Whooaaa there cowboy!!!!

[stress]

ooops a little bit too much compound there

 

[Ugmore Baggage]

My "dream shim" would have a picture of flames radiating from the center. Just to freak me out every time I had to remove it. "Ahh I burned it! oh sheesh"

 

[fletch]

You could have diamond heatsinks - best thermal conductor known to man

Although I think the price would be just a *bit* much

 

[I.M.O.G.]

That picture is riddled with problems, only the most obvious of which is the thermal compound application.

1. Thermal compound: It's sloppy and uses a bit too much compound obviously. This doesn't need explained.
   
2. Thermal probe: The placement of the thermal probe, at best, plain sucks.
   
   I attached an enlarged image of the probe placement - the end of the probe is not trimmed, and the wire that actually observes temperatures is a good distance away from the die. If you are going to use a probe like this, atleast trim the probe.
   
   However, even if the probe were trimmed, the accuracy of the readings are still junk - the probe doesn't just read temperatures from the small fraction of the probe touching the CPU die, it reads temperatures from the entire exposed end of the probe. Observed temperatures will be a composite reading of the air, thermal paste, CPU substrate, and the small portion of the probe that actually comes close to touching the die.
   
   Making an assumption that the die temperature is consistent from it's edge, through it's center, to the other edge, is also ridiculous... The temperature of the die can vary greatly from one area to another.
   
3. Shim: Don't buy one, they can hinder contact and there is a better way of making your own.
   
   

   There is a really cool way of making a shim like support that bonds to your HS but not the chip and fits EXACTLY right. I will see if I can find the info. It involves using JBweld I believe. BRB.
   
   Buying a ready made shim for safe insurance is not a great idea, and there is a much better alternative.
   
   EDIT: I don't think I'm going to find the guide to making your own shim, but it basically goes like this:
   
   1) Cover the CPU substrate with a thin sheet of paper.
   
   2) Put a small glob of JBWeld near the corners of the paper on the CPU substrate.
   
   3) Intall the Heatsink and let the JBWeld harden.
   
   4) Remove your heatsink with the new JBWeld "feet" that are custom made for your CPU. Remove the paper from the feet and/or the CPU and throw it away.
   
   5) That's it. You've got a shim substitute which gives you all of the insurance of a shim, but with none of the possible risk of a shim not fitting your CPU correctly.

   
   These are my comments in shims:
   
   

   A few statements about shims. A shim will NOT lower your temps. A shim is cheap insurance against cracking your core. All shims are NOT the correct height for your core, so they can restrict contact.
   
   So it is safe to not use a shim, and I wouldn't ever use one, but you may reduce your risk by using one. Shims cannot actually improve your temps, but if they are not the right size, they can make your temps worse by interfereing with core contact.
   
   Here's a little extra material that I gathered by request:
   
   I went ahead and gathered the most relevant quotes from these links and wrote them out for you. These links and excerpts illustrate some of the issues associated with shims:
   
   First, Let's see what's behind door number 1:
   http://www.overclockers.com/articles313/
   
   

   • This was the tell-all test for me - the use of the copper spacer on the Slot 1 package yielded no usable gains and in the case of the PGA package, it was a disaster. Not only was the effective cooling of the CPU core substantially reduced, the remarkable cooling of the back of the motherboard would surely result in major long term condensation problems.

   Door number 2:
   http://www.overclockers.com/tips278/
   
   

   • SUMMARY: Adding a copper CPU spacer to protect a fragile AMD CPU makes a lot of sense. However, even though it does transfer heat from the CPU package to the heatsink - it looks like there is little or no net CPU cooling benefit.
   • Anyhow, first thing I did was to "eyeball" the fit - looked like it was a little lower than the CPU core on my Duron 800, so I proceeded to grease it up with a non-silcon grease and slap it on the CPU.
   • Intuitively, you would think adding a copper spacer of the kind pictured above would add to CPU cooling. Increase the cooling area of the heatsink to the CPU package and it should run cooler. Anything that can help to tame the flaming AMD CPUs has to be a good move.
   • What I did not expect to see were CPU core temps HIGHER (as measured by the thermistor in the heatsink above the CPU core).
   • The moral of all this: Save your money and don't buy a CPU spacer expecting a temperature benefit. If you want additional cooling, aim some fans at the back of the motherboard. Treat the motherboard like another heatsink and cool it accordingly.

   Door number 3:
   http://www.overclockers.com/articles665/
   
   

   • The shim does not fit the T-Breds and requires modding to get it to fit correctly, but I'll get to that in just a second.
   • The shims provided do not fit the T-Bred CPUs correctly and need cutting to get them to fit. As you can see with the picture below, I had to mod the shim used or it would conflict with the components on top of the CPU:

   Door number 4:
   http://www.overclockers.com/articles410/
   
   

   • SUMMARY: If you're concerned about CPU cracking, this is a good way to protect your CPU's core - don't expect any cooling benefits.
   • Steve 'Supercoolin' Foster did a fair amount of research on what copper CPU shims do to CPU cooling and concluded that it does not help and in fact, it is a detriment. This is due to the additional loading of secondary cooling on the heatsink and the disruption of the secondary heat path. Copper shims transfer heat efficiently, while "The Cool-Shim does not assist and/or help in additional cooling."
   • CONCLUSION
     CPU shims are a great idea if you remove your heatsink often and you have one that is a bear to take on and off. The Cooler Guys shim is very precise in its width and is as non-conductive as a rock, so there's no possibility of shorting out the CPU. Expect some minimal CPU temp degradation which may or may not be obvious with in-socket thermistors.

