![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
TEC
- Thread starter Legion-
- Start date
Used to be a big thing back in the day I hear but from what I understand, not so much anymore. Especially since you still have to use another form of cooling to remove the heat from the hot side of the plate, so you might as well remove the middle man (the TEC) and stick with air or go water or go extreme closed loop refrigerant system of some sort.
I know i will need a water setup with the TEC. Ive researched the refrigant systems and decided on going for TEC. But with TEC i still have a few questions about it. And doing research most of the info ive found is dated. Was hoping some here had used TEC.
The heat density of modern CPUs make direct-TEC cooling a huge challenge.
It may work on Ivy-Bridge, but you'd need to use the largest TECs (often rated for about 545w on ebay), and then use them near their rated voltage. You'd then need a custom waterblock with a high clamping force (using bolts) between the cold-plate and the waterblock.
The efficiency will be wretched.
The efficient way to use TECs, which can actually be more efficient than phase-change cooling, is just to chill the water using the same large TECs, but multiple in parallel at low voltage (generally less than 8v). This will limit you to about ~20C below ambient water.
http://www.xtremesystems.org/forums...in-the-process-of-making-a-new-3x50mm-chiller - this guy, UltraSonic2 - has done it all. And he has done it well. If you search through his history, and some of the blocks he has had for sale in the past, you can get an idea of what was possible with past setups.
I bet those guys could tell you what is and isn't worthwhile.
My opinion is that the TEC-chiller would be worthwhile if it were less expensive, and easier to produce. It can certainly be quieter than the phase-change units. I personally wouldn't bother anymore because we now get very limited benefits. It's not like I'm going to get anything close to my TEC-cooled Celeron-A-mendocino at 744mhz. That, and some evaporative coolers can get you close to a sustained 10C below ambient using just a lot of silver and colloidal silver as the biocide (so you're not inhaling garbage - and who doesn't want a silver lining within their respiratory system!).
http://www.xtremesystems.org/forums...l-vs-Krow-All-In-One-Liquid-Chiller-Build-Off! - these guys seem to have some competition going on. I recognize some of the names in that thread, but back when I was researching this stuff, most people were just playing catchup with UltraSonic.
Reposted Edit from http://hardforum.com/showthread.php?t=1719754 :
Edit: I just realized that since Sandy-Bridge, CPU power has decreased a bit. Perhaps the power-density is still high, but the overall heat load has decreased. Direct-TEC cooling is likely possible again: http://www.xtremesystems.org/forums/...)&daysprune=-1 - these guys claim to have made an ~ambient temp solution. I bet one could go with a little more power, and even cool the hot-side with an evaporative cooler (probably a lot of evaporation though) for an additional decrease in temps.
Final recommendation: Go hang out over at the XS forum and absorb as much as you can. Re-assess when you know more. Also consider the "power-density" of the TEC - that is the cooling power per mm^2. The gigantic 545 ebay TEC, which may be listed as maybe 437w from the manufacturer...not sure... may be a 62mm or 50mm square, for instance. I'm sure the designers will take into account both the efficiency at the desired voltage range and the cooling power per unit area.
Also remember that you'll need a dedicated PSU for the TECs. In the past I had actually used a PC PSU. I adjusted the 12V rail up to 13v, and then used 6.5v for an even number of TECs. I used multiple power and ground wires. I bet this isn't recommended practice, but it works on PSU's with easy potentiometer voltage-rail adjustment.
It may work on Ivy-Bridge, but you'd need to use the largest TECs (often rated for about 545w on ebay), and then use them near their rated voltage. You'd then need a custom waterblock with a high clamping force (using bolts) between the cold-plate and the waterblock.
The efficiency will be wretched.
The efficient way to use TECs, which can actually be more efficient than phase-change cooling, is just to chill the water using the same large TECs, but multiple in parallel at low voltage (generally less than 8v). This will limit you to about ~20C below ambient water.
http://www.xtremesystems.org/forums...in-the-process-of-making-a-new-3x50mm-chiller - this guy, UltraSonic2 - has done it all. And he has done it well. If you search through his history, and some of the blocks he has had for sale in the past, you can get an idea of what was possible with past setups.
I bet those guys could tell you what is and isn't worthwhile.
My opinion is that the TEC-chiller would be worthwhile if it were less expensive, and easier to produce. It can certainly be quieter than the phase-change units. I personally wouldn't bother anymore because we now get very limited benefits. It's not like I'm going to get anything close to my TEC-cooled Celeron-A-mendocino at 744mhz. That, and some evaporative coolers can get you close to a sustained 10C below ambient using just a lot of silver and colloidal silver as the biocide (so you're not inhaling garbage - and who doesn't want a silver lining within their respiratory system!).
http://www.xtremesystems.org/forums...l-vs-Krow-All-In-One-Liquid-Chiller-Build-Off! - these guys seem to have some competition going on. I recognize some of the names in that thread, but back when I was researching this stuff, most people were just playing catchup with UltraSonic.
Reposted Edit from http://hardforum.com/showthread.php?t=1719754 :
Edit: I just realized that since Sandy-Bridge, CPU power has decreased a bit. Perhaps the power-density is still high, but the overall heat load has decreased. Direct-TEC cooling is likely possible again: http://www.xtremesystems.org/forums/...)&daysprune=-1 - these guys claim to have made an ~ambient temp solution. I bet one could go with a little more power, and even cool the hot-side with an evaporative cooler (probably a lot of evaporation though) for an additional decrease in temps.
Final recommendation: Go hang out over at the XS forum and absorb as much as you can. Re-assess when you know more. Also consider the "power-density" of the TEC - that is the cooling power per mm^2. The gigantic 545 ebay TEC, which may be listed as maybe 437w from the manufacturer...not sure... may be a 62mm or 50mm square, for instance. I'm sure the designers will take into account both the efficiency at the desired voltage range and the cooling power per unit area.
Also remember that you'll need a dedicated PSU for the TECs. In the past I had actually used a PC PSU. I adjusted the 12V rail up to 13v, and then used 6.5v for an even number of TECs. I used multiple power and ground wires. I bet this isn't recommended practice, but it works on PSU's with easy potentiometer voltage-rail adjustment.
Last edited:
- Joined
- Mar 18, 2010
- Messages
- 14,451
TECs are notoriously inefficient and usually not worth the cost, both upfront and on the electricity bill.
With your 3960x, you'll most likely run into a frequency wall long before you run into a thermal wall on a good watercooling setup.
With your 3960x, you'll most likely run into a frequency wall long before you run into a thermal wall on a good watercooling setup.
Update, i found a tec/waterblock i'm gonna get and test on my i7 2600k. I will be using a haf xm case. Can i get some ideas on parts i will need. Like what rad, looking for a 360 rad in push/pull. Just gonna cool cpu for now. Also will be adding 2 liquid chillers to loop down the road. Whats a good pump and rez?
- Joined
- Mar 18, 2010
- Messages
- 14,451
If you're going to go with liquid chillers, skip the radiator and go straight to those. Skip the TEC as well, as the chillers will be more efficient.