It is currently Fri Aug 22, 2014 8:42 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 49 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: DIAMOND-based heat transfer compound
PostPosted: Thu Apr 16, 2009 8:58 am 
Offline
Site Admin

Joined: Sun Aug 11, 2002 3:26 pm
Posts: 11828
Location: Vancouver, BC, Canada
DIAMOND-based heat transfer compound

_________________
Mike Chin,
Editor/Publisher, SPCR
Support SPCR with your donations!


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 9:19 am 
Offline

Joined: Sun Apr 20, 2008 3:27 am
Posts: 91
Location: Sheffield, UK
Last time I bought some thermal paste I bought the following:

http://www.innovationcooling.com/ Innovation Cooling Diamond 7 carat

It dropped my temps by about 2*C compared with Artic Silver 5.

That was about 6 months ago, so that diamond thermal compound isnt anything new!

_________________
Q9550 @ 3.6Ghz (1.15v) | Gigabyte X38 DS4 | 4Gb DDR2 800mhz 4-4-4-12 | 8800GT OC 512mb
WD SE16 640Gb
Enermax Modu82+ 525w | Antec P182 | Xigmatek HDT-S1283 with Noctua P12 | Noctua S12 800rpm rear and front


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 9:35 am 
Offline
Site Admin

Joined: Sun Aug 11, 2002 3:26 pm
Posts: 11828
Location: Vancouver, BC, Canada
It was new to me... and thank you for the link. We've ignored the raging debates about TIM for a long time as the variances and margins of error in testing seemed to be at least as big as that claimed by various makes... but perhaps these diamond-based ones are really better. We'll try them out, then post a report in a while.

_________________
Mike Chin,
Editor/Publisher, SPCR
Support SPCR with your donations!


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 9:42 am 
Offline

Joined: Wed Apr 04, 2007 12:25 pm
Posts: 72
Location: USA
Quote:
...the penultimate material for thermal conductivity has been turned into thermal paste...

So, it's the next to best material for thermal conductivity? What's the best? :?

_________________
"One day, I didn't have to lift a finger, and my wife was thrilled."


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 9:53 am 
Offline

Joined: Sun Apr 20, 2008 3:27 am
Posts: 91
Location: Sheffield, UK
MikeC wrote:
It was new to me... and thank you for the link. We've ignored the raging debates about TIM for a long time as the variances and margins of error in testing seemed to be at least as big as that claimed by various makes... but perhaps these diamond-based ones are really better. We'll try them out, then post a report in a while.


To be honest, I agree mostly, when it comes to TIM then provided you buy one of the better ones I don't think it really matters all that much.

I believe there probably is a worthwhile difference between "Generic TIM 1" and say "Artic Silver 5" or one of these diamond ones. But comparing AS5 with AS3 or ceramique or diamond or whatever, you can get enough variance just with slightly different application methods!

What has a bigger impact, and is widely known by the overclocking communities, is lapping your heatsink and CPU's IHS, to provide completely flat even surfaces to mate with TIM.

_________________
Q9550 @ 3.6Ghz (1.15v) | Gigabyte X38 DS4 | 4Gb DDR2 800mhz 4-4-4-12 | 8800GT OC 512mb
WD SE16 640Gb
Enermax Modu82+ 525w | Antec P182 | Xigmatek HDT-S1283 with Noctua P12 | Noctua S12 800rpm rear and front


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 10:08 am 
Offline
Friend of SPCR

Joined: Tue Oct 05, 2004 1:59 am
Posts: 1630
Location: Portugal
Capsaicin wrote:
Quote:
...the penultimate material for thermal conductivity has been turned into thermal paste...

So, it's the next to best material for thermal conductivity? What's the best? :?



Uhm... it IS diamond. The second best is silver.

