PSU for cooling the CPU??

[digicat]

Hey all. I want to get a new PSU, and have been checking out my options. I've been reading the forums, as well as doing other research to see what I might need to pick up. At the same time, I'm also trying to keep the noise and temperature in the case down.

I see a lot of PSU's on the market today with fans under them to pull heat off of the CPU and exhaust through the back of the PSU. At first this seemed like a handy thing to have to help get rid of excess hot air in the case, but then I got to thinking, a PSU needs to be kept cool too, why would you want to have a constant flow of heated air going through it?

I see these PSU's all over the market, from all the major PSU companies. Either the heat isn't an issue, or it's a marketing ploy, or something else. Which is it?

So, on paper, it seems like a bad idea at first, but does it really matter? Does anyone with some real-world experience have some input here?

 

[BillR]

Hey all. I want to get a new PSU, and have been checking out my options. I've been reading the forums, as well as doing other research to see what I might need to pick up. At the same time, I'm also trying to keep the noise and temperature in the case down.

I see a lot of PSU's on the market today with fans under them to pull heat off of the CPU and exhaust through the back of the PSU. At first this seemed like a handy thing to have to help get rid of excess hot air in the case, but then I got to thinking, a PSU needs to be kept cool too, why would you want to have a constant flow of heated air going through it?

I see these PSU's all over the market, from all the major PSU companies. Either the heat isn't an issue, or it's a marketing ploy, or something else. Which is it?

So, on paper, it seems like a bad idea at first, but does it really matter? Does anyone with some real-world experience have some input here?

It’s one of those things that fall under the heading of “functionally compromised” or you could term it “compromised functionality”

Manufactures use the technique because it’s the cheap and easy way out. Lot’s of us do it because it’s all we know or have been taught.

You are right though; it’s a fundamentally bad idea.

If you read deeper into this and a few other threads many of us have started using internal ductwork to redirect heat to where it belongs, outside the computer. Thin foam board, cardboard, plastic and metal are all materials in use to re-engineer the airflow inside your machine.

I even have one machine where I have built the PSU into a box outside the main computer tower.

A cool PSU is a happy PSU, the same goes for CPU’s, Hard drives, ram etc.

Good observation on your part

 

[gee]

design flaw? hell no.

Don't forget that the great majority of PCs don't have any case exhaust fans other than the one in the power supply... In this case it doesn't matter wether the PSU gets its intake air directly from the CPU or from the complete opposite end of the case, it's still going to be experiencing the same temperature.

If you want to keep your PSU cool, add more exhaust fans to your case. But don't go to any great lengths until you're sure you need it.

If you really want to verify that your "PSU is happy", then measure your case's inside air temperature by taping a medical thermometer somewhere inside - don't use your mobo 'case temp' sensor. Then look at the derating curve for your power supply, and make sure that the load you're drawing from it is less than the derated power at that particular point.

If your power supply doesn't have a derating curve published anywhere, then it sucks

 

[Ice Czar]

derating curve

see its when you toss out little terms like that
that broad new vistas of search queries happen

and returns like
Clearing the Fog that surrounds derating curves @ Power Design 365
(registration required)

strangely I have to agree with both gee and BillR
(what they are saying isnt as diametrically opposed as it would seem at first glance)

the PSU has always exhausted the case, however, unlike OEM manufacturers that run Computational Fluid Dynamics software to model the thermal gardients in a PC, we get to use trial and error

the adoption of a bottom fan to directly exhaust the cpu heatsink dates back to the Athlon
when AMD and Enermax cooked up the new builders recommendation
AMD Thermal, Mechanical and chassis Cooling Design Guide
(Enermax was the only enthusiast manufacturer that had a bottom fan at the time, and I bought one simply to conform to the builders guide)

these forums in particular are famous for cramming as much high draw, high heat components as possible into a box
so dealing with that heat, and the draw of those components does lead to alot of creative thermal solutions, you can increase the capacity of your supply with a cooler operating temperature, if it needs it.

there has been a few interesting mainstream developments Ive recently seen on a few mobos, including a collector shroud around the CPU socket that exhaust the heatsink
getting rid of that air as quickly as possible just makes damn good sense, but it is starting to impact the PSU, especially when you consider the output of a new P4 is over 100 watts of heat

as gee suggested, mapping your case is a good idea, I have a digidoc to do that, but dont use a single spot

a cut and paste
------------------------------------------------------------------------------------------------------------------
there are three basic aspects of any thermal solution

1. the heat transfer interface, how large it is, it design and thermal resistance
(in this case the heatsink in the PSU)

2. The transfer medium be it air or water, what volume (and for air static pressure)
for a given period of time (with additional variables like turbulence, and reistance)

3. The Temperature Differential

while the efficiency of a given supply at a given load will determine the high side of a temperature differential, by employing a cooler airstream you have the option to lower the lower side of the differential

a real good computer oriented guide to thermodynamics
General Heat Transfer Guide

Heat transfer is all about temperature differentials. Conduction through materials and convection away from surfaces is proportional to the temperature differential that exists. Basically the thermal resistance of a given solution changes with the temperature differential, which is why thermal solutions are rated in °C/W which leaves that variable ambient open but gives you a formula

The heat transfer through the wall follows a simple equation:
Q=k/L(T1-T2)

We can draw some interesting conclusions from this equation. First, heat transfer is proportional to the temperature difference on the object. If the temperature differential doubles, the heat transferred doubles. Second, the conduction coefficient "k" is proportional to heat transfer. If the conduction coefficient doubles, the heat transfer doubles Alternatively, for the same differential temperature, twice as much heat may be transferred. The final observation is "L". As thickness increases, heat transfer decreases. Alternatively, to maintain the same heat transfer through a material twice as thick requires twice the temperature differential.

while that addresses conductance, there is a corrallary for convection (transfer from the heatsink to a fluid (air\water\ect)
its known as Newton's Law of Cooling

your pretty much stuck with the heatsink a supply has, but altering the airflow isnt the only tool left to you

the first senerio you see alot of duct building is with folks trying to silence a PC
they start be compromising the thermal solution of a PSU by reducing the airflow and normally pressure
http://forums.silentpcreview.com/viewtopic.php?t=8450
next you see it with paranoids like me, that are looking for as much capacity and longevity as possible, like any mechanical or electronic device, the PSU will benefit from a lower operating temperature (within reason)
the Arrhenius equation
roughly translating to this rule of thumb*
Each 10°C (18°F) temperature rise reduces component life by 50%
Conversely, each 10°C (18°F) temperature reduction increases component life by 100%.

* additional variables of course apply

 

[BillR]

Something that has always been a source of fascination with me is the battle between the engineer and the technician.

The engineer will with great care and all the tools of his trade design a project and hand it to the technician for final build. The technician will sometimes look at the drawings and want to ask the engineer; “Are you on drugs?”

It’s the endless and timeless battle of the “Theoretical” and the “Practical”

I tend more toward the practical. Math is not my forte. I learned a long time ago that most published specs are as a rule worthless if taken as is, that is “unqualified”. Unfortunately that is the way most people take specs, as gospel truth.

Ice Czar seems to have acquired the ability to blend both methods into something (as a rule) understandable as well as usable.

This is one of the better forums on this site where one can actually learn something.

I do however know that when my PSU is outputting air at 80 degrees F in a 75 degree room and my CPU is running at 100% load at 100 to 105 degrees F in that same system, I’m probably close to being on the right track