. AP is tied up performing the amplifier measurement and can't be used to simultaneously measure the AC. Standard tools I have for voltage measurements (e.g. meters) are too slow to track the activities of AP trying to determine power.

I tried looking at the P12 meters but they are ridiculously slow and as such, useless for this purpose.

At the end of the day, it doesn't matter "why" but the results we achieved.

You don't have a multimeter?

 

[audio_tony]

I doubt it can regulate up. My bet is that the output voltage is always lower than the input.

I tend to agree with this view. Given Amir's comment about the weight of it, it's likely that a standard transformer / rectifier arrangement is used to power the internal amplifier.
Therefore the output voltage 'swing' will be limited by the input line voltage - however - if it's using a step up transformer, then the output 'swing' could be increased, but probably at the expense of available output current (which appears to be what Amir is seeing in this case).

So whilst it can probably regulate the output voltage at low loads, high loads will result in voltage droop.

 

[Edgar]

He is a very cute little snake. If there are no copyright issues, I agree. Amir could/should use it. He even has an airhead look on his face like there is not much between the ears. Kind of the look like "I can hear the magic!" Perfect specimen!

Should we call it Danny?

 

[CtheArgie]

Yes, to ever find out you are wrong you first have to care about whether you are wrong and have some way of "finding out how I'd be wrong."

This isn't something the purely subjective epistemology offers.

Because you can always imagine a difference and that will trump any and all objective evidence to the contrary. And even on the purely subjective paradigm you can't be "shown wrong" by someone else's subjective experience of not hearing a difference. That just means they don't have your magic powers...er...powers of perception (or resolving system...etc).

It's a Dogmatic system of Faith that one doesn't recognize, and the inevitable psychological projection is that anyone who challenges it is being "Dogmatic."

Maddening.

"Imagine the difference". That is the key issue. I don't want to "imagine the difference".

It's like Cookie sending her email the other day asking: Do you trust math or your ears? I was simply stunned. My ears tell me these tiny musicians with their tiny instruments are inside my head playing when I listen to headphones. I know what I hear! I don't need math or any STEM to explain that to me! How did they get there inside my head? (sarcasm, just in case).

 

[Omar Cumming]

What's so appalling is that he's a fine engineer, so he has to know that the stuff he's putting out there is bullshit and his products are pure snake oil.

I believe it has been pointed out on other PS Audio threads here, that he has no technical background whatsoever

Cheers

 

[restorer-john]

@voodooless @restorer-john Suppose instead the voltage was LOWER in the mains and the P12 regulated it up some % using stored charge in the caps?. Would it provide more power instead of less ? It's the same logic applied to what restorer-john said, but in reverse, in which case I ask again, what's the point of this testing if what restorer-john says is correct ? If he's correct, the test is meaningless, unless I didn't understand what restorer-john said, which is a distinct possibility. Let's wait to hear from him again....it's his hypothesis, not mine.

The Outlaw is a traditional amplifier with a big fat toroid and what looks like a conventional design:

With such an amplifier, the output power depends on the input mains voltage. If the mains voltage goes up by 5% from 120V to 126V, the secondary transformer voltages supplying power to the amplifier will also go up by 5%. As power is the square of voltage over resistance, the change in ouput power is large for a small change in rail voltage. Say the rails were 45-0-45V after rectification filtering. That would mean an approx 33V CT transformer. At 126V, that transformer puts out 34.65V per winding and we now have 50.1-0-5.1V rails to play with.

All things being equal, our amplifier will go from 253W to 313W for a 5% change in mains voltage. If it went the other way, 5% lower (42.6V rails), the power output would go from 253W to 227W. The numbers are simplistic, allowing for a fixed rectifier drop and no OPT stage losses but fine for demonstrating the concept.

WIthout knowing the input and output voltages from the PS Audio P12, we cannot attribute losses to ouptut stage impedance, regulation or whatever.

I doubt it can regulate up. My bet is that the output voltage is always lower than the input.

It uses a DDS/FPGA to generate the sine. That sine can be any desired amplitude and therefore the amplifier it drives can produce any voltage (within the rail ranges). As it regulates to within 0.5%, there is probably an A/D line (feedback from the output sockets) that the micro reads to ensure regulation is maintained.

 

[voodooless]

It uses a DDS/FPGA to generate the sine. That sine can be any desired amplitude and therefore the amplifier it drives can produce any voltage (within the rail ranges). As it regulates to within 0.5%, there is probably an A/D line (feedback from the output sockets) that the micro reads to ensure regulation is maintained.

Yes, but the “within rail ranges” is the key part here. What are those? Are those actually higher than the input?

 

[LEFASR160]

As a data point, a random power outlet in my house has a voltage differential between loaded and unloaded of 9V (238V to 229V @ 9.234A) The source impedance of that Active-Neutral return is therefore 0.975R. That is quite a distance from the breaker panel on a daisy chained circuit (bedroom/kitchen/laundry) with standard cabling in a 20+ year old house (nothing else pulling current at the time on that circuit).

That makes the active/hot only circa 0.487R.

I see no reason to doubt @amirm device for the initial measurement of 0.26R for his active/hot only.

@restorer-john yes but I doubt that the output impedance of the P12 as measured with that device is anywhere near 3 ohms. That's over 300 times spec.

 

[restorer-john]

Yes, but the “within rail ranges” is the key part here. What are those? Are those actually higher than the input?

It's a high voltage amplifier capable of swings over 200V. It would need 170V to do 120V RMS with no losses.

The easiest way to do that would be with a bridge configuration to keep voltages low and costs reasonable. The picture I found suggests that too.

Bear in mind, this is guesswork, but I'd wager a bridged class B pair with tons of feedback to lower the output impedance and a narrow bandwidth.

 

[restorer-john]

@restorer-john yes but I doubt that the output impedance of the P12 as measured with that device is anywhere near 3 ohms. That's over 300 times spec.

Exactly, that's why Amir said what he did. It's clearly not a true number.