Diamond has 5 times better thermal conductivity than silver and 6 times more than copper. Just imagine a diamond heatsink :lol:

_________________
Fractal Define R4 | Corsair AX750 | MSI Z97 Gaming 5 | Intel Core i7 4770K w/TRUE 120 Rev. C | 16GB G.Skill Sniper DDR3-1866 | Sapphire HD 7970 3GB OC w/Arctic Accelero Xtreme IV 280(X) | Crucial m4 128GB + Crucial MX100 512GB + WD Caviar Blue 1TB + WD Caviar Red 4TB | JVC HA-RX900 | Dell U2312HM + BenQ G2400WD | SteelSeries 7G | Logitech G400s | SteelSeries QcK+


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 10:20 am 
Offline

Joined: Sat Mar 17, 2007 3:17 am
Posts: 780
Location: Croatia
Capsaicin wrote:
Quote:
...the penultimate material for thermal conductivity has been turned into thermal paste...

So, it's the next to best material for thermal conductivity? What's the best? :?

Toothpaste! http://www.dansdata.com/goop.htm


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 10:32 am 
Offline

Joined: Mon Jan 12, 2009 4:06 pm
Posts: 27
http://benchmarkreviews.com/index.php?o ... 8&Itemid=1

This is my favorite Thermal Paste article -- it is a 33 way comparison of all types of thermal paste. It has "Innovative Cooling 7 Carat Diamond" Thermal paste in it --- not sure if it actually has diamonds or not.


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 10:38 am 
Offline

Joined: Mon Apr 25, 2005 10:00 am
Posts: 2131
Location: Grand Rapids, MI
rpsgc wrote:
Capsaicin wrote:
Quote:
...the penultimate material for thermal conductivity has been turned into thermal paste...

So, it's the next to best material for thermal conductivity? What's the best? :?



Uhm... it IS diamond. The second best is silver.

Diamond has 5 times better thermal conductivity than silver and 6 times more than copper. Just imagine a diamond heatsink :lol:


Actually, it's Graphene which is at a minimum over twice as conductive as diamond. However it's a little hard to make a atom-thin layer of carbon into a paste...

_________________
HTPC: OrigenAE X11|Gigabyte GA-MA785GPMT-UD2H|Phenom II x3 740BE w/AC Freezer 7|150GB Velociraptor|Corsair VX450
Main: Antec 300 (SlipStream @ 800rpm/140mm @ 5v)|Asus M4A88TD-M|Phenom II x4 945 (Mugen2 pass.)|Asus EAH6850|Samsung 830 128GB|Antec TP750
WHS: DF-85|P8H67-M Pro|I5-3450S/Hyper 212+|Corsair AX650|Sandisk Extreme 240GB, 2xWD20EARS, 2x WD15EARS, WD15EADS


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 11:23 am 
Offline

Joined: Tue Dec 13, 2005 1:08 pm
Posts: 1407
Location: Michigan
jhhoffma wrote:
Actually, it's Graphene which is at a minimum over twice as conductive as diamond. However it's a little hard to make a atom-thin layer of carbon into a paste...

I would take it in atom thick sheets. :)


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 12:36 pm 
Offline
Friend of SPCR

Joined: Wed Dec 17, 2003 6:14 am
Posts: 1235
Location: Finland
If you were to use this sheet as a thermal pad, then your cooler and cpu mating surfaces would need to be flat on the atomic level. Not really feasible. The sheet would also probably break when installing the cooler, even though graphene is one of the strongest materials in the world.

_________________
my general purpose desktop system: HP LP3065 30" LCD 2560x1600 pixels, Q9450, 8GB DDR2 ECC 800MHz, EVGA 460GTX 1GB SC (OC@800MHz/2000MHz), WD Velociraptor 300GB, Samsung 2TB, Gigabyte EX38-DS4, Antec P182b, Corsair VX450, Thermalright Ultra 120 Extreme, Scythe S-Flex fan, <90W AC idle, 200W AC gaming


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 12:46 pm 
Offline

Joined: Mon Mar 17, 2008 1:49 pm
Posts: 1192
Location: UK
xafier wrote:
What has a bigger impact, and is widely known by the overclocking communities, is lapping your heatsink and CPU's IHS, to provide completely flat even surfaces to mate with TIM.