   Door number 5:
   http://www.amdmb.com/article-display.php?ArticleID=76
   
   

   • Everyone has heard horror stories about chipped or cracked CPU cores when either installing or removing an HSF. Enter on to the scene, spacers or shims whichever name you prefer.
     
     There has been an ongoing argument as to whether shims/spacers aid in cooling; there are proponents on both sides of the fence. The writer has found NO advantage or disadvantage relating to cooling when using one.

   Door number 6:
   http://www.amdmb.com/wiki/index.php?full=The Basics of Overclocking
   
   

   • Something else that can be a topic of debate is shims. A shim is a product made of copper or aluminum which can sometimes be anodized. A shim is used to sit on top of your CPU and take up the space between the silicon base of your CPU and the Heatsink. It has helped users to make sure not to chip off any corners of the CPU core (the small silver square in the center of your CPU). However, there can be two drawbacks to these products. Copper, or any metal for that matter, conducts electricity. If the shim is not properly seated on the CPU it may touch one of the components mounted on the top of the CPU and short it out. Using an anodized version of a shim lowers the chances of this happening, but does not rule out the possibility of it happening. The other drawback is metal conducts heat, just like you Heatsink. The only portion on the CPU that you want to conduct heat away from is the core. But with a shim, some of that heat may be getting transferred back to the CPU through the use of a shim.

   Door number 99874983423458392813487374:
   http://www.amdmb.com/article-display.php?ArticleID=94&PageID=5
   
   

   • This brings us to a product who’s value as an aid in heat reduction has led to much heated (pun intended) debate with champions on both sides of the question. It is my view, after years of experience that shims offer little to no advantage, when it comes to heat dispersal.

 

[eobard]

[*]Thermal probe: The placement of the thermal probe, at best, plain sucks.

I attached an enlarged image of the probe placement - the end of the probe is not trimmed, and the wire that actually observes temperatures is a good distance away from the die. If you are going to use a probe like this, atleast trim the probe.

However, even if the probe were trimmed, the accuracy of the readings are still junk - the probe doesn't just read temperatures from the small fraction of the probe touching the CPU die, it reads temperatures from the entire exposed end of the probe. Observed temperatures will be a composite reading of the air, thermal paste, CPU substrate, and the small portion of the probe that actually comes close to touching the die.

Making an assumption that the die temperature is consistent from it's edge, through it's center, to the other edge, is also ridiculous... The temperature of the die can vary greatly from one area to another.

Ture that! The probe on the backside of my CPU gives numbers about 15-17°c above ambient full load. When I had the probe trimmed down and touching the side of the core I was getting load numbers about 5°c above ambient, [Amish guy on Simpsons] tis a fine dream but sure tis no reality English [/Amish guy on Simpsons].

 

[Radical]

That's kind of what my first TIM application looked like.

 

[squeakygeek]

You could have diamond heatsinks - best thermal conductor known to man

Although I think the price would be just a *bit* much

Aren't heat pipes better?

http://forum.oc-forums.com/vb/showthread.php?threadid=250286&goto=newpost

 

[I.M.O.G.]

I don't see what either of those things have to do with this thread - we weren't talking about heatsinks, and we weren't talking about anything that has to do with heat pipes... So I won't mention Carbon Nanotubes.

 

[Hookem]

What are my real temp's??? Who know's???

Ture that! The probe on the backside of my CPU gives numbers about 15-17°c above ambient full load. When I had the probe trimmed down and touching the side of the core I was getting load numbers about 5°c above ambient, [Amish guy on Simpsons] tis a fine dream but sure tis no reality English [/Amish guy on Simpsons].

However, even if the probe were trimmed, the accuracy of the readings are still junk - the probe doesn't just read temperatures from the small fraction of the probe touching the CPU die, it reads temperatures from the entire exposed end of the probe. Observed temperatures will be a composite reading of the air, thermal paste, CPU substrate, and the small portion of the probe that actually comes close to touching the die.

What are my real temp's??? Who know's???

 

[squeakygeek]

huh?

 

[I.M.O.G.]

I feel like I'm playing the guessing game - what are you saying with those quotes hookem?

 

[Silversinksam]

Shims stink and I still dont understand this product but found a pic without the 12 pounds of thermal compound on it



http://www.pc-mac2000.de/shop_fr/fr/dept_141.html


XMPT Universal AMD CPU shim Euro7,90



XMPT Universal AMD CPU shim
Anodized aluminum
Non conductive
Non absorbent material
Superior heat transfer vs. other insulating materials used in the market place.
Cut out slot for on DIE thermistor sensor temperature readings
Accurate precision material height rolled under 50,000 lbs of pressure for uniformity.
Easy access to L1 bridges
Non interference with surface mounted resistors on older and later AMD CPUs

CPUFX Spacer Universel

 

[Hookem]

I feel like I'm playing the guessing game - what are you saying with those quotes hookem?

I guess what I am saying is, who really knows what their temperatures are???

Do you know of a way that you can get absolutely correct temperatures???

 

[Korndog]

I guess what I am saying is, who really knows what their temperatures are???

Do you know of a way that you can get absolutely correct temperatures???

nothing is really 'absolute'
Thats why u always have to use the ~

 

[eobard]

Maybe whoever had the chip in the first pic was trying to sneak thermal compound across the border by hiding it on top of CPUs. (one pound at a time)

 

[stan03]

LOL

 

[I.M.O.G.]

I guess what I am saying is, who really knows what their temperatures are???

Do you know of a way that you can get absolutely correct temperatures???

Nope, thats just the thing, temperature readings give roughly relative information.

There are documents explaining how internal on-die sensor readings are considerably affected by their placement within the die even.

The only temperatures that mean anything to me are 70C, 35C, and -10C... beyond that our readings don't mean anything IMO.