• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

TEC or Phase-Change? Which uses more electricity? (with identical cooling results)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

blackjackel

Member
Joined
Oct 18, 2002
Location
Los Angeles
TEC or Phase-Change? Which uses more electricity? (with identical cooling results)

If you don't know the topic of the thread yet, you are either slow or you clicked the link too fast :D




While i am at it, why not ask another question...eh? here it comes:

What would be the best peltier (or peltier wattage) to get when using my 350 Watt ANTEC PSU to power it...

(3.3V & 5V combined output 230W......... also.............3.3V & 5V AND 12V combined output 330W)


And my third and final question: According to this:

(3.3V & 5V combined output 230W......... also.............3.3V & 5V AND 12V combined output 330W)

Does that mean 12V output would be 100W?



P.S: If it helps in answering the first question, i do run watercooling at he moment, and a very nice watercooling setup at that =P Oh, and i'm running a 1700+ OC'd at 2300mhz with my watercooling, i wanna push it even higher with peltier, how much more would i get out of it?


Thanks!
 
Last edited:
OP
blackjackel

blackjackel

Member
Joined
Oct 18, 2002
Location
Los Angeles
also, if i DO get a TEC, should it be just big enough to fit over the CPU core? Or should it be large enough to cover the surface of my waterblock?

Whats the best place to get TEC's from? Considering price/performance, best price for the most performance?

I will do some further searching to find this out incase it goes unanswered =) thanks in advance though to anyone answering any of those questions...
 

Paxmax

Member
Joined
May 8, 2002
The TEC consumes far more electricity than phase change for the cooling we need.
I'm not 100% sure how much more but it'll be about twice as much or more.

You'll be wanting at least a 172W TEC or more to have a chance of some cooling. I would rather aim at the 226W TEC, just to be on the safe side.

The TECs size will be accordingly to it's cooling power. So 226W TEC will be big, count on 50 x 50 mm, so you need a peice of copper between cpu and peltier to duct the coolness to the cpu-die.
Currently there is no way to manufacture a TEC with 226W cooling with the size of a cpu die.

The performance inrease of your new TECed setup will be somewhere between 0 and maybe even 200 MHz more, it's really hard to tell.

You'll be definetly need a really good watercooling setup to reap benefits with a TEC. Make sure your waterblock covers the whole TEC and you need ALOT of clamping pressure between coldplate and waterblock.

Also remember, the TEC is really a heatpump, so it removes the cpu heat at one side and dumps it on the other side + the peltier heat it created with the power it consumes to move heat. So most of the heat is from the peltier itself.
 

gusgizmo

Disabled
Joined
Aug 10, 2003
Location
Kona, Hawaii
yes, most definetly. TEC's are very inefficient in their operation and usually require more energy for worse results than phase change. for the same results as a TEC in direct-die, you would probably use a 1/10 hp compressor, which is 73 watts, versus a 220 watt pelt. for a worthwhile waterchiller, you would want to use a 1/4-1/3 hp compressor, and you might reach temps of -25c and below. so that would be from 185watts to 250watts for much better results.
 

Ven0m

Member
Joined
Sep 7, 2003
Location
Warsaw, Poland
According to my knowledge, Phase Change (efficiency of 40-45%)is about 8-9x more efficent than TECs (efficiency of around 5%, not these new prototypes which have efficiency of around 50-55%, but used now).

This means that you to have same temp as 200W phase change, using TECs you'd need about 1.6-1.8kW (LOL). How much energy goes to moving energy - is alredy written above, rest goes to heat.

TECs are used as water chillers because they produce lots of heat and it's easier to chill them in better place than on CPU, other aspect for water chillers is size that they can be - more than mobo could resist.

I'd not recommend using phase change as "water" chiller, it's better to have phase change block. Advances of using water chillers with TECs don't appear with phase change, so in this case there's no use in having them.
 
OP
blackjackel

blackjackel

Member
Joined
Oct 18, 2002
Location
Los Angeles
Ven0m said:

TECs are used as water chillers because they produce lots of heat and it's easier to chill them in better place than on CPU, other aspect for water chillers is size that they can be - more than mobo could resist.

Easier to chill them? whats them? Chill the TEC? why would you want to chill the TEC?

Also what do you mean about size and resistance? If i make a waterchilling thing it would be seperate from the case, it would be an outside thing, not in the case.

and what do you mean more than the mobo can resist?

Ven0m said:

I'd not recommend using phase change as "water" chiller, it's better to have phase change block. Advances of using water chillers with TECs don't appear with phase change, so in this case there's no use in having them.

Please elaborate on "advances", why dont they appear with phasechange? What advances?
 

gusgizmo

Disabled
Joined
Aug 10, 2003
Location
Kona, Hawaii
This means that you to have same temp as 200W phase change, using TECs you'd need about 1.6-1.8kW (LOL). How much energy goes to moving energy - is alredy written above, rest goes to heat.

TECs are used as water chillers because they produce lots of heat and it's easier to chill them in better place than on CPU, other aspect for water chillers is size that they can be - more than mobo could resist.

