Impressive thermo electric(peltier) air cooler out. The Titan Amanda

[Praxis1452]

Originally posted by: Bobthelost

Originally posted by: Madellga
The cooling is done by the heatsink, not the peltier. The peltier helps one move the heat from one face to the other. I didn't do the math to check this case, but this is the principle behind a heat pump system. If you look to an air conditioning system, you use less power to feed the system than the heat it is actually moving from one source to another. Typical thermodynamics...

Almost:

1) The amanda draws 60W, including two 120mm fans. Assuming they draw a mere 1A each on the 12V rail that's 24W gone, 36W left to move the heat.

2) Peltiers are horrifically inefficent. 50% is a good rule of thumb equation. In other words to shift 100W you need to put in 100W, all 200W now needs to be cooled by the heatsink. In our case you're not even managing a third of a high end P4D.

A mere 1 amp for fans? umm Only Delta's may draw that much. A panaflo uses pushing 110CFM only draws 6 watts. In reality those fans may be drawing .1 amps or a little more. They are only 92mm and go to 2400rpm which is I would say medium speed.

Peltiers are only aorund 5% efficient I believe not 50%.

The delta T formula works like this.
Delta T = (1 - (Heat Load/Max Cooling Power)) * Max Temp Difference General estimate.
Heatload= processor.

Cooling power=TEC wattage. Max temp difference is the max temp differential of the TEC. Max is 69C. It is possible for a TEC to have a max temp differential of 69C however you will have no load...

 

[Bobthelost]

Originally posted by: Praxis1452
A mere 1 amp for fans? umm Only Delta's may draw that much. A panaflo uses pushing 110CFM only draws 6 watts. In reality those fans may be drawing .1 amps or a little more. They are only 92mm and go to 2400rpm which is I would say medium speed.

Yeah, you may be right on that after all, vantec fan drawing a mere 6W or so for that speed range (LED though)
http://coolerguys.com/840556019121.html

So the Peltier is looking more at a 45-50W Pelt.

Peltiers are only aorund 5% efficient I believe not 50%.

The delta T formula works like this.
Delta T = (1 - (Heat Load/Max Cooling Power)) * Max Temp Difference General estimate.
Heatload= processor.

Cooling power=TEC wattage. Max temp difference is the max temp differential of the TEC. Max is 69C. It is possible for a TEC to have a max temp differential of 69C however you will have no load...

Peltiers are more efficent than 5%! For that to be true you'd need a 2000W PSU to cool a 100W CPU!

Maybe you're thinking of it the wrong way, that you only expend 5% of the power transported. In which case you're completely and utterly wrong.

 

[markymoo]

i here what you saying its inefficient i will be providing better aircooling inside the case and around the cpu area. i just want that bit more and its the next thing to try without switching to water. im not going to be put off by one bad review. incidentally no reviewer will say a certain product is bad if the manufacturer is sending them ample products they want.

it seems alot better than that coolit solution. with that the temps shot up dramatically on
load.

@cmdrdredd
what temps do you get using the thermalright Ultra-120 & Panaflow H1BX ?



ocz showed off a promising phase change cooler for $200 early this year. does anyone know anything further on it?

http://www.legitreviews.com/article/282/1/

Originally posted by: cmdrdredd
You don't know how ignorant you sound... "all Conroe can hit 4Ghz" do you really believe that trash you just spewed out?

My friends have p5b deluxe with good hsf and they all get 3.4 3.6 and have high temps... the temp is preventing them from reaching further. they on a low voltage and some have the p5wdh deluxe and H2O and they get 410fsb 3.69 the board being the limiting factor. we talking 300-400mhz more in both cases. so lower temps and change of board be the success in this case.

 

[cmdrdredd]

Originally posted by: markymoo
i here what you saying its inefficient i will be providing better aircooling inside the case and around the cpu area. i just want that bit more and its the next thing to try without switching to water. im not going to be put off by one bad review. incidentally no reviewer will say a certain product is bad if the manufacturer is sending them ample products they want.

it seems alot better than that coolit solution. with that the temps shot up dramatically on
load.

