EPTIm not saying liquid cooling is great, but it has its uses.
Oh and btw, in summer my ambient temps reach over 40 C quite often. I think last years max was 47 C :D.
Big Cooling Thread - Page 2
| Forum Index > Tech Support |
|
iloveav
Poland1481 Posts
Im not saying liquid cooling is great, but it has its uses. Oh and btw, in summer my ambient temps reach over 40 C quite often. I think last years max was 47 C :D. | ||
|
JingleHell
United States11308 Posts
On March 27 2012 16:50 iloveav wrote: I think you forgot that liquid cooling is also a way to keep your system quiet. At least thats my reason to use it. Im not sure i agree with your statement about the 10 Celcius drop as a max in CPUs. I know its far more in GPUs, as i changed my old ati 4850 cooling twice, first to aftermarket cooler, and i got a 45 C drop (not kidding, anyone who had a 4850 with stock cooler knows how bad it was). and then i got a 13 C drop when i moved to liquid. My Q6600 on stock runs on a kandalf LCS case (modified thermaltake big water kit http://www.overclockersclub.com/reviews/tt_kandalf/) I get 21 C on idle and 31 on Load. I dont know how hot runs q6600 on air, but im quite sure its a bit higher :D. Look again at which section the 10 Celcius drop was in. TIM. I was specifically talking about thermal paste and the benefits of it. And I didn't forget that liquid can keep you quiet, I just considered it outside the purview. It CAN be quieter, but it won't be with closed loop as a rule, since GPU cooling is where it makes the biggest difference, and full custom liquid, as I stated, needs to be researched more completely before using. If you want to write up a quick section on it, feel free, but as I've only studied it, not done it, I didn't feel qualified to say more than I did. | ||
|
TheToast
United States4808 Posts
 One thing though, I would disagree with your opinion on thermal paste/compound. Thermal compound is going to greatly increase the heat trasmission between the CPU and heatsink, reducing any "hot spots" where there may be an air gap between the CPU and heatsink. Given that you can pick up a tube of Artic Silver (a decent name brand) on New Egg for less than $10, there's really no reason at all not to use it. I personally would not run a PC without thermal paste/compound even when not overclocking, you are increasing the risk of damage to your CPU especially if you live in a warm climate or the PC is in a warm place such as next to an air vent. Given that the stuff is so cheap, there's really no downside to using it. As for how to apply it, again I disagree. I've heard all kinds of crazy methods, including putting a big glob on the middle of the CPU and pressing the heatsink into it. (I had a co-worker who argued that this reduced "air bubbles" and was therefore the most effective). However, I'm inclined to disregard the Hardware Secrets website test for two reasons: first in their testing method they did not mention anything about allowing for a break in period. This is key, as after some time the compound changes consistancy and actually becomes more effective; Artic Silver does note this in their documentation. The break in period of the Artic Silver Céramique they used is rated as around 25 hours. Using so much less thermal compound will actually result in a lower break in period, and that could explain their results. Second, every heatsink is different, some have well machined smooth surfaces; others may not. So it's hard to make any sort of blanket statements about the effectivity of a specific setup. Artic Silver actually has their own list of recommendations of which method of application to use for each different type of Intel CPUs. As you can see, for the most part a vertical line or full surface spread is reccommended. There are links on that page to some of Artic Silver's other documentation which has some other interesting information as well, including info on the break in period. However I would point to this list as the best means for applying thermal compound. For myself, I usually default to a full, very thin, surface spread in most situations. A post it note is the perfect size and works well to spread the stuff across the CPU, though an old credit or gift card would be ideal. One thing I will agree with the Hardware Secrets thread on, less is better, no matter which method you use. Overall, thermal compound does help if you apply it correctly; and since most of it is so cheap there's no reason not to. | ||
|
Wabbit
United States1028 Posts
I'm pretty sure JH meant that it doesn't matter much what TIM is used for most people as the difference in temps is going to be minimal as long as it's applied properly - not that no TIM should be used, lol. Yeah there are some trashy TIM's that if you replaced with something like PK-1 or MX-4 you'd maybe get ~5*C or so better temps under load, but usually it'll be less than that, especially when there isn't a ton of heat to dissipate in the first place. I also disregard everything HardwareSecrets does regarding cooling. Probably the best TIM review is this: http://skinneelabs.com/2011-thermal-paste-review-comparison/ Also X-Bit's roundup that includes TIM's used on retail products: http://www.xbitlabs.com/articles/coolers/display/thermal-interface-roundup-1_12.html#sect0 Also nobody should be buying Arctic Silver 5 unless it's $5 for a tube and everything else costs twice as much (almost never the case, other good stuff is pretty inexpensive too) and you need paste and don't care about spending the $5 more for ~3-4 degrees in the right circumstances. | ||
|
JingleHell
United States11308 Posts