This can be done badly though. Noticed this recently:
Quote:
Lapping Warning!
Prolimatech does not condone any type of lapping done to the CPU or to heatsink base. Every Prolimatech's heatsink base is designed on a pin-point scale of how the base is to be flat and/or curved where it's needed to be. We have programed our machines to machine the surface in a very calculated way. Any after-manufacture lapping or modding done to the base will alter the design, hence negating its performance factor as well as its warranty.

http://www.prolimatech.com/index.html

The manufacturer of one of SPCR's highest rated heatsinks clearly doesn't like people messing them up.

_________________
Silverstone SG03B: E8200, Asus P5E-VM HDMI, 2Gb RAM, Leadtek 9600GT+S1 rev. 2, Samsung 500Gb, Seasonic X-400, 2x Akasa 120mm, Scythe Zipang 2 fanless


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 1:03 pm 
Offline

Joined: Sun Apr 20, 2008 3:27 am
Posts: 91
Location: Sheffield, UK
edh wrote:
This can be done badly though. Noticed this recently:


Yes I realise that, and thats exactly why I personally have never done it.

It really depends on your CPU and your heatsink, some heatsinks are more concave to match the convex shape of most IHS's.

It's really a last resort for those people that really like to overclock to silly levels.

But as I said, as long as you buy an aftermarket TIM as opposed to the sub-standard gunk that ship with most heatsinks, then you're unlikely to see much difference between the market leaders of TIM products.

_________________
Q9550 @ 3.6Ghz (1.15v) | Gigabyte X38 DS4 | 4Gb DDR2 800mhz 4-4-4-12 | 8800GT OC 512mb
WD SE16 640Gb
Enermax Modu82+ 525w | Antec P182 | Xigmatek HDT-S1283 with Noctua P12 | Noctua S12 800rpm rear and front


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 1:40 pm 
Offline
Patron of SPCR

Joined: Thu Jun 29, 2006 8:37 pm
Posts: 376
Location: Mississauga, ON
quick! someone call dan from dan's data!

_________________
..::AMD Opteron 165 2.25GHz 1.25V:.:Scythe Ninja 2:.:ASUS A8R32-MVP Deluxe S939:.:2GB Corsair XMS PC3200 RAM:.:eVGA GeForce 460GTX 1GB GDDR3::..
..::200GB Seagate 7200.8 HDD:.:Plextor PX-755SA DVD±R/RW-DL:.:Seasonic S12II-380 PSU:.:Yate Loon D12SL-12:.:Antec P180 Case::..


Top
 Profile  
 
 Post subject: Re: DIAMOND-based heat transfer compound
PostPosted: Thu Apr 16, 2009 3:22 pm 
Offline
*Lifetime Patron*

Joined: Sat Dec 01, 2007 2:04 am
Posts: 601
Location: UK
MikeC wrote:


I read somewhere recently (link) that Fraunhofer have been touting diamond as a possible TIM component, but then they've been hawking diamond everything these last few months, right down to ball bearings, presumably in an attempt to secure further research cash. So my interest was piqued when this came up, as I'd be surprised if someone had beaten them to the punch with a product on the market.

This is Electrospell's business address. Seems it might be a bit of a garden shed operation...

Image


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 3:40 pm 
Offline
*Lifetime Patron*

Joined: Sat Dec 01, 2007 2:04 am
Posts: 601
Location: UK
Just reading all the stuff on their website, all the home made PDF flyers with random microscope images, I think it's odd that there is this sheen of it being aimed at industry, when the language and applications discussed are clearly aimed at PC enthusiasts. I think I will file this one under snake oil until someone proves me wrong.

I don't doubt there are thermal pastes out there for which some enterprising souls have found a marketable use for industrial diamond dust. I'm only guessing here, but I think there is a world of difference between some stuff that happens to contain diamond, and paste which properly harnesses the thermal properties of diamond.


Last edited by blackworx on Thu Apr 16, 2009 3:46 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 3:42 pm 
Offline

Joined: Sun Oct 09, 2005 7:53 pm
Posts: 81
How much of the thermal resistance between CPU and ambient air is typically accounted for by the TIM? If, for example it is 1%, then reducing this by a factor of three is only going to decrease your total thermal resistance by 0.7%, which is unlikely to be worth while. On the other hand, if it is (say) 10% of thermal resistance, then reducing it would be useful.