@cmdrdredd
what temps do you get using the thermalright Ultra-120 & Panaflow H1BX ?



ocz showed off a promising phase change cooler for $200 early this year. does anyone know anything further on it?

http://www.legitreviews.com/article/282/1/

Originally posted by: cmdrdredd
You don't know how ignorant you sound... "all Conroe can hit 4Ghz" do you really believe that trash you just spewed out?

My friends have p5b deluxe with good hsf and they all get 3.4 3.6 and have high temps... the temp is preventing them from reaching further. they on a low voltage and some have the p5wdh deluxe and H2O and they get 410fsb 3.69 the board being the limiting factor. we talking 300-400mhz more in both cases. so lower temps and change of board be the success in this case.

The problem is not cooling, but that some CPUs regardless of any cooling used just don't get 100% stable at a certain speed. my Ultra-120 keeps my E6400 @3.2Ghz under 58c at load. I have a week24 and I can't get stable at 3.4 without close to 1.45v (actual after vdroop) and I'm unwilling to use that voltage regardless of temps.

Some of it has to do with memory not liking a higher FSB. Depends on brand etc.

 

[markymoo]

@cmdrdredd
look at this. the noctua nh-u12 and themalright Ultra-120 the results are so close.
http://www.neoseeker.com/Articles/Hardware/Reviews/thermalright_ultra_120/3.html

Originally posted by: cmdrdredd
Some of it has to do with memory not liking a higher FSB. Depends on brand etc.

theres brands of good quality memory can both do 1000mhz why would one be unstable if both brands are both stable on the p5b and well within the capabilities of there speed be the cause of the unstability? the memory is not being overclocked. regardless of its high fsb the memory running within its limits. hmmmm

 

[markymoo]

@cmdrdredd
your ultra 120 wins but we are talking slight and even then they used a superior thermal paste and better fans. are coolers are passive work good without fans but my fans cool it further and the better the fans...

im not going to 1.7v 1.65 its debatable whats too much. im sure but 1.575 on air is my limit and what i consider the MAX pushing it hot but not to its breaking point. keeping it under 1.6 is ok for 24/7. my friend on H2O + p5wdh had to take it to 1.6 to get 410fsb and stable, all the other voltages were raised in unison. a different board im sure he would of not had to raise it as high. he was using top quality rad thermochill pa160 and storm apogee.

ok so i knock off 4 years off the life of my cpu. i will of upgraded well before then. at least i get the best out of it and have the better speeds. incidentally the core can take 140c before it physically damaged.

i will retract not all will do 4gig but not everyone got the hardware to get there so they wont know but all i know alot can get to 3.6 and then past there its been a heat issue so they dont know whats the limit but they all think theres more potential. in some case some can only goto 3.2 no matter what they try but are trying hard enough.

 

[thilanliyan]

Originally posted by: markymoo
incidentally the core can take 140c before it physically damaged.

Where did you get that information?

 

[markymoo]

heres some encouraging benchmark results from the titan amanda

link

link

@thilan29
the core has been reported from intel to damage permanent at 140c and the solder that keeps it attached to the intel heatsink melts at 77c and below 61.5c the safe limit to run it 24/7

 

[markymoo]

titan amanda now at performance pcs - $89

http://www.performance-pcs.com/catal...ducts_id=20692

 

[smopoim86]

Originally posted by: Bobthelost
How does a 60W pelt cool a 150W chip, not to mention how does it do that while keeping the heatsink merely warm? I smell bullshit here.

exactly, also it works against its self, half the heatpipes are attached to the cold plate(thats how it keeps from frying the cpu).

For a pelt to cool a processor the pelt has to be a higher wattage and then the heatsink has to cool the combined wattage of both the pelt and the processor.

