Test results
Our quartet of PSUs all have different power ratings, so we must take this into account for our results. The table below shows the current draws we reached and the voltages provided at those loads. We've then calculated output power and efficiency. Where an output power is shown in green, the PSU has reached its rated output. Red means this wasn't the case.
Readings under load | Power & efficiency | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
PSU | +12V1 | +12V2 | +5V | +3.3V | -12V | +5VSB | Input power | Ideal output | Actual output | Efficiency | |
NorthQ NQ-4800-400 | Voltage | 11.935 | 11.92 | 4.8825 | 3.08 | -12.785 | 4.89 | 620 | 352 | 347.42 | 56.8% |
Current | 13 | 10 | 7 | 10 | 0.25 | 1 | |||||
be quiet! Dark Power Pro 430W | Voltage | 11.73 | 11.715 | 4.8 | 3.0775 | -12.3 | 4.89 | 556 | 446.2 | 432.64 | 80.3% |
Current | 13 | 13 | 16 | 14 | 0.25 | 1 | |||||
be quiet! Dark Power 470W | Voltage | 11.8275 | 11.8125 | 4.785 | 3.09 | -12.65 | 4.9 | 608 | 483.5 | 471.50 | 79.5% |
Current | 14 | 14 | 18 | 15 | 0.25 | 1 | |||||
FSP Zen 400W | Voltage | 11.632 | 11.725 | 4.905 | 3.095 | -12.8 | 4.8575 | 463.6 | 411.7 | 400.24 | 88.8% |
Current | 14 | 13 | 10 | 9 | 0.25 | 1 |
Both be quiet! Dark Power PSUs reach their rated outputs and therefore pass our tests. They do this at ~80% efficiency, which if you compare against our figures for the 34-way round-up, is a good figure.
FSP's Zen PSU also passes our load testing, proving to be the most efficient design too, coming in at a very good 88% efficiency. It's not just good for your power bill either. In the Zen's fanless design it's important that as little power of possible is wasted as heat. There was some ripple on the two 12V rails, but nothing major enough for concern.
The NorthQ PSU doesn't quite live up to expectations, however. We had tried to draw 11 amps from the +5V rail on the PSU, but this would cause the PSU to shut itself off, possibly overcurrent protection or some other failsafe. Dropping this down to 7, but tweaking both +12V rails to 12 and 13 amps, respectively, gave us the same problem. We were finally able to get the PSU to stay functioning at the currents shown in the above table. However, after a whole 10 seconds the PSU's input power started to rise dramatically... a sign of increased resistance from rising temperatures and impending doom. Then, something blew in side the PSU and the NorthQ 400W was laid to rest.
All of this happened at just 350W, 50W off the stated output power. So, we can assume that the 400W rating is a peak rating, not a sustained output. Further, it can't cope with that for very long, heating up and falling over in a short space of time.
Temperatures
With the NorthQ dead, we were left with the FSP and be quiet! PSUs for temperature readings. Here's a lovely table depicting what our readings revealed:
PSU | Ambient (ºC) | PSU idle (ºC) | PSU after load testing (ºC) |
---|---|---|---|
be quiet! Darkpower Pro 430W | 30.5 | 31 | 39.5 |
be quiet! Darkpower 470W | 30.5 | 31 | 40 |
FSP Zen 400W | 22 | 22 | 38.5 |
A temperature delta between idle and post-testing of 9.5-10ºC for both be quiet! PSUs, then. It is worth considering that a PSU in a 'real' environment isn't likely to have the luxury of ambient temperatures at its inlet. Instead it will have air from the PC's innards, which could well have been warmed by all of the other components in the chassis. Once again, a quick flick over to our Taoyuan results show that the be quiet! PSUs do a pretty good job of keeping themselves cool, and they do it fairly quietly too (as the name would have you hope).
The Zen was tested on a different day, thus the different ambient temperature. As you might expect, its temperature delta is higher, but the absolute figure is in the same region as the be quiet! PSUs'. We also took a reading off the fins atop the PSU. These were at 57ºC; clearly where at least some of the heat is being drawn to. Given the PSU passed our load tests, we'd say the cooling is adequate. However, as with any passively cooled component, we recommend installing it into a case with good ventilation.