It wouldn't be hugely difficult to measure TIM thermal resistance scientifically. Here's one idea: we have various chunks of Big Fat Copper Rod (BFCR). The ends are polished flat, and we have fittings to hold two sections of rod very close to each other. The fittings have poor thermal conductivity. Two of the BFCRs have heatpipes attached. All BFCRs have a temperature probe. The basic experimental setup is to have the heatpipe BFCRs at each end of a 'stack' of BFCRs, with different TIMs at each join. (Typically we'd have just two interfaces - one for the TIM being tested, one for a standard TIM. Other configurations would be useful for callibration - figuring out how much thermal resistance is in the heatpipes and BFCRs.) Then the heatpipes would be put into water baths at two different temperatures. For example: if we have 3 BFCRs (two joins), bath temperatures 0 C and 60 C, and the middle BFCR turns out to stabilize at temperature 40 C, then the join on the hot side has half the resistance of the join on the cold side.
(The BFCRs also need to be insulated against the air, which adds complication.)

Anyhow, I don't think SPCR is up for this level of complication. Here's a simpler set up:
A rod of highish thermal resistance metal sits with base in a ice water bath, and projects some distance above. A temperature probe is inserted just below the top end. The top end is where we apply TIM to attach to a copper block, which is connected by heat pipes to the hot bath. The probe sees thermal resistance between its location and the cold bath due to the poorly conducting metal it is in. It sees resistance between its location and the hot bath due to the TIM, copper block and heatpipes. The better the TIM, the hotter the probe will be - but it will be difficult to callibrate into a degrees-per-watt measurement.


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 4:40 pm 
Offline

Joined: Wed Apr 28, 2004 5:16 am
Posts: 777
Location: canberra, australia
rpsgc wrote:
Capsaicin wrote:
Quote:
...the penultimate material for thermal conductivity has been turned into thermal paste...

So, it's the next to best material for thermal conductivity? What's the best? :?



Uhm... it IS diamond. The second best is silver.

Diamond has 5 times better thermal conductivity than silver and 6 times more than copper. Just imagine a diamond heatsink :lol:


i think he means that, given diamond is the best material for thermal conductivity, and its diamond that has been turned into a thermal paste here, then penultimate is the wrong word to use. the sentence "...the penultimate material for thermal conductivity has been turned into thermal paste..." would seem to suggest that its silver that had been turned into a thermal paste, which of course was done many years ago..


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 5:29 pm 
Offline
*Lifetime Patron*

Joined: Mon May 23, 2005 7:05 am
Posts: 618
Location: State College, PA
now I'm no engineer (phew!) but surely any part in the chain between silicon and air (via IHS>TIM>hs base>heatpipes>fins) is going to be at equilibrium once temperatures have stabilised? i.e., presumably the TIM can only transfer heat as fast as it can be evacuated from the base of the heatsink and at a steady state (idle, full load) that will equilibrate. Otherwise you would see a steady increase in CPU temperature at a steady state, which you don't.
The only difference I can see this making is that it will marginally increase the speed at which that equilibrium is reached - so the CPU will take longer to reach steady temp at full load and less time to cool down. The only things that will reduce CPU temp at steady state is how much energy is going in (CPU heat) and how much is going out (heatsink>air transfer efficiency).

Then again, I'm no engineer :)


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 7:00 pm 
Offline
Site Admin

Joined: Sun Aug 11, 2002 3:26 pm
Posts: 11828
Location: Vancouver, BC, Canada
mattthemuppet --

You're right, but timing in cooling, as in life, is sometimes everything. ;) Uh where do I go after that throwaway quip?! :lol:

If the TIM represents a 10% impedance in the heat transfer of the whole system, then improving it to 0% would actually lower temp in real conditions, where the thermal load does not stay steady, but instead bounces up and down depending on what the user is asking of the computer.

Better TIM simply means that it's less of an impedance or bottleneck. It doesn't means something else isn't even more of a bottleneck and hence you might not see the benefit. For example, with some heatsinks, the slow quiet fan we prefer is probably the biggest bottleneck to performance. With say a Thermalright HR01 and any 120mm fan at 500rpm, surely the difference in TIM will tend to be minimized because the airflow is the bigger bottleneck. With a 125W CPU instead of an 65W CPU, the difference in TIM would show up bigger again.

