Load testing the PSU
Testing a power supply is definitely a challenge, you’ll need professional load testers to check the PSU’s behavior the right way.
The first basic test is performed with a simple power supply tester. It doesn’t load the PSU more than a couple of Watts, but it helps to determine if the power supply unit is operational at all. There’s a self-check indicating if the voltages are at the proper level. As you can see, there have been no issues with the Deepcool DQ750-M unit.
The following setup was used for the remainder of the testing:
- Voltcraft VC-870 Digital Multimeter - voltage measurement (+ Fluke 97 scope meter)
- Oscilloscope - GW INSTEK GDS 3154
- DC load - original solution
- Fluke 97 Scope meter (a measurement of voltage and ripple),
- Voltcraft SL-451 decibel meter (volume measurement)
- ACUVIM-IIRF – for measuring active power (input from the socket).
- PCE Instruments PCE-DT 50 tachometer – for the measurement of the rotational speed of the fan.
Measurements are taken only using the 230 V input voltage.
Efficiency
In our power supply reviews, we emulate real-world usage by creating such loads as a power-hungry, multi-GPU setup with a powerful CPU would have. Older components used to require more power, but there has been some progress in this area. In a typical setup, reaching over 500-600 Watts is a rather difficult task. Let’s focus on the PSU’s efficiency first. Checking efficiency is actually a relatively easy thing to accomplish.
It does look good even at low load (75 W) which is quite typical for the idle in modern PC’s.
Maximum load
Next up is the maximum wattage that this unit can hold up before switching off.
| Maximum Load | Efficiency (in %) |
| 767W | 88.04 |
Even with the above 100% load efficiency is looking good. 767 Watts that’s an impressive value, but you wouldn’t want to come close to that for a longer period.
ErP Lot 6 Power Off wattage
We look at the powered-off status (ErP/EuP), productivity mode (we stress the CPUs), and finally gaming. The lower the wattage, the more efficient the PSU. It is that simple.
| ErP Lot 6 Power Off | value in Watts |
| - | 0.21 |
This is well below the 0.5 W requirement.
Ripple testing
First up the ATX12V V2.2 specification for DC output ripple:
| ATX12V Ver 2.2 Noise/Ripple Tolerance | |
| Output | Ripple (mV p-p) |
| +3.3 V | 50 |
| +5 V | 50 |
| +12 V | 120 |
Deepcool DQ750-M achieved the following results:
| AC Ripple (mV p-p) | +3.3 V | +5 V | +12 V |
| 75 W (10%) | 3 | 12 | 14 |
| 185 W (25%) | 5 | 14 | 18 |
| 375 W (50%) | 9 | 17 | 37 |
| 565 W (75%) | 13 | 20 | 48 |
| 750 W (100%) | 16 | 25 | 63 |
Results are well within the tolerance. Even in the worst case scenario (100% load), it was about half of the allowed result, not at all worrying.