After all, if they buy the best, and then glop it on, it's wasted money. I don't usually suggest a pre-spread for application myself, just because there's a bit more technique to it than most people will want to spend the time on, assuming you do it right, and if you do it wrong, you end up with air bubbles. | ||
|
TheRabidDeer
United States3806 Posts
On March 27 2012 16:50 iloveav wrote: I think you forgot that liquid cooling is also a way to keep your system quiet. At least thats my reason to use it. Im not sure i agree with your statement about the 10 Celcius drop as a max in CPUs. I know its far more in GPUs, as i changed my old ati 4850 cooling twice, first to aftermarket cooler, and i got a 45 C drop (not kidding, anyone who had a 4850 with stock cooler knows how bad it was). and then i got a 13 C drop when i moved to liquid. My Q6600 on stock runs on a kandalf LCS case (modified thermaltake big water kit http://www.overclockersclub.com/reviews/tt_kandalf/) I get 21 C on idle and 31 on Load. I dont know how hot runs q6600 on air, but im quite sure its a bit higher :D. Fans can be just as quiet if not more quiet than liquid, especially if you dont run them at 100% all of the time by using fan controllers. | ||
|
JingleHell
United States11308 Posts
On March 28 2012 08:04 TheRabidDeer wrote: Fans can be just as quiet if not more quiet than liquid, especially if you dont run them at 100% all of the time by using fan controllers. Yep. And anyways, GPU coolers tend to be the loudest fans in the system most of the time, and a large enough loop for CPU and GPU, or dual loops, costs a small fortune. Even worse if you're running SLI/CFX. High end liquid can be more efficient, but it's only similar or slightly better if it's not high end. Added Wabbit's links to OP. Keep the resources coming, gents. | ||
|
TheToast
United States4808 Posts
On March 28 2012 07:16 JingleHell wrote: Added the Arctic Silver guide to OP. I didn't actually suggest NOT using thermal paste, I suggested using the pre-applied stuff for most people. Since you agree that not overdoing it is vastly important, then hopefully you understand why I'd suggest people leave the particularly thin layer that tends to be pre-applied. After all, if they buy the best, and then glop it on, it's wasted money. I don't usually suggest a pre-spread for application myself, just because there's a bit more technique to it than most people will want to spend the time on, assuming you do it right, and if you do it wrong, you end up with air bubbles. Ahh, okay I misunderstood then. When you say leave the pre-applied stuff, are you talking about the thermal pads that come installed by default on most pre-built systems? Just as a note, the artic silver documentation actually recommends removing those even before using the PC, as they claim most have a paraffin wax base that, once it melts, gets into the microscopic pores in the metal of the heatsink and CPU and can never be fully removed. Now, granted, they are trying to sell their thermal compound above all else; so you can take that advice with whatever emphasis you like. Also you should probably link Artic Silver's guide to applying thermal compound for AMD's as it's quite different than their recommendations for Intels: http://www.arcticsilver.com/amd_application_method.html# These are their recommendations for applying thermal compound to heatsinks with pipes or ridges: http://www.arcticsilver.com/pdf/hp/hphs_method.pdf (.pdf file) Also might want to add "Break in Period" to the glossary, this is how Artic Silver's instructions explain it: Break-In Period Explained: Due to the unique carrier fluid used and the shapes and sizes of the thermally conductive particles in Arctic Silver's thermal compounds it will take multiple thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to metal cap interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop slightly to significantly 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. Also from the same instructions, this is their list of the break-in time ratings for their various products + Show Spoiler + Break-In Period by Thermal Compound: Arctic Silver 5: Break-in period: 200 hours (Break-in period will occur during normal use.) Temps will drop several degrees over the break-in period measured with a thermal diode in the hottest part of the CPU core. Céramique: Break-in period: 25 hours (Break-in period will occur during normal use.) Temps will drop several degrees over the break-in period measured with a thermal diode in the hottest part of the CPU core. Arctic Alumina: Break-in period: 36 hours (Break-in period will occur during normal use.) Temps will drop several degrees over the break-in period measured with a thermal diode in the hottest part of the CPU core. Matrix: Break-in period: 300 hours (Break-in period will occur during normal use.) Temps will drop several degrees over the break-in period measured with a thermal diode in the hottest part of the CPU core. For what it's worth, I think you should unlink that Hardware Secrets article, it's just plain wrong and has a lot of mis-information. | ||
|
JingleHell
United States11308 Posts
| ||
|
TheToast
United States4808 Posts
On March 28 2012 09:21 JingleHell wrote: Eh, I added a caveat, since it does have some useful info. Also added the AMD TIM link, and the thermal pad advice. I had utterly forgotten that detail. Well it's up to you obviously. I'd actually forgotten about that thermal pad thing too, I do vaguely recall hearing that somewhere before; but I don't think I've ever seen a trusted source with that info before. | ||
| ||