I'd not recommend using phase change as "water" chiller, it's better to have phase change block. Advances of using water chillers with TECs don't appear with phase change, so in this case there's no use in having them.

blackjackel said:
Easier to chill them? whats them? Chill the TEC? why would you want to chill the TEC?

i think he meant cool them.

Also what do you mean about size and resistance? If i make a waterchilling thing it would be seperate from the case, it would be an outside thing, not in the case.

i think he meant that a waterchiller is better because the size and weight of a pelt, coldplate, waterblock, mounting bracket, and insulation.

and what do you mean more than the mobo can resist?

hes talking about the motherboard possible being damages due to the weight or whatever.

Please elaborate on "advances", why dont they appear with phasechange? What advances?

      I think that he meant advantages here. Cut the man some slack, english is his second language if I inferred correctly from his location.

      Anyway, considering the cost of running a peltier setup, any advantage of running a TEC waterchiller is quickly, and thoroughly, diminished. I don't agree with Ven0m on the argument that a TEC waterchiller is better than running the TEC on the processor.

      Also, the point on it being recommended to run a direct-die block instead of a waterchiller, there is more than one view on this point. While it is true that you can get better performance with the direct-die block, for the average user that does not have experience with refridgeration it can be a daunting task to get all the parts necessary, empty the system, make the modifications, recharge the system, and install it properly.

      A "water"/liquid-chiller using a phase change loop removed from an already built system such as a refridgerator or air conditioning unit is much easier to build, since you dont need to modify the system at all beyond removing the loop from whatever housing it has, and bending the pipes into an appropriate shape.
 
OP
blackjackel

blackjackel

Member
Joined
Oct 18, 2002
Location
Los Angeles
Thanks for clearing that up. Helped alot....


I just bought this off ebay for 8.99 plus tax and shipping:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2583963166&category=4660

Its the same as this:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2582761183&category=4660

says its 130-160W but can do 180W. I plan on powering it with my old 350 Watt Antec PSU's 12 volt rale that gives out 16A max (the TEC requires 15)



Would that be nice to put on my 1700+ (currently OC'd to 2.3 with water)

Do you think ill be able to hit 2.5 maybe even 2.7?
 

gusgizmo

Disabled
Joined
Aug 10, 2003
Location
Kona, Hawaii
i wouldnt plan on it. it may reduce your temps, but its 50 watts behind the curve, so i dont think it will be able to hit the sub-zero temps you need to hit 2.5 or 2.7. might be worth another 100MHz though.
 
OP
blackjackel

blackjackel

Member
Joined
Oct 18, 2002
Location
Los Angeles
what if i use the sucker to cool down my videocard with a huge heatsink and large fan? its a 9500 np unsucessfull softmod. Maybe i can get a sucessfull softmod with this pelt?

Maybe i can even hit 9800 speeds.... no?
 

__TRONIK__

Member
Joined
Sep 28, 2003
Location
washington d.c.
lol @ softmod with cooler temps! the softmod enables pipes - one of your pipes is broken and even fairy dust ain't fixing it! I think I am going to dive into pelt cooling the same gpu - i have heard 500gpu with an 80watt pelt is possible. Me likie.
 

Paxmax

Member
Joined
May 8, 2002
Ven0m said:
According to my knowledge, Phase Change (efficiency of 40-45%)is about 8-9x more efficent than TECs (efficiency of around 5%, not these new prototypes which have efficiency of around 50-55%, but used now).

How did you (or someone) come up with an efficiency of just 5% ?
I'd like to see the calculus for that...
Also, almost no one uses the TECs of full blast, can you give me a rough estimate of efficiency of a TEC operating at 80-90% of Vmax please?
 

L337 M33P

Member
Joined
Jun 5, 2003
Location
TEH INTERNETS
For a TEC operating at specified Vmax and Imax:

http://www.computermods.com/Danger_Den_226_Watt_Potted_Peltier.html

226W peltier = 226W of heat moved from cold-->hot

Imax = 24 Amps
Vmax = 15.2 Volts

Power used by the TEC is: 24x15.2=364.8 ~360W

Efficiency = Useful work done/power input to achieve work

=226/360
~63%

The peltier will be more efficient at a lower voltage, but not by much. I estimate 70% efficiency max at a decent delta T.
 

Paxmax

Member
Joined
May 8, 2002
My thoughts exactly m33p.

I'm no ace on phasechange, but can't you get maybe around 300-400W of heat moved (*extreme case, special conditions*) with a 100W compressor?

So, that would mean an efficiency of 300%+.

With a 226W TEC operating at 80-85% of Vmax you can get (*rough numbers, extreme case*) 80% efficiency.
You input 240W(12V, 20A) energy and it moves about 200W heat.

With extreme case I mean low Delta temperature, goes for both TEC and phase change.

These are just numbers in the ballpark, I can't nail 'em down with 100% accuracy. I just can't belive phasechange would be 8 times more efficient overall.