Case fan as CPU cooler?

  • Thread starter Thread starter Dave Smith
  • Start date Start date
D

Dave Smith

I've seen a couple of articles for adapting an 80mm fan to a 60mm
heatsink, but it seems to me that the most effective way to do this
would be to turn a rear fan around and duct the air flow to the
heatsink. It seems to me that this setup would supply more air and
cooler air to the sink. It also seems to me that there must be a good
reason why this isn't a common setup because I'm not the sharpest tool
in the shed and others must have tried this before and there must be a
reason why it doesn't work. So, why not?

Thanks.
 
Dave Smith said:
I've seen a couple of articles for adapting an 80mm fan to a 60mm
heatsink, but it seems to me that the most effective way to do this
would be to turn a rear fan around and duct the air flow to the
heatsink. It seems to me that this setup would supply more air and
cooler air to the sink. It also seems to me that there must be a good
reason why this isn't a common setup because I'm not the sharpest tool
in the shed and others must have tried this before and there must be a
reason why it doesn't work. So, why not?
Click to expand...

Where is all this air going to go once past the heatsink? You still need a
decent exhaust fan someplace.

I was considering something similar, except leave the back fan exhausting.
Suck the case air through the heatsink and out the back of the PC.
 
Where is all this air going to go once past the heatsink? You still need a
decent exhaust fan someplace.

I was considering something similar, except leave the back fan exhausting.
Suck the case air through the heatsink and out the back of the PC.
Click to expand...
My case has two rear exhaust fans. I figured to reverse one of them
and maybe move the other to the front. There's also the power supply
fan, though it doesn't seem to move much air.
 
The fan in power supply is typically more than sufficient to
exhaust case air. Case air temperature relative to room air
temperature is a function of chassis fan CFM.

Bottleneck in CPU cooling is not interior case temperature.
Bottleneck is mostly where thermal conductors change medium.
For example, CPU to thermal compound. Thermal compound to
heatsink. Heatsink to air. As airflow drastically increases,
heatsink cooling improvements become marginal. Most heatsink
cooling is with almost no airflow. As airflow increases, the
overall improvement in cooling decreases exponentially. As
airflow increases, then heatsink cools better, but not by that
much more. Therefore almost as much cooling is obtained
blowing air across heatsink as when driving airflow into
heatsink (as determined by how heatsink fins are designed for
that airflow).

Heatsink cooling is more affected by LFM across heatsink and
not by massive CFM increases of interior to external chassis
airflow. AFter the first 80 mm fan, that additional cooling
fans provide only marginal advantage. Direction of airflow
across heatsink also is irrelevant.

Why not mount a bigger fan on heatsink? A maximum weight
spec limits how much can be attached to CPU. Too much weight
can mean less cooling by compromising another thermal
bottleneck. CPU to heatsink interface is degraded by too much
weight and vibration.
 
The fan in power supply is typically more than sufficient to
exhaust case air. Case air temperature relative to room air
temperature is a function of chassis fan CFM.

Bottleneck in CPU cooling is not interior case temperature.
Bottleneck is mostly where thermal conductors change medium.
For example, CPU to thermal compound. Thermal compound to
heatsink. Heatsink to air. As airflow drastically increases,
heatsink cooling improvements become marginal. Most heatsink
cooling is with almost no airflow. As airflow increases, the
overall improvement in cooling decreases exponentially. As
airflow increases, then heatsink cools better, but not by that
much more. Therefore almost as much cooling is obtained
blowing air across heatsink as when driving airflow into
heatsink (as determined by how heatsink fins are designed for
that airflow).

Heatsink cooling is more affected by LFM across heatsink and
not by massive CFM increases of interior to external chassis
airflow. AFter the first 80 mm fan, that additional cooling
fans provide only marginal advantage. Direction of airflow
across heatsink also is irrelevant.

