dhanson865 wrote:I firmly believe that the chart is worthy of its own sticky in the forums and/or a permanent spot on the recommended PSU article.
Fair enough. Will see about getting it done... but one problem is some of the PSUs on the lists were tested in the anechoic chamber, some were not.
dhanson865 wrote:Essentially Mike took the 10 scale and changed it to a 3 scale.
3 = Editor's Choice
2 = SPCR Recommended
1 = Everybody Else
More or less.
guises wrote:So what I'd suggest instead of this is to split it up by the maximum wattage at which the power supplies will remain quiet.
The point is that this maximum wattage is not a singular point. but rather, a sliding scale. Your very thinking about this suggests a linear scale -- it's not. It depends heavily on ambient temp, case design (specifically PSU air intake), user perception of "quiet" and of "cool", etc. The simple fact is that among the high end cases that many SPCR readers are buying and using, a huge number have some kind of independent fresh air intake for the PSU. They come from every brand, it's not like 4 years ago when the P180 and one or two half-thought out LianLi's were the only ones with this feature. It's quite commonplace.
So is SPCR's PSU testing system still appropos? For the cases that don't have a separate PSU intake, yes, but not for the ones I refer to. And it's my semi-educated guess that most SPCR readers, if they're really into silent computing, seek out such cases. For them, the old PSU ranking system was only about half right. The noise level at low loads was right... and despite that, most would not have heard much difference between the PSUs that idle <~18 dBA/1m. Everything else makes more noise in most systems! And above 150~250W, which is where a lot of PSU fans start speeding up in the test rig, the data would have been plain wrong. If the case used had an independent PSU intake vent, the fan ramp up point might extend up by 200W... in which case, for most systems (which don't take more than 300W), the mid/high power sonic differences seen in the SPCR tests simply would not have shown up.
Yes, you guessed it, the PSU test system is also undergoing careful review.