be quiet! DarkPower 750W - BQT P6-PRO 750W |
---|
AC Input | 100V-240V, 12A-6A, 50 Hz |
---|
DC Voltage | +3.3V | +5V | +12V1 | +12V2 | +12V3 | +12V4 | -12V | +5VSB |
---|
Max. Output | 24A | 30A | 20A | 20A | 20A | 20A | 0.8A | 3A |
---|
170 W | 600 W | 20W |
750 W |
Wow! That's some massive power delivery capability. 20 Amps on each of the four rails should be enough for any video card that will come out in the near future. It is definitely enough for G80 and R600.
Tested on: AMD Athlon64 3000+ @ 2000 MHz, ABIT AT8, Radeon X1900 XTX + Radeon X1900 XTX Crossfire, 2x 512 MB DDR400, WD Raptor 36 GB.
All voltages are very stable, especially the 12V line, which goes to almost exactly 12.0V under load. The 3.3V line seems to be fluctuating a bit, but since no vital components need that voltage it does not matter at all. Nowadays all components draw most of their supply voltage from the 12V line.
Above image shows the Ripple Voltage measurement (5 mV per vertical division, 5uS per horizontal division). With 17 mV from top to bottom, the result is only about average. Ripple voltage was measured on the 12V line at idle.
Standard deviation 12V | 50.62 |
---|
Standard deviation 5V | 10.92 |
---|
Standard deviation 3.3V | 13.70 |
---|
Power Factor | 0.96 |
---|
Ripple Voltage 12V | 17 mV |
---|
Power Efficiency | 78% (320W:413W) |
---|
Standard deviation is a statistical term, which tells how far away from the average the measurements are. In other words it's the average of the average.
A large standard deviation indicates that the data points are far from the average and a small standard deviation indicates that they are close within the average.
With an efficiency of 78% the Dark Power Pro claims the title of being the most energy efficient PSU we tested so far.
10% of efficiency come down to a yearly difference of $4.08 in power cost. This is assuming $ 0.10 per kWh, which is an accepted average value for power cost in the United States. We assumed 8 hours of PC usage per day, 5 days a week at 200W system average power draw.