PC Power & Cooling CPU Cooler?

  • Thread starter Thread starter Philip Wright
  • Start date Start date
P

Philip Wright

I am a big fan of PC Power & Cooling power supplies
and am considering using one of their CPU-Cool 478L
cpu coolers on a Prescott 3.2E processor. Does anyone
have any experience with these coolers on a Prescott?

Thanks
 
Philip Wright said:
I am a big fan of PC Power & Cooling power supplies
and am considering using one of their CPU-Cool 478L
cpu coolers on a Prescott 3.2E processor. Does anyone
have any experience with these coolers on a Prescott?

Thanks
Click to expand...

That is a low profile cooler intended for rack mount computers.
Thermal resistance is 0.44C/W. You want something with a
thermal resistance closer to 0.25C/W. Plug the numbers
into this equation, to see what temp the CPU would run at.
Max_Power (TDP) is 89 watts for the 3.2E.

CPU_temp = Case_Air_Temp + (Thermal_Resistance_degC/Watt * Max_Power)

With that cooler, you'll be in permanent thermal throttle
mode.

HTH,
Paul
 
That is a low profile cooler intended for rack mount computers.
Thermal resistance is 0.44C/W. You want something with a
thermal resistance closer to 0.25C/W. Plug the numbers
Click to expand...

Is this C/W value affected by the fan used? It would seem intuitive
that the higher the fan cfm, the lower this ratio should be?

Is there a formula that calculates how much the C/W changes according
to fan cfm?
 
Is this C/W value affected by the fan used? It would seem intuitive
that the higher the fan cfm, the lower this ratio should be?

Is there a formula that calculates how much the C/W changes according
to fan cfm?
Click to expand...

I don't know if there is a formula as such, as the effectiveness
of increased CFM, will depend on the surface finish, the shape
of the object etc. Turbulent or laminar flow. Boundary layer.

If you check this review site, you can find some info:
http://www.overclockers.com/topiclist/index15.asp#HEATSINK REVIEWS AND TIPS

The plot at the bottom of this page, shows the general form to
expect. The CFM would have to be measured with an air flow device,
in order for this plot to be done versus CFM, instead of RPM.
The CFM of a fan depends on how much resistance to air flow
there is, so the CFM rating on the fan itself is not that
precise when a load is applied to it.

http://www.overclockers.com/articles1043/

A thermal engineer knows the answers to this stuff, and I don't :-)

HTH,
Paul
 
Paul said:
I don't know if there is a formula as such, as the effectiveness
of increased CFM, will depend on the surface finish, the shape
of the object etc. Turbulent or laminar flow. Boundary layer.

If you check this review site, you can find some info:
http://www.overclockers.com/topiclist/index15.asp#HEATSINK REVIEWS AND TIPS

The plot at the bottom of this page, shows the general form to
expect. The CFM would have to be measured with an air flow device,
in order for this plot to be done versus CFM, instead of RPM.
The CFM of a fan depends on how much resistance to air flow
there is, so the CFM rating on the fan itself is not that
precise when a load is applied to it.

http://www.overclockers.com/articles1043/

A thermal engineer knows the answers to this stuff, and I don't :-)

HTH,
Paul
Click to expand...


Thanks for all the info. What cooling devices are group readers
using on Prescott cpu's? This will be going into a P4C800-E Deluxe.

Thanks again!
 
Is there a formula that calculates how much the C/W changes according
to fan cfm?

Yes, the formula for C/W according to fan cfm is specific to a given heat
sink design and a given fan size (80x25, 92x25, 70x15). The formula is
derived from empirical lab data and is a best guess based on the lab data.
Here are a few of the variables; flow direction of the fan, does the fan
operate at a constance speed, are there dead air spots, air pressure behind
the fan, thermal properties of the HS material, air resistance of the fin
design, homogenous composition of the HS material, fin design of the HS, air
flow properties of the HS (lam or turb and at what point in the HS), ect.
All these variables can change with time. This is why a vantec tornado fan
(80x38) may not cool any better than a high speed sunon (80x25) or why the
thermaltake V12 uses a three blade fan (higher pressure). There are allot
of factors involved and most are dynamic. A good rule of thumb is to use a
high pressure fan on a microfin HS that will produce a constance cfm. For a
non-microfin HS use a high flow cfm fan. It is always a good ideal to use a
variable resistor to control the fan speed to optimize the noise/cooling
efficiency ratio.

A high cfm may not always be the answer to better cooling. For instance I
have a TT slientboost HS with a vantec tornado fan on a variable resistor at
12V and 5600 RPM the differential cooling is the slightly better (<1°C) than
~9V and 4500 RPM but the noise level of the fan increases exponential. A
low pressure high cfm (80x25 sunon or TT) fan does very poorly on this HS
due to the micofin design where as a high pressure fan relatively lower RPM
works great. The TT V7 which has a wide spaced fin design and copper core
works best with a low pressure high cfm fan (92x25). The swiftech HS design
appears to work better with a high cfm fan and high pressure although some
people have gotten good results with a high cfm low pressure fan pulling air
through the swiftech HS rather than pushing air though the HS due to it's
radial pin design.
 
Back
Top