I'll explain some things most people may not know about mains and drawn power. Audio gear is quite different from resistive loads (used in the 'proper test' Paul did
)
I'll try to explain this as simple as possible but for people not knowing about ohms law it may be too technical.
The internal amp can deliver about 35V AC (most likely, possibly somewhat less even) and has peak current limiting of 30A (so 1kW short momentary peaks) and about 350W output average at full load (assuming PF around 1).
For the EU version about 70V and 15A peaks, same Wattage.
The 30A peak current limit is
my assumption based on the specs from PS Audio and my understanding of how the PPP works.
There is never going to be a real answer to that if its correct as PS Audio is not going to give the numbers and extensive and specialist gear would be needed to check this.
The amp inside is not working in a similar fashion as a speaker amp. This has to do with the way power is transferred into an amp.
Most of the time the internal amp does not have to deliver any current (energy) but its output voltage is changing.
Only near the peaks of the mains voltage the diodes in the rectifiers of electronics (as all audio electronics works on DC as in rectified AC) the internal amp needs to deliver current. As can be seen below about 3/4 of the time NO current is drawn and in 1/4 of the time current is drawn and the biggest chunk in about 1/3 of that time. THIS is what causes the flat-tops of mains by the way. All electronic devices in your whole street/village all draw power at the same time which sags the voltage at the tops only. Because of 3 phase mains the 'pain' to mains is spread over 3 different time moments.
An illustration is needed I suppose. A real world scope shot of a DC power rail at the smoothing caps that feeds electronics inside audio gear.
What we see above the current in the secondary (so not in the mains) of a transformer. The rectified output voltage of the DC is
25V average under a load that draws
3.3A so 82W.
What is observed is that in order to draw 3.3A (in this case 82W) the transformer has to deliver
24A in short bursts.
The loaded AC voltage into the rectifier is 19.9V and peak currents drawn are 24A peaks.
Now the output is a transformer so at 230V the winding ratio is 1:11.5 (lets forget about losses in the transformer for a moment) and 1:6 for 120V AC.
This means (for 120V) using the winding ratio of 1:6 means peaks of 4A at 120V is drawn in order to draw 82W on the DC side.
Now that DC feeds an amplifier that has a 6 ohm speaker connected to it. Class AB amps have a max efficiency of about 75% in practice so about 62W into the speaker. For this 62W output power the amp draws 82W from mains, so the current in 120V systems is 4A in very short bursts and not the 'expected' 0.7A (120V and 82W)
So a
5.7x higher peak currents than one would expect based on drawn power at the DC side compared to say an incandescent lightbulb or heater or dummy load.
What we learned here is that audio amplifiers draw short but high current peaks in short bursts and not act as a resistor or speaker.
Lets assume we use a P12 at max. output power and we use similar numbers (so ideal transformer and amplifiers driven near clipping levels) and connect 2 amps (stereo) that draw about 400W each continuous (which is a serious amp and pretty loud SPL). They are class AB amps with 75% efficiency at full power so draw 530W each meaning 1060W from the P12.
A resistive load (such as used in the wonderful demo from PS audio) at a load of 1000W (so equivalent to 2x400W into the speaker) would draw 8A.
No problem for a device that is specified for 1250W continuous.
The thing we learned above is that in order to draw the 1000W from the DC power rails the PP12 does not have to deliver 8A (like in the resistive load they use to show 'real measurements') but would have to deliver the power into the rectifier in short moments.
We saw from real world measurements of the 82W example above that in order to do so we need 5.7x higher currents in short bursts repetitive in 120Hz. 120Hz because a bridge rectifier draws power at both peaks (positive and negative).
Lets go with 5x higher peak currents than expected =
40A peaks. This would have to be delivered (and pass through) the internal amp.
The peak power it can handle is 3.6kW = 30A so while the device can handle 1.2kW continuous in a resistive load it cannot handle 1.2kW continuous in power amp output.
Now for the fortunate part.. Folks running their 400W power amps will never draw this continuously as music isn't continuous. When the amps are driven near clipping points in peaks in music the amp will draw about 50W average with music. Only in short bursts, determined by the peaks in the music which will be mostly bass transients, that 400W power will be drawn. This power comes from the internal reservoir caps and the only thing that needs to happen is the replenishing of the DC rails in the period mains provides the short current peaks.
So even when 800W peaks are drawn in music the needed 40A peaks even when limited to 30A will only cause a slight sag in the DC rails at very short moments. The MW function (as well as the flat-tops of mains) will allow a bit more time for the diodes in the bridge rectifiers to conduct so current can be a bit lower.
This means the PP12 can supply the needed power in audio systems. Maybe not continuous on a test bench but no issues with music. Even when the peak power output of very high power amps sags a tiny bit this will go unnoticed.
Here's another thing though that has to be taken into consideration. Which is the line voltage of mains will sag when amps draw this kind of power too. When the mains voltage is 120V and the connected power amps draw 40A peaks directly from mains that voltage will sag.
The HC outputs will sag more than the non HC outputs. The HC outputs are intended for high power speaker amps.
Something Paul refuses to address.
The 'power amplifier' inside the PowerPlants thus has to work very different from a speaker amp. It has to be purpose designed for this.