I think you can safely say that using better TIM ensures that your cooler does the best it can under whatever conditions you give it. The degree of difference may be negligible in may "silent PC" scenarios, however.

_________________
Mike Chin,
Editor/Publisher, SPCR
Support SPCR with your donations!


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 11:17 pm 
Offline

Joined: Tue Nov 04, 2008 6:03 am
Posts: 364
Location: Sweden
I’m a bit sceptic about using diamond dust in a TIM, at least when it comes to the expectations proclaimed.

Diamond is a very good thermal conductor (2.5 to 12.5 times that of silver, depending on exact composition) because of its crystalline properties where heat is easily transferred between the atoms. If you turn it to dust and try to transfer heat through the dust there will be lots of non crystalline contact surfaces where the thermal resistance is considerably higher.
Mixing diamond dust with a liquid to produce TIM will result in a mix where the thermal and mechanical properties of the liquid will be a, if not the, deciding factor in its overall performance.
The claims that this TIM is many times better than “conventional TIMâ€


Top
 Profile  
 
 Post subject:
PostPosted: Thu Apr 16, 2009 11:24 pm 
Offline

Joined: Mon Nov 28, 2005 1:01 am
Posts: 106
Results of the IC diamond paste, average 3.38deg improvement to AS5 based on 125 users.

http://www.ocforums.com/showpost.php?p= ... tcount=433

I've used the stuff, improved my temps by a a couple of degrees. As said by mikec it makes a bigger difference with high power cpus or gpus, hence some of the >10deg improvements.

_________________
Rig: AMD Opteron 146 @ 2.9ghz | DFI NF4 Ultra-D | 2x512mb OCZ Value VX @ 2-2-2-5 242mhz @ 3.3V | X800 PRO @ XT 560/525 | 160+250GB | Zalman 360B-APS | Chaintech av-710 | A-C freezer 64 pro / ATI silencer


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 12:31 am 
Offline

Joined: Sat May 31, 2008 2:33 am
Posts: 124
Location: Australia
Matija wrote:
Capsaicin wrote:
Quote:
...the penultimate material for thermal conductivity has been turned into thermal paste...

So, it's the next to best material for thermal conductivity? What's the best? :?

Toothpaste! http://www.dansdata.com/goop.htm


Cheese is no slotch either:
http://enthusiast.hardocp.com/article.h ... VzaWFzdA==


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 1:49 am 
Offline

Joined: Thu Oct 16, 2008 10:42 am
Posts: 205
Location: Italy
It would be interesting to compare this VS ocz freeze, arctic silver 5 etc (that ocforums.com post is not very trustworthy).


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 3:21 am 
Offline

Joined: Tue Dec 13, 2005 1:08 pm
Posts: 1407
Location: Michigan
mattthemuppet wrote:
now I'm no engineer (phew!) but surely any part in the chain between silicon and air (via IHS>TIM>hs base>heatpipes>fins) is going to be at equilibrium once temperatures have stabilised? i.e., presumably the TIM can only transfer heat as fast as it can be evacuated from the base of the heatsink and at a steady state (idle, full load) that will equilibrate. Otherwise you would see a steady increase in CPU temperature at a steady state, which you don't.

By definition things in stead state don't change, so most of what you just said is meaningless.

The TIM does effect the "steady state" CPU temperature. A higher resistance TIM increases the temperature difference required between the CPU and the heatsink to transfer the same amount of heat energy. This means a hotter CPU and a cooler heatsink. Also since the heatsink isn't the only heatsink you are also increasing the temperature of everything else connected to the CPU (i.e., the motherboard.)

Resistances like TIMs don't effect transient response much. Capacitance effects transient response, which for CPU heatsinks looks something like using water for a coolant or lots of copper mass close to the CPU. Neither will help steady state at all, but both can hold more transient heat.


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 4:34 am 
Offline

Joined: Mon Nov 28, 2005 1:01 am
Posts: 106
ekerazha wrote:
It would be interesting to compare this VS ocz freeze, arctic silver 5 etc (that ocforums.com post is not very trustworthy).


