new to TEC

[lookmomnobrains]

Hello people from the TEC sub forum lol

I’m thinking of getting started at TEC cooling, and when I get my masters in engineering and a good job jump to phase chance building (don’t think I will be able to pay for phase chance at the moment).

My current system:
CPU: E6600 @ 3.4ghz 1.4V core
CPU cooler: Thermaltake Big Typhoon
VGA: Asus 8800gts
MB: Asus p5b-dlx
MEM: ddr2 800mhz 2gb
PSU: Corsair HX620

Thinking of putting a peltier one the Thermaltake Big Typhoon one 1 side and a copper plate on the other. And put that on the cpu agene. Use some Dielectric Grease and Neoprene to Insulate the mb. And run the peltier from my Corsair HX620.

Questions:
Is the idea any good at all?

What kind of peltier should air-cooling / a Thermaltake Big Typhoon be able to handle?

What kind of peltier can hold the heat load of a ocd c2d to 3.6-4ghz 1.5-1.6vcore?

Am I missing something of what I need / thinking of using?

What kind of temperatures should I think of getting whit this?


Any advice is always welcome, and also does anyone now any suppliers in Holland?

[Cronos]

TEC aircooling is never a good idea.
To cool 100W load you will need at least 200W TEC, which will consume 300W of power. So, you will have to dissipate 400W total with air cooler. You will get better results with plain water cooling.

[Holst]

Is the idea any good at all?

To be honest, no its an awfull idea.

If you read seedomos results using C2D and 220watt pelt you can see that at these speeds the peltier is barely big enough for the heat load.
He also had some problems keeping water temps low enough (untill he switched to two radiators)
With aircooling it just wont work at all, everything will overheat.

I think you would be better off thinking about phase change. As you have studied thermodynamics you should be able to understand the system designs.
Plus if your university have a workshop they can probably help you out with tools and stuff. parts for a single stage phase change are quite chep, the expensive bit is the tools. If you can use the universitys equipment then you can build phase change for the same parts cost as watercooling (using seond hand parts and making some stuff)
Go and talk to the guys at the university workshop and your lecturer in thermodynamics, if its a good university they should be able to give allot of help.

Good luck.

[Mr. Peanut]

I would also like to add that, unless you have a very powerful TEC and have assembled/tuned it well, a very hot processor won't benefit very much from it. If you're OCing like a madman, I would go with a 62mm^2 TEC or none at all. This is because there is a certain point for each TEC setup where better temps are achieved without the TEC. This point occurs from pumping lots of heat into the TEC. Since there is no way you are going to cool a 62mm^2 TEC on air, I wouldn't try this.

[lookmomnobrains]

To be honest, no its an awfull idea.

If you read seedomos results using C2D and 220watt pelt you can see that at these speeds the peltier is barely big enough for the heat load.
He also had some problems keeping water temps low enough (untill he switched to two radiators)
With aircooling it just wont work at all, everything will overheat.

I think you would be better off thinking about phase change. As you have studied thermodynamics you should be able to understand the system designs.
Plus if your university have a workshop they can probably help you out with tools and stuff. parts for a single stage phase change are quite chep, the expensive bit is the tools. If you can use the universitys equipment then you can build phase change for the same parts cost as watercooling (using seond hand parts and making some stuff)
Go and talk to the guys at the university workshop and your lecturer in thermodynamics, if its a good university they should be able to give allot of help.

Good luck.

Was thinking the head load could be problem, but then agene did see some factory made air tecs but couldn’t really find much specs on them and any good reviews from them so I was thinking I will just ask some lol…

I think I’m gone do some phase chance building then but still gone wait some time before I’m gone start whit that hopefully next year in my BEng year … sins in really don’t have to money at the moment :(

[NaeKuh]

okey from the basics i picked up from NOL.

1. TEC's are very inefficient at cooling modern processors.
2. You need a really big tec to get decient temps. NOL said, a 226W wouldnt cut it for a quadcore. After getting mine, i seriously believe him.
3. I think your still assuming a TEC is some magical cooling compoent like a evaporator head.

The hot side on the TEC needs to be cooled REALLY REALLY EFFICIENTLY. I dont think the BT will do that.

So yeah, after NOL was nice to pointing out all my high idealistic noob concepts on TEC, I gave up. :D

Thanks NOL for saving me from spending a TON of money on something completely uselss