Testing results and temperature musings
Like in our 400W-500W roundup, we calculate output power and efficiency based on the pass or fail settings the PSU made it to during torture testing on PSU-testing harness. The results are as follows.| Readings under load | Power & efficiency | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PSU | +12V1 | +12V2 | +5V | +3.3V | -12V | +5VSB | Input power | Ideal output | Actual output | Efficiency | |||||
| Akasa GreenPower 80+ 500W | Voltage | 11.895 | 11.58 | 4.782 | 3.075 | -12.192 | 4.835 | 611W | 516W | 502.21W | 82.2% | ||||
| Current | 18 | 17 | 11 | 10 | 0.25 | 1 | |||||||||
| FSP Blue Storm II 500W | Voltage | 11.992 | 11.85 | 4.935 | 3.105 | -12.022 | 4.927 | 617W | 507.7W | 502.53W | 81.4% | ||||
| Current | 18 | 17 | 10 | 9 | 0.25 | 1 | |||||||||
Both the FSP and Akasa supplies reached their rated output, which means they pass our tests. They also manage to do this at around 80% efficiency, which we also consider to be decent performance
However we must point out that the Akasa does manage to live up to its claims of 80 per cent efficiency, whereas the FSP didn't manage the 85 per cent-plus efficiency that the blurb stated it would. As you can see from the table, both supplies were almost identical, and produced their power at just over 80 per cent efficiency.
[advert]The FSP also displayed a small amount of rippling on the second PEG connector, which is connected to the primary 12V rail. This isn't a good sign, but it did manage to stay within specification and pass our tests. Both of the PEG connectors on the FSP were wired to separate rails, however.
The Akasa placed both of the PEG connectors on the same rail. In fact, the ATX, PEG, SATA, and the second half of the EPS connector were all wired to the primary 12V rail! The second 12V rail has the first half of the EPS, which also serves as the P4 plug, attached.
This layout may put a heavy demand, through the PEG connectors, on to the primary 12V rail. We doubt if it could handle the power draw of a pair of GeForce 7950GX2 or Radeon X1950XTX cards in a dual-GPU setup. However, we tested the Akasa with a pair of GeForce 7900GTXs, and an Intel Core 2 Extreme X6800 and it ran for a couple of hours without any problems. Akasa also claims to have run a pair of GeForce 7800GTXs with and AMD Athlon 64 FX-62 for extended periods without any problems.
When we simulated a heavy load with the benchmark rig, though, we quickly encountered problems. The secondary rail was fine, only needing a 5A draw for the P4 socket. The primary rail had problems with overload, as it was loaded up with 11A to simulate a heavy load from the ATX socket, SATA, and Molex connectors, as well as a further 10A to simulate the two loaded PEG connectors. This adds up to an unwieldy 21A, which is 3A more than the supplies maximum load. Unsurprisingly, the supply shut down very swiftly when we started dishing out this punishment.
Temperature Testing
The table below shows the peak temperatures the supplies reached during testing. Ambient temperature is a recording of the temperature in the room during the test, which allows us to make a direct comparison between the two supplies although they were tested at different times.| PSU | Ambient (°C) | PSU idle (°C) | PSU after load testing (°C) | Calculated Delta (°C) (load - idle) |
|---|---|---|---|---|
| Akasa GreenPower 80+ 500W | 21 | 23 | 31 | 8 |
| FSP Blue Storm II 500W | 21.5 | 23.5 | 28 | 4.5 |
Fortunately the temperature in the test environment stayed relatively stable over the course of the testing. It only rose half a degree between the testing of the Akasa and the FSP.
The supplies themselves were tested outside of a case, while attached to a synthetic testing rig, so these temperatures may not be what you'd see when they are in a system, which is likely to be at least 10°C higher, unless you have a blow-hole exhaust fan close to the PSU.
The results in our table do not suffer from these complications. You can see there is a minor difference of half a degree in ambient temperature, which would suggest that the Akasa and FSP perform identically at idle. However, under our load testing the Akasa runs three degrees warmer than the FSP. This could be because the Akasa is sharing the primary 12V rail between the two PEG connectors, while the FSP puts on each 12V rail, or because the fan on the FSP is running slightly faster. Nonetheless the difference is minimal, and neither PSU is running dangerously hot.