 

[cmdrdredd]

Originally posted by: markymoo
@cmdrdredd
look at this. the noctua nh-u12 and themalright Ultra-120 the results are so close.
http://www.neoseeker.com/Articles/Hardware/Reviews/thermalright_ultra_120/3.html

Originally posted by: cmdrdredd
Some of it has to do with memory not liking a higher FSB. Depends on brand etc.

theres brands of good quality memory can both do 1000mhz why would one be unstable if both brands are both stable on the p5b and well within the capabilities of there speed be the cause of the unstability? the memory is not being overclocked. regardless of its high fsb the memory running within its limits. hmmmm

I've seen memory rated at DDR2-800 fail at DDR2-533 at 6-6-6-18. Why? simply because nothing is guranteed and had to RMA it. Plus no motherboard manufacturer will ever gurantee more than stock. They offer adjustments as a use at your own risk deal. They offer no gurantee you'll get 1Mhz more.

 

[cmdrdredd]

Originally posted by: markymoo
@cmdrdredd
your ultra 120 wins but we are talking slight and even then they used a superior thermal paste and better fans. are coolers are passive work good without fans but my fans cool it further and the better the fans...

im not going to 1.7v 1.65 its debatable whats too much. im sure but 1.575 on air is my limit and what i consider the MAX pushing it hot but not to its breaking point. keeping it under 1.6 is ok for 24/7. my friend on H2O + p5wdh had to take it to 1.6 to get 410fsb and stable, all the other voltages were raised in unison. a different board im sure he would of not had to raise it as high. he was using top quality rad thermochill pa160 and storm apogee.

ok so i knock off 4 years off the life of my cpu. i will of upgraded well before then. at least i get the best out of it and have the better speeds. incidentally the core can take 140c before it physically damaged.

i will retract not all will do 4gig but not everyone got the hardware to get there so they wont know but all i know alot can get to 3.6 and then past there its been a heat issue so they dont know whats the limit but they all think theres more potential. in some case some can only goto 3.2 no matter what they try but are trying hard enough.

Again this is using water which is not the same as air at all so your argument means nothing. And no the core can't do 140c...if it could then you'd never need to buy a heatsink and intel's stock junk would overclock to 5ghz :roll:

 

[markymoo]

@cmdrdredd
lyou misunderstand i never said the core can do 140c i said it gets permenantly damaged physically

sure not every cpu and memory made the same and incompatibilitys etc.

 

[markymoo]

anyway even if its inefficient. people had nightmares of pelts in the past this seems a better success. the tec isnt always on all the time. it comes into its own when its overclocking hot.

the results speak for themselves. i prefer a titan over a infinity. they both big but the titan will cool better.

 

[0roo0roo]

how many watts does a water pump take?

 

[aka1nas]

Originally posted by: markymoo
@cmdrdredd
lyou misunderstand i never said the core can do 140c i said it gets permenantly damaged physically

sure not every cpu and memory made the same and incompatibilitys etc.

61C is the max safe rated temperature for a 65-watt core2Duo as per Intel's datasheet:
http://download.intel.com/design/processor/datashts/31327802.pdf

Anything above this has the potential to do damage to the processor.

 

[Elfear]

Originally posted by: 0roo0roo
how many watts does a water pump take?

Depends on the pump. The 50Z is very efficient for the amount of head and water flow. I think it uses ~7W while some of the larger AC pumps, like those from Iwaki, use 40-50W.

 

[markymoo]

@aka1nas
you misunderstand also. i already had said 61.5c is the safe limie to run it stable. people have run them upto 90-100c and its still posted. i said to actually damage it beyond repair like melt it inside is 140c! thats nothing to do with 61.5c safe running it.

 

[cmdrdredd]

Originally posted by: markymoo
@aka1nas
you misunderstand also. i already had said 61.5c is the safe limie to run it stable. people have run them upto 90-100c and its still posted. i said to actually damage it beyond repair like melt it inside is 140c! thats nothing to do with 61.5c safe running it.

85c = throttle kickin

 

[markymoo]

heres thread on a user who bought one. result and pics included.

http://www.hardforum.com/showthread.php?t=1099746&page=1&pp=20