PSU Fans


Limp Gawd
Aug 17, 2003
I think that main source of nosie in my case is comming from the PSU fans, its one of those that has one on the bottom and one at the back, can I just replace them with a normal 3 pin fan? I was thinking a zalman one since they are meant to be quiet. Do I really need the one on the bottom, this may be the noisy one?
Keith130 said:
BUMP, so no one has replaced their PSU fans?

They have a 3-pin connector but it will probably be a different plug than your standard size.
Thanks, if they are a differnt size will brute force work or just solder the wires to the connections and make sure they cant short out?
I've replaced the fan in one of my PSU's before when the stock fan motor burnt out.
All I had to do is solder the red and black wires where the old ones were.
So the interal PSU fans are two pin then? Are two pin fans cheaper than 3 pin ones since they dont have a speed control?
replacing the fans is a common mod
however by reducing the efficiency of the thrrmal solution,
your reducing the power output the PSU will produce,
(all the way up to some threshold of reliable performance)
in addition to shortening its life expectancy

PSU Fundamentals and Recommended Units article
Simulate an inside-PC thermal environment while the PSU is under load: When the PSU is tested, the temperature of the environment is realistically warm, and tied directly to its actual output.
*UL/CSA do load testing in a controlled steady-state temperature environment -- whatever the manufacturer specifies. These are highly technical environmentally controlled testing chambers (usually quite small like a few cubit feet at most) in which you can "dial in" the temp & humidity and atmospheric pressure you want. This is NOT realistic.... let me explain.

You know that the AC power drawn by a PC is equal to the heat generated inside that PC.

So you have this PSU. You put it under load on the test bench. The room temp is 25C.

The PSU draws 240W from the 120VAC line and pumps 180W DC into the load tester. This means you have a 75% efficiency PSU. The heat generated inside the PSU is 60W. The PSU handles it fine.

It is a 300W PSU. You set the load to 300W and the PSU draws 400W from the wall to degenrate this power in DC. the voltages all look stable, the PSU looks like it could do this all day. Great, right?

Not really. Inside a PC, even at the first power level of 180W DC output, the PSU would not only be dealing with the power it generates within itself (that 60W) but also the heat from all the other components. In a tower case, probably at least 50% of the power drawn by the components actually rises up as HEAT to the top of the inside of the case where the PSU is. Hmmm. half of 180W is 90W... so the PSU is dealing with around 150W of heat within itself. And I'd say that's a conservative estimate.

How hot would that be? Depends on the size of the PC case, and the airflow through it....

In thermal testing setup, the DC power output of the PSU runs through the resistor banks of the load tester. The heated air from these resistors is fed into a box that has about the same volume as a mid-tower PC case with components in it. So as the PSU load goes higher, the temperature of its environment -- the box or simulated PC case -- gets hotter.

With 400W output the temperature at the intake side of the PSU has gone above 40C in my thermal simulation case.

Q: Why is this important?

A. Because the power output capability of a PSU is dependent on its operating temperature.

You may have seen specifications like this, which are common: 0ºC ~25ºC for full rating of load, decrease to zero Watts O/P at 70ºC

Examine what that says. Full power (let's say 400W) is available when the unit is operating in a temperature of 0ºC ~25ºC. Hmmm. Think about this.

Have you ever felt air blown out of a PSU in a PC running absolutely full tilt (which it would have to do to get anywhere near 400W output) that felt cool to the fingers? 25ºC airflow would feel exactly that: Cool, given that normal body temperature is 37 °C.

So this PSU cannot deliver full rated power when its temperature goes over 25ºC. OK, what happens to the max power output capacity above that temp? Max power output decreases gradually so that by the time the PSU temp reaches 70ºC, the PSU cannot deliver any power at all. So if you assume that this power drop as temp rises is linear, then max power capacity will drop by ~9W for every degree over 25ºC.

In every PC I have ever examined, the air temp around the PSU would measure at least 30~35ºC -- under almost any load. And if/when it is pushed, 40ºC is nothing at all, especially just above the CPU area where the PSU sites.

So... here's a simple table of max power output at different temps, using that linear ~9W drop for every degree over 25ºC...

25C - 400W
30C - 355W
35C - 310W
40C - 265W
45C - 220W

Hmmm... that macho 400W PSU can only produce 265W at 40C!! And if you have a really hot loaded PC, well, you could be even further down -- when you need it most! That's why testing of PSUs in realistic thermal conditions is of interest to Overclockers and High Performance PC enthusiasts. It's what I do with my PSU tests at SPCR.
Keith130 said:
So the interal PSU fans are two pin then? Are two pin fans cheaper than 3 pin ones since they dont have a speed control?
I can't generalize for every power supply, but I've seen a lot of 2-pins. These fans are cheaper, but it doesn't mean that the fans aren't speed-controlled. The "speed control" may be external to the fan, and the power supply just controls the supply voltage according to thermal conditions or manual input.