There is something I don't quite understand. Most people who post in this forum are constantly downplaying the need for a higher powered psu. Fine- one does'nt need a 500w PSU to run their system, but based on the article by MikeC - with "all other things being equal" (meaning efficiency and so on) a higher powered PSU seems to be more capable of handling a fan running at 5v than a lower powered one:

From MikeC's article on PSU basics:

"The main benefit of higher power PSUs is when the airflow in the PSU is deliberately set very low in order to minimize noise. This means the PSU components will run hotter. All other things being equal, a higher rated PSU will be a better choice in such an application because its parts are generally rated for higher current and heat than a lower rated model."

IF that is the case, AND I plan on running my psu fan at a very low speed, would'nt it be wise to get an honestly rated high powered psu over a lower one if PSU failure is what I am worried about because I modded it with a very low spinning fan?

If the answer is yes.. then I don't understand why so many users on these forums advocate buying psu's that are only "sufficient" for the system, rather than overpowered for it.

Betty.

The truth is, most of the higher wattage PSUs aren't really equipped to deal with more heat, it's just that the lower watter PSU are overequipped to deal with what they've got. That's why there's generally no problems swapping in a slower fan.

But as I say again.. and as MikeC has written: "All other things being equal.."

It is not as if there are NO higher powered PSU's that are not honestly rated.

I see your logic Betty, and a 500w PS should have MUCH bigger HS's than a 300w... but usually don't.

It's a common misunderstanding that a higher wattage model would run cooler under light load than a lower wattage model. This is not the case, however.

If we take two PSUs, a 300W and a 400W one, and attach them to a 200W load, both PSUs will have to evacuate same amount of heat (caused by the AC-DC conversion), assuming that the efficiencies are the same. Usually efficiency rises to the peak value when running a PSU at ~2/3 of the max. rated power, so the 300W model would have a slight advantage compared to the 400W model.

Even if the components in the 400W PSU are rated for higher wattage, the situation is usually that if You take a 300W and 400W model from same manufacturer, the heatsinks will be more or less identical.

If assumed that the efficiencies are the same for both PSUs, running both supplies at their max. rated output means that the 400W model will have to dissipate more heat (from AC-DC conversion). In general this means that the manufacturer will have to opt for a faster spinning fan to keep the temps in spec. The manufacturer could also opt for larger heat sinks with same fan, but most likely the PCB layout would then have to be revised, so the fan switch will be far more profitable.

Which then means that the noise will be louder throughout the entire power range when compared to the 300W model, as the fan is a louder one.

Also, the higher-wattage units are more expensive than your-average-350W model. The price difference could be used for a replacement fan.

Cheers,

Jan

Since I'm being quoted here, I feel obliged to step in.

I stand by this comment:

With most decent brands, the highest rated model in a series DOES have higher rated parts than the lowest rated one. This means the higher power one will tolerate more heat longer than the lower power one. It does not always mean that the higher power one runs its fan at a lower for a given power output; that really depends on whether the PSU designers went to that level of detail to minimize noise.

Seasonic has mentioned that they use the higher heat headroom in their higher power models to shift the "turnover" or ramp-up point in the fan controller to a high temp on the higher power models to keep noise to a minimum up to a higher temp/power level on those models. They are the only ones who have mentioned such a strategy; it's also very difficult to veryfy without an absolute thermal-controlled test enviroment.

My original comment is most relevant for those who want to mod their PSUs with a slower fan. So if you take a PSU that normally runs its fan at 20 cfm at 40C, and replace it with one that gives only 10 cfm at the same temp, all other things being equal, a 500W model should withstand the higher heat considerably better than a 300W model.

But there's no clear evidence that in most PSU lines, the 500W model will run quieter than the 300W one -- at any temp. And as Jan Kivar mentions, with some higher power PSUs, a more powerful fan is used to accommodate the higher heat output potential at max power, which further negates the "quieter" potential of the bigger PSU.

Thanks for all of your comments. The conclusion seems to be the same though- a higher powered PSU (honestly rated) would be less prone to failure if one mods it with a very low spinning fan.

For example, currently I have a pc power and cooling turbo-cool 300 modded with a panaflo at about 6v. Sometimes I get worried about overheating and PSU failure. It seems clear that I should be less worried about overheating if I had the PC P&C 510 in my box since the components in that psu are rated for higher wattage (since this psu for example, is honestly rated)

Even more, it has been said in these forums that the effect on noise of a fan inside a PSU is much more pronounced than a case fan for example, because of air cavitation, obstacles in the air path, etc etc.. More of a reason why one would want to mod a psu with a low spinning fan (I'm not talking about a thermistor controlled one). And then the argument becomes circular of course.. if you want a low spinning fan then you need a psu that can handle becoming very hot.. and hence, a higher powered one.

Well worry is more the user than the PSU, no?

I've got two Fortron FSP300-60PNs modified with slow fans, and a generally hot running Antec modded with an AF80CT; being that I've run all these PSUs for quite a while now without a single hiccup, I'm not particularly worried about it. I've also yet to see another member of SPCR report their PSU dying from excessive heat due to a fan mod (haven't gone looking for it, specifically, but I don't recall anyone reporting such an event since I joined the community). The Antec is the 380 included with the Sonata; it's got a Northwod o/c'd to 3GHz, three 7200rpm hard drives and a 15000rpm hard drive attached, is under full CPU load 24/7 without any A/C in the hot room for a good several hours everyday while I'm work; works flawlessly. The modifed Fortron in Sigma One was my first to be modded and it barely even runs warm (of course it's only a single 4200rpm hard drive, a Northwood 2.2 and a Radeon 9600). Finally I have the very same PSU in Gamma Two, which has a Mobile Barton at 2556MHz and 2.00volts, as well as an overclocked Radeon 9800 nonPro and a single Samsung SP80; that PSU does run hot, but it has yet to show any signs of failure.

Any good brand of PSU with efficient operation will serve just fine, even with a slow fan, unless under really extreme circumstances; circumstances more extreme than those that I put my PSUs under, in terms of both heat and load. Perhaps a 6800 Ultra, overvolted, plus an overvolted, overclocked Prescott, plus several 15000rpm hard drives, would overdo it for an underfanned 400watt PSU from a good brand (SeaSonic, Enermax or Fortron, for example); but who's to say that's not already too much for an unmodified 400watt PSU? And what SPCR member would kill themselves by trying to get silent/near-silent acoustic performance out of such an insane monster rig?!?

-Ed