Overclock.net banner

1 - 9 of 9 Posts

·
Premium Member
Joined
·
4,596 Posts
Discussion Starter · #1 ·
Hi All,

Haven't posted on the forum in quite some time, but I'm running into a few challenges for a work related project and thought maybe you guys could help out.

We are using a TEC to cool a CCD camera chip to -25C with air cooling (alum barrel as heatsink and fans). Specifically, we are applying Arctic Ceramique to the cold side of the TEC (CCD side) and AS5 to the hot side (heatsink plates). Here are some questions I'm facing:

1) Is this combination viable? I know that AS5 hardens and flakes at temps below 0C and hence we put it on the hot side, and it is more efficient at temps above 0 than Ceramique. For the cold side, we're using Ceramique since it is cheaper, hardens, lasts longer without drying out, and cleaner to apply. What if we used the same paste for both sides of the TEC?

2) Is burn-in time necessary? I read that Ceramique doesn't need a burn-in time to reach optimal temps, but AS5 does. The website suggests a 200 hour break-in period to reach a 2-5C temp drop, but from my experience in the past only 1-2 days is needed for the biggest impact. Because we want to assemble these products as fast as possible, the burn-in time vs temp output might not be worth it. Originally I was thinking of a 20 minute break in time, but don't know which temperature intervals I should fluctuate to get best results - i.e. for CPU applications, it is recommended to turn on/off the computer in intervals until 200 hours with stress tests in between.

3) OCN member "tuchan" wrote (back in May 2006 from a thread):
Quote:
When applying the paste with peltier use I have also found that using a fair amount seems to be better than less. This is for the "TEC to plate" cold side.
I have done much experimentation with different degrees of torque applied to both the TEC and pressure on the CPU. All with varying success. As long as the CPU has good and flat contact with the cold plate it seemed to be good for me. Too tight a contact and it would seem to run a lot hotter, to the tune of 10C. Firstly I just placed the block\\TEC assembly on the CPU with the computer on 100% load and just fiddled with the hold down pressure making sure the block is all sitting flat on the CPU. As many water coolers might know if you screw the block on too tight you loose cooling capacity for some reason, the same applies here
Is this true? Do I need to apply more AS5 and Ceramique when using with TEC rather than a regular heatsink? Does pressure really make that much of a temperature difference?

Thanks for your help everyone.
 

·
Premium Member
Joined
·
4,018 Posts
Quote:
Originally Posted by wudaddy View Post

Hi All,

Haven't posted on the forum in quite some time, but I'm running into a few challenges for a work related project and thought maybe you guys could help out.

We are using a TEC to cool a CCD camera chip to -25C with air cooling (alum barrel as heatsink and fans). Specifically, we are applying Arctic Ceramique to the cold side of the TEC (CCD side) and AS5 to the hot side (heatsink plates). Here are some questions I'm facing:

1) Is this combination viable? I know that AS5 hardens and flakes at temps below 0C and hence we put it on the hot side, and it is more efficient at temps above 0 than Ceramique. For the cold side, we're using Ceramique since it is cheaper, hardens, lasts longer without drying out, and cleaner to apply. What if we used the same paste for both sides of the TEC?

2) Is burn-in time necessary? I read that Ceramique doesn't need a burn-in time to reach optimal temps, but AS5 does. The website suggests a 200 hour break-in period to reach a 2-5C temp drop, but from my experience in the past only 1-2 days is needed for the biggest impact. Because we want to assemble these products as fast as possible, the burn-in time vs temp output might not be worth it. Originally I was thinking of a 20 minute break in time, but don't know which temperature intervals I should fluctuate to get best results - i.e. for CPU applications, it is recommended to turn on/off the computer in intervals until 200 hours with stress tests in between.

3) OCN member "tuchan" wrote (back in May 2006 from a thread):
Is this true? Do I need to apply more AS5 and Ceramique when using with TEC rather than a regular heatsink? Does pressure really make that much of a temperature difference?

Thanks for your help everyone.
@Puck You're expertise is needed.

More is needed for the increased surface area. With a CPU heatsink only the few square CM in the center of the IHS are generating heat. With a TEC the ENTIRE TEC is generating heat, you need proper coverage of the entire TEC.

-Z
 

·
TEC Veteran
Joined
·
2,007 Posts
I would just use Ceramique2 on both sides - like you said, its cheap and loves low temps. I used it for a while on my setup until switching the IHS side to Thermene, which performs better then the Ceramique 2 and also resists shrinking (at least down to 0c, havent tested below that). Keep in mind that there is also an older original Ceramique formula that performs much worse then the newer 2, which will be clearly labeled on the tube. I know some places still have ancient tubes of the original Ceramique floating around.

Never noticed any burn in with Ceramique, but yes the AS5 will drop a few *C after heat cycling and running for a few days.

It is expensive and probably overkill, but CLP works great for TECs as well since it freezes into a solid at +8c for a nearly perfect bond, and reflows once it gets above that so it can be removed.
Quote:
Originally Posted by ZytheEKS View Post

@Puck You're expertise is needed.

More is needed for the increased surface area. With a CPU heatsink only the few square CM in the center of the IHS are generating heat. With a TEC the ENTIRE TEC is generating heat, you need proper coverage of the entire TEC.