Why not mount a bigger fan on heatsink? A maximum weight
spec limits how much can be attached to CPU. Too much weight
can mean less cooling by compromising another thermal
bottleneck. CPU to heatsink interface is degraded by too much
weight and vibration.
Click to expand...

Thanks very much for your input. If I read your reply correctly,
ducting air from outside the case VIA an 80mm fan for CPU cooling
would work, but would not provida a significant cooling advantage. I
also was hoping to reduce noise as low RPM case fans are quieter than
typical 5000 RPM CPU fans. Is there anything inherently wrong with
this approach?
 
Thanks very much for your input. If I read your reply correctly,
ducting air from outside the case VIA an 80mm fan for CPU cooling
would work, but would not provida a significant cooling advantage. I
also was hoping to reduce noise as low RPM case fans are quieter than
typical 5000 RPM CPU fans. Is there anything inherently wrong with
this approach?
Click to expand...

If you reverse that fan to cool the CPU you need to reassess the
entire chassis cooling, since having the air intake there completely
alters chassis airflow. You'll have less air intake in the front and
even if you added a fan in front, you then have a singular point of
air entry instead of an even flow across many components. "Sometimes"
this single-point can be advantageous, if it's a fan in front of the
HDD bay, and even better if the air also has an unimpeded path towards
the video card, but on the other hand you still decrease airflow
around CDRW drive... some run hot and others don't. Also a front fan
"can" be the one most easily heard in an otherwise quiet system,
especially if that system is sitting under a desk. Only you can test
each change and note the results, but don't just look at the CPU temp
sensor and assume it's better.

Blowing air towards a part with a temp sensor on it will seem to help,
but it only tells you what that single point temp reading is, and that
part, the CPU in this situation, is designed to tolerate a little
heat. By having all fans in the rear blowing out you maximize the
airflow though the chassis. If you really need a cooler or quieter
CPU then you might consider buying a better heatsink, it is the
time-tested and proven course of action, and should help even if you
did end up running that rear fan reversed.


Dave
 
Dave Smith said:
I've seen a couple of articles for adapting an 80mm fan to a 60mm
heatsink, but it seems to me that the most effective way to do this
would be to turn a rear fan around and duct the air flow to the
heatsink. It seems to me that this setup would supply more air and
cooler air to the sink. It also seems to me that there must be a good
reason why this isn't a common setup because I'm not the sharpest tool
in the shed and others must have tried this before and there must be a
reason why it doesn't work. So, why not?

Thanks.
Click to expand...

i've built quite a few machines using a smaller case fan as a cou cooler
with excellent results.
it only takes a few minutes to make a mounting bracket
and i;ve been able to keep the cpu relatively cool...
plus the larger fans tend to be quite a bit quieter
 
Larger, less noisy 80 mm fans will move more air across the
heatsink. But they are heavier. Therefore they cannot be
mounted on CPU. All the better. If they blow air across
heatsink, then they are not mounted on heatsink and CPU, and
if fan fails, the heatsink still provides cooling. Less
cooling, but heatsink still cools CPU.

80 mm fans come in normal noise and lower noise versions,
with lower noise versions moving less air. An 80 mm fan
manufacturer may have between 10 and 20 different fans in that
same form factor. Each moving different amounts of air (CFM)
and making different levels of noise (dB).
 
i've built quite a few machines using a smaller case fan as a cou cooler
with excellent results.
it only takes a few minutes to make a mounting bracket
and i;ve been able to keep the cpu relatively cool...
plus the larger fans tend to be quite a bit quieter
Click to expand...

I've been using this way for near 2 years on my O/C XP1800,
http://superfan.freewebspace.com/
Not the,"Mummy" version.Quiet and cheap.Max under load temps just over
50 Deg C :P



--
Free Windows/PC help,
It's a G not a J in jmx to reply :)
http://www.geocities.com/sheppola/trouble.html
Free songs download,
http://artists.mp3s.com/artists/17/sheppard.html
 
Back
Top