To water or to Peltier, or both that tis the question

[Papsomax]

I've as much exhausted the air-cool methods, since my goal is to get my XP 2500 oc'd stable at 200x12.5 (the best now is 12.5x187 with air 38/43 at 1.85 cVolts, I know I need to go extreme. I am leaning towards the peltier way. What do you guys think. H20 vs pelt or both?

Out here,
Paps

 

[specific]

people use a peltier without water?

such a waste of time

what board? ram?

Getting a 2500 to 2.5 at 200+ stable w/ cpu interface enabled is luck of the draw, especially with anything less than a good water setup.

 

[Papsomax]

luck as it shall be deemed

My mobo is an MSI KT6 Delta with SATA. I am running 512 OCZ PC3200 at 2-2-2-6, with Radeon 9500 Pro. At 190x12.5 I cannot post and even at 187x12.5 with most voltages pushing max it is about 90% stable. 100% stable with all aps at 184x13.5. Yes I have read reviews of pelts alone with an external power source keeping the temps in the 10-20 range.

Out here,
Paps

 

[hafa]

What are your temps at 187x12.5? Take a look at my sig; my temps run around 37 at 31 ambient. System is 100% stable. The moral of the story is that you don't need to have temps in the 10-20 range to have a good O/C. As specific said, part of it is luck of the draw.

Before going extreme, you may want to haunt the MSI forums at http://www.ocforums.com/forumdisplay.php?s=&forumid=46 and check your PSU voltages to be sure they're in spec. You may even want to try the higher FSB with different memory settings. Experimentation is the heart and soul of overclocking...

 

[Ven0m]

Possible setups:
1. water
2. water + peltier (unrecommended)
3. phase change

Air cooled paltier is not an option.
1. the easiest - you get better temps than with air cooling, and rig can be really silent if made towards "stealth technology"
2. as above + pelt - uses lots of power, pelt produces lots of heat - higher bills and needs more efficient cooling (water <-> air). I don't recommend it because of high costs of using, because pelt needs some power supply (additional cost) and it's harder to have running really silent.
3. Decent temps and crazy (good) efficientcy ratio comparing to pelt systems. Bad sides - starting cost, uses more energy than water cooling (but far less than peltiers), weight, needs some experience or more money if you want to buy ready rig (prometeia).

Depending on your money, free time and how crazy you are, I'd choose option 1. or 2.

 

[Joe Camel]

seems many of us are going to the extreme. (see my post: AMD diy kit) in the extreme section. if i didnt have the H2O setup already, i would have thought more about going prometeia. (damn they cost alot new)
IMHO: water and water+pelt is more about the fun of creating your own setup... yes you will get a beter OC but your talking 100's not 1000's of MHz like you MIGHT get going phase change.

personally, i went to just water first (skipped HIGH end air) and have been very happy with it for months....now i want pelt too. and if the pattern holds true, ill have a Promy in 6-8 months LOL.

good luck

 

[CCW]

To be honest, I personally dont think Thermo-Electric cooling is worth it. For less money, you could run a chilled water cooling system .this, like TECs would need condensation protection as you will going below ambient room temperatures but it will save money and should cool just as well if not better. You can buy a proper chiller or just use a bucket filled with ice blocks, the only thing about the seocn doption is its not designed for 24hr use, only benchmarking sessions so for 24h use youd need to buy a chiller.

Craig