I don't see why not. its the compiled results of users from a number of forums. Having a large number of results reduces potential for error. if i give you an OCAU post for example is it suddenly trustworthy? eg.
http://forums.overclockers.com.au/showt ... p?t=714640

If there's major flaws in methodology that make the results not useful (see OCAU link for details) sure point them out, but i don't think its helpful to dismiss a test purely on association.

_________________
Rig: AMD Opteron 146 @ 2.9ghz | DFI NF4 Ultra-D | 2x512mb OCZ Value VX @ 2-2-2-5 242mhz @ 3.3V | X800 PRO @ XT 560/525 | 160+250GB | Zalman 360B-APS | Chaintech av-710 | A-C freezer 64 pro / ATI silencer


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 6:16 am 
Offline
*Lifetime Patron*

Joined: Sat Oct 25, 2003 3:21 pm
Posts: 1252
Location: 15143, USA
MikeC wrote:
...then improving it to 0% would actually lower temp in real conditions, where the thermal load does not stay steady, but instead bounces up and down depending on what the user is asking of the computer.

MikeC, I'm disappointed to see you write this. Better heat transfer certainly affects steady-state operation, not just transient heating. Take the "no steady-state difference" conclusion to the next level, and you'll believe that the heatsink doesn't matter once temperatures stabilize. Hmmm... Something may be wrong with that idea.

QuietOC's later post is a more accurate description of thermal resistance.


Top
 Profile  
 
 Post subject: what about "liquid metal"?
PostPosted: Fri Apr 17, 2009 6:32 am 
Offline

Joined: Thu Mar 26, 2009 10:56 am
Posts: 5
Location: NYC
No one mentions the stuff called "Liquid Metal"! First, it sounds cool. And I've read that it does a very good job as a TIM. I suspect that its benefits are not really about its conductivity (though that must be important too), but about its ability to make good atomic-scale contact with the surfaces you are trying to connect.

There were some downsides, at least to the original product I read about 6 months ago. It had a tendency to harden over time and eventually bond your cpu to your heatsink. Not so cool. And I seem to recall that some of the original formulations were highly toxic (I'm thinking mercury). But I wasn't planning on drinking the stuff. The links I can find now all claim to be non-toxic, so maybe that has changed (or maybe my original information was incorrect; net is full of nonsense and rumour mongers like myself).

I think there are a couple of competing products, but this is the first one I read about:
http://www.coollaboratory.com/en/product_pro.shtml

Note that I haven't tried it, so I'm not really recommending it. I have tried the diamond stuff (results were inconclusive) and lapping my heatsink (results were very impressive).


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 7:17 am 
Offline
Site Admin

Joined: Sun Aug 11, 2002 3:26 pm
Posts: 11828
Location: Vancouver, BC, Canada
HammerSandwich wrote:
MikeC wrote:
...then improving it to 0% would actually lower temp in real conditions, where the thermal load does not stay steady, but instead bounces up and down depending on what the user is asking of the computer.

MikeC, I'm disappointed to see you write this. Better heat transfer certainly affects steady-state operation, not just transient heating. Take the "no steady-state difference" conclusion to the next level, and you'll believe that the heatsink doesn't matter once temperatures stabilize. Hmmm... Something may be wrong with that idea.

QuietOC's later post is a more accurate description of thermal resistance.

I didn't say better heat transfer does not affect steady-state operation, simply mentioned the more typical/realistic condition for a heatsink: The load varies, it's not steady at maximum load as in stress testing (although it's often more of less steady, as during most operations that call for extended typing like this). I agree that QuietOC's later post is a more accurate description of thermal resistance.

_________________
Mike Chin,
Editor/Publisher, SPCR
Support SPCR with your donations!


Top
 Profile  
 
 Post subject:
PostPosted: Fri Apr 17, 2009 8:32 am 
Offline
*Lifetime Patron*

Joined: Sat Oct 25, 2003 3:21 pm
Posts: 1252
Location: 15143, USA
MikeC wrote:
I didn't say better heat transfer does not affect steady-state operation...
Fair enough. My apologies.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 49 posts ]  Go to page 1, 2  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: Bing [Bot] and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group