Why 99% of PC Power Supply Reviews Are Wrong

[Squilliam]

nordlead said:

Squilliam said:

But, but the PC has to have an internatl temperature fewer than 5 degrees above ambiant under load!

Btw, shouldn't a 80+ PSU have sustained efficiency throughout the wattage range? Thats what the certification is about.

I've seen some studies that showed the best cooling accomplished was with 2 fans (plus the PSU fan). They used identical computers with varying fan setups. The theory why it outdid the other combinations was the over abundance of turblance caused by the manny fans slowed the overall cooling rate. All I know, is that so long as you have decent airflow over the hot components you should be good. Adding 5 more fans generally doesn't help. I can't find the test I found before, or I'd link you to it.

I just looked up the 80+ (which I understood wrong) and you get sustained 80% minimum efficiency from 20% to 100% load, but some PSU's get 85% or better in the 50-80% load range. Obviously your best bet would be to look at the efficiency chart for the PSU you are considering and map it to your estimated power requirements.

Thanks for the info on the turbulance. Thats something I certainly hadn't though of. I suspect the number of case fans is essentially an enthusiast circle jerk. When I was into car modification it was the same deal, people even going as far as to cryo treat their components. Too much money and not enough (engineering) sense?

The main consideration when considering power supplies is probably cost. If you get a big and impracticle one which costs heaps with little benefit to the setup you're using you've paid too much. I suspect this is still a bigger factor than ~15W of wasted heat under load.