-Z
Wow jinx, I was replying at the same time lol!
 

·
Premium Member
Joined
·
4,018 Posts
Quote:
Originally Posted by Puck View Post

I would just use Ceramique2 on both sides - like you said, its cheap and loves low temps. I used it for a while on my setup until switching the IHS side to Thermene, which performs better then the Ceramique 2 and also resists shrinking (at least down to 0c, havent tested below that). Keep in mind that there is also an older original Ceramique formula that performs much worse then the newer 2, which will be clearly labeled on the tube. I know some places still have ancient tubes of the original Ceramique floating around.

Never noticed any burn in with Ceramique, but yes the AS5 will drop a few *C after heat cycling and running for a few days.

It is expensive and probably overkill, but CLP works great for TECs as well since it freezes into a solid at +8c for a nearly perfect bond, and reflows once it gets above that so it can be removed.
Wow jinx, I was replying at the same time lol!
ph34r-smiley.gif


-Z
 

·
Registered
Joined
·
12 Posts
Quote:
Originally Posted by Puck View Post

It is expensive and probably overkill, but CLP works great for TECs as well since it freezes into a solid at +8c for a nearly perfect bond, and reflows once it gets above that so it can be removed.
Wow jinx, I was replying at the same time lol!
When you say CLP I presume your mean Coollaboratory Liquid Pro? I've was worried being Gallium based and thus loving to eat aluminium, it wouldn't be safe to use of peltiers, as most use alumina (aluminium oxide) ceramic plates.

I plan to use Coollaboratory Liquid ultra, between the naked die and cold plate, but was undecided on what to use on each size of the peltier itself, I normally use Artic Silver 5 for everything, but as said here It doesn't appear to work well at very low temperatures.

I want to use what ever will give the best performance, but am worried what effect Liquid Pro or Liquid Ultra would have on peltier.
 

·
Premium Member
Joined
·
4,018 Posts
Quote:
Originally Posted by TAZ-NZ View Post

When you say CLP I presume your mean Coollaboratory Liquid Pro? I've was worried being Gallium based and thus loving to eat aluminium, it wouldn't be safe to use of peltiers, as most use alumina (aluminium oxide) ceramic plates.

I plan to use Coollaboratory Liquid ultra, between the naked die and cold plate, but was undecided on what to use on each size of the peltier itself, I normally use Artic Silver 5 for everything, but as said here It doesn't appear to work well at very low temperatures.

I want to use what ever will give the best performance, but am worried what effect Liquid Pro or Liquid Ultra would have of peltier.
Gallium eats aluminum because they're dissimilar metals. Aluminum oxide is not a metal, and is also an electrical insulator so it can't react with gallium. Maybe I'm missing something there, so if anybody wants to correct me feel free.

-Z
 

·
Premium Member
Joined
·
4,596 Posts
Discussion Starter · #8 ·
Quote:
Originally Posted by Puck View Post

I would just use Ceramique2 on both sides - like you said, its cheap and loves low temps. I used it for a while on my setup until switching the IHS side to Thermene, which performs better then the Ceramique 2 and also resists shrinking (at least down to 0c, havent tested below that). Keep in mind that there is also an older original Ceramique formula that performs much worse then the newer 2, which will be clearly labeled on the tube. I know some places still have ancient tubes of the original Ceramique floating around.

Never noticed any burn in with Ceramique, but yes the AS5 will drop a few *C after heat cycling and running for a few days.

It is expensive and probably overkill, but CLP works great for TECs as well since it freezes into a solid at +8c for a nearly perfect bond, and reflows once it gets above that so it can be removed.
Wow jinx, I was replying at the same time lol!
Thanks Puck. I think I'm going to apply Ceramique to both sides instead of using AS5 on one side, and just skip the burn-in phase altogether due to time constraints on our builds.
 

·
TEC Veteran
Joined
·
2,007 Posts
Quote:
Originally Posted by TAZ-NZ View Post

When you say CLP I presume your mean Coollaboratory Liquid Pro? I've was worried being Gallium based and thus loving to eat aluminium, it wouldn't be safe to use of peltiers, as most use alumina (aluminium oxide) ceramic plates.

I plan to use Coollaboratory Liquid ultra, between the naked die and cold plate, but was undecided on what to use on each size of the peltier itself, I normally use Artic Silver 5 for everything, but as said here It doesn't appear to work well at very low temperatures.

I want to use what ever will give the best performance, but am worried what effect Liquid Pro or Liquid Ultra would have on peltier.
Oops, probably should have clarified that I meant that it works great in a TEC setup in general because of its 8c freezing point, did not mean to infer to use it on the TEC itself. Ceramique2 is what is recommended for that.

I use it under the IHS as well, and haven't tried to put it on a whole TEC...I imagine that would be expensive and messy, with too much room for error. CLP/U really needs perfectly flat surfaces, careful application, and even mounting - I would be worried about it moving around too much while torquing down something with as much surface area as a TEC.

I doubt it would actually damage it since most TECs are ceramic and insulated along the edges, I would just be worried about getting a good application.
 
1 - 9 of 9 Posts
Top