Solved Cooling

[seekermeister]

Yes, good thought. I just checked it in the BIOS and it is set at Full On, and it's running at 1939 RPM, which is over 300 RPM past it's rated speed.

 

[A Guy]

AS5 also takes up to 200 hours to reach maximum cooling. But it certainly won't drop a major amount.

Due to the unique shape and sizes of the particles in Arctic Silver 5's conductive matrix, it will take a up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink or with a low speed fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop 2C to 5C over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired.

The thought is that the cooler isn't seating correctly. Can you manually push down on the cooler while testing to see if there is a dramatic drop in temps?

A Guy

 

[seekermeister]

It's seated properly.

 

[seekermeister]

I found this page:

Spire AMD Heatsink Temperature Comparisons - Eclipse II SP984B1-V2 FrostyTech Review

Which has an interesting comparison chart. It makes it easy to compare on the basis of temperature and sound, but one thing I have figured out yet is that since some of the better coolers are so massive, is how to predetermine whether they would fit properly in my rig. It isn't that there isn't enough case space, but the fact that even with my Cooler Master Hyper TX3, there is only ~ 1/2" of space between it and the chipset HS/F sitting next to it. Anything less than that, I wouldn't be able to manage taking the fan off and on with the little wire clip that hold it on. Eyeballing it, it appears that even this cooler would not fit over the HS/F, it it were wider.

 

[seekermeister]

Due to the size/space considerations that I mentioned, my thinking has turned toward using a water cooler for the CPU, but after looking at some of the ads for the Corsair H50 and H70, which are the only ones that I found that states that they are suitable for an AM3 rig, I still have some questions. Looking at the photos of the hardware that comes with them, I didn't see any attachment that appeared to be designed for an AMD processor. One on the H70 stated that it would require buying the attachment separately from Corsair...is that true of all Corsair water coolers? Is there some other brand that actually includes the proper
attachment?

EDIT: Also, I tried to find a chart that would give actual temperatures with water coolers, something like the one that I linked in post #24, but didn't find it. Anyone know where to find one?

 

[pparks1]

It is my understanding that the H50/H70 will fit an AM2/AM3 layout right out of the box.

 

[A Guy]

This has your current cooler and the H50 as well as other air choices.

Corsair H50 CPU cooler review

Q: Is it compatible with my motherboard and CPU?
A: The first release of the H50 had a part number of CWCH50 and it shipped with Intel 775 and 1366 brackets, AMD AM2/AM3 and Intel 1156 brackets are available separately on our webstore if you need them. The current model of the H50 is the CWCH50-1 and it now includes all brackets (Intel 775, 1156, 1366, AMD AM2, AM3) in the box.

Source

A Guy

 

[pparks1]

Seekermeister,

There is also a possibility that you have "too many" case fans installed in your box. They might be pulling so much air out of the box that you are no longer getting sufficient airflow through your heatsink to pull away the warmer air generated by the CPU. This is further reinforced by the fact that when you turned around your CPU fan, it made things worse. Sucking air through your cooler seemed to work better than pushing it through...which leads me to believe your exhaust fans are pulling the air out too fast for it to effectively cool.

If you can turn down the speeds of these case fans, you might want to try that. I run 2 x 120mm exhaust case fans in my Antec P182. One at the top blowing out, and 1 in the top back blowing out. I leave them on their lowest setting. I run an Intel Q9550 overclocked from 2.83ghz to 3.40ghz on a stock Intel cooler and I've dropped my vCore from 1.25v to 1.175 I believe and I idle in the low 30's and max out around 62-65...which is well within spec on my CPU.

Common sense says, "you can never have too many fans", but the laws of thermal dynamics aren't exactly governed by common sense notions.

 

[seekermeister]

This has your current cooler and the H50 as well as other air choices.

Corsair H50 CPU cooler review

Q: Is it compatible with my motherboard and CPU?
A: The first release of the H50 had a part number of CWCH50 and it shipped with Intel 775 and 1366 brackets, AMD AM2/AM3 and Intel 1156 brackets are available separately on our webstore if you need them. The current model of the H50 is the CWCH50-1 and it now includes all brackets (Intel 775, 1156, 1366, AMD AM2, AM3) in the box.

Source

A Guy

I guess that all of the vendors on eBay are selling the older versions, but it is good to know what to look for. That leaves one thing that bothers me about using these particular water coolers, being the apparent length of the plastic tubes running between the water block and the radiator. They are not long enough to extend through the holes provided in my case, to locate the radiator outside. Even if they were, since the system comes as a sealed loop, it wouldn't be readily easy to disconnect the tubes to run them through the holes.

There are only two locations that the radiator would be able to fit internally with my case, being either to replace the rear exhaust fan, which it would seem that it would block general airflow of the rest of the case, or to replace my 200mm blowhole fan on top, which seems like a waste. If I did put it at the blowhole, that would put the radiator directly above all of the other electronic components, which I would dislike, because of the possibility of a leak, destroying the entire computer. If I go to a watercooled system, I would like to get as much of the system outside of the case as possible.

 

[pparks1]

I think it's fairly rare for the Corsair systems to leak, but it certain can and does happen.

Personally, I don't think I would go for a liquid cooling solution. Seems like an unnecessary risk for many builds. I understand the desire to get temps as cool as possible, but there is a point where effort does not equal return on investment.

 

[seekermeister]

Normally, I would agree, but I'm having enough of a heat problem, that I can not do with the computer the very thing that I updated to this build for...encoding. Therefore if it requires spending more to make it work, I shall.

 

[pparks1]

Did you have any thoughts about my comments about the possibility of their being too much airflow, thus impeding proper cooling of your CPU? Can you turn down the fans or shut at least one down and then test to ensure that you aren't starving the CPU heatsink from the air it needs?

 

[seekermeister]

I considered it, but since I'm a retired pilot, with a good knowledge of aerodynamics, which relates to this issue better than thermo dynamics, I do not tend to feel that the idea will work. Yet, I have not discarded it, simply put it on hold for the moment, and will experiment with it some later. For thermo dynamics to work in the fashion needed in this case, it would require sufficient airflow through the cooler. I do not believe that excessive airflow would hinder cooling, if anything, it would improve it. The Bernoulli Principle states that when a liquid or gas is sped up, that the pressure is decreased, and in turn, the temperature decreased. The only way that would not necessarily be true is if the mass is sped up to near supersonic speeds, where a heating due to compression comes into play.

The only thing that I can see where airspeed could play a part in the problem, is if somehow, in a fashion that I do not understand, the airflow is being obstructed or detoured to a path other than intended.

 

[LiquidSnak]

It's less aero and thermodynamics in an enclosed space such as a tower. It becomes a strange matter of fluid dynamics. Something such as the venturi principle could very well apply if there is more air moving around your cooler, thus making it unavailable for the CPU fan to suck in.

 

[seekermeister]

Theoretically, fluid and aerodynamics operate the same. EDIT: With the exception that a gas is compressible and a fluid is not.

EDIT: I suppose that if there was a very high speed airflow across the intake of the fan, it could disturb airflow into the fan, but airspeeds in a case are not that high. Once air enters the fan, it would cross the cooling fins, and would receive thermal energy as designed. The only exception that I can think of, is if after passing through the fan, the air is diverted from taking a path directly across the fins, and go at an angle that shortens the path of exposure. This is a normal aspect of airflow that is near the edge of an airfoil (at least with an airfoil shaped as a wing. With a flat fin, not so much so, but still to a point), but not across the majority of the surface.

 

[LiquidSnak]

Yes, but aerodynamics apply better in an open space, and fluid applies better in an enclosed space, but yes, they are very similar

 

[seekermeister]

The only way that I can see that it matters as far as whether the space is open or enclosed, as far as liquid/gas behavior, is the compressibility factor that I mentioned in the edits above. However I believe that compression is a minor factor in the topic of discussion.

EDIT: Another thought occurred to me...with masses such as these, another trait that differs is viscosity, which may be a factor in surface adhesion, but if the surface is smooth, such as a cooler fin, that viscosity shouldn't impart anything strange to the airflow pattern.

 

[LiquidSnak]

hmm... There might be an easy way to test this. Do you have any incense?

 

[seekermeister]

Sorry, never use the stuff.

Another type of cooler that I had hoped would be more readily available by now, is the Danamics liquid metal coolers. However, after Googling, it seems that they have not gone fast track in the market yet. I think that is partly due to the fact that they restricted their market to Europe, and it is not available in the USA, which never made much sense to me. The part that I like best, is that it doesn't require any fans. In my case, it is not so much of a noise issue, as it is space. It appears that it would sit along side my chipset HS/F without any problems.

From what I have found with regular air coolers, the only ones that appear would fit the space better are those shaped in a circular array, which seems would allow more space from the other HS/F to fit under it.

 

[sygnus21]

I considered it, but since I'm a retired pilot, with a good knowledge of aerodynamics, which relates to this issue better than thermo dynamics, I do not tend to feel that the idea will work. Yet, I have not discarded it, simply put it on hold for the moment, and will experiment with it some later. For thermo dynamics to work in the fashion needed in this case, it would require sufficient airflow through the cooler. I do not believe that excessive airflow would hinder cooling, if anything, it would improve it. The Bernoulli Principle states that when a liquid or gas is sped up, that the pressure is decreased, and in turn, the temperature decreased. The only way that would not necessarily be true is if the mass is sped up to near supersonic speeds, where a heating due to compression comes into play.

The only thing that I can see where airspeed could play a part in the problem, is if somehow, in a fashion that I do not understand, the airflow is being obstructed or detoured to a path other than intended.

Ouch..... that made my head explode

Oh the science of cooling a PC

Good luck.