Pump that soca, cool that chip

[W1nTry]

Interesting:

Chips cooled with noise

Do you mind speaking into the bag sir?

By Ambrose McNevin: Friday, 25 January 2008, 10:10 AM

LIQUID COOLING of computer chips and other heat intensive technology can be boosted by blasting sound waves at the coolant.

The New Scientist reports that researchers at the Georgia Institute of Technology have come up with sound enhanced liquid cooling.

Liquid coolants are being explored as an alternative to fans and heat sinks as processing power increases and becomes more heat intensive. Liquids are most efficient for heat transfer when allowed to boil away excess energy but this can lead to the formation of bubbles which act as insulation thus negating the effect.

But Ari Glezer and his colleagues at the Georgia Institute of Technology have been working on the problem and have come up with a new way to get rid of the bubbles. By placing a speaker a few millimetres from the origin of the heat - in this case a chip - with cool liquid in between and hitting it with sounds near 1 kilohertz they could get rid of 147 per cent more heat.

The approach which Glezer describes as “elegant” is being touted as a possible answer to heat build up in everything from chips to hybrid vehicles and spacecraft. µ

[Crixx_Creww]

hmmm
makes sense , except yu would then have to use more power for the speakers...
not the most energy efficient solution

but a solution none the less

[W1nTry]

how so crixx? you not using a big speaker or anything
it'll be a really small speaker with a 1Khz signal, that's like a phone even with a lower resposne... hardly a Z5500 meng

[Crixx_Creww]

i know its not ALOT of power, but yur still introducing MORE energy into the equation
its not the MOST energy efficient i said, but it is a good solution

passive is always best

otherwise we might end up like the bugatti veyron with its 5 radiators
1 main radiator for the engine, and then 4 radiators to cool the radiator...

[W1nTry]

Well considering they were talking about liquid cooling to begin with, the idea is cooling which would already involve a pump, so using more than a passive arrangement is ALREADY out the window... now from an effeciency point of view crix, performance per watt would be the measurement. So if by introducing a miniscule amount of extra power allows you to clock higher and run cooler, then performance would increase with nominal increase in total enenery input, so by that thinking, its actually more efficient than passive, since you just CAN'T get that kind of head room from passive cooling.

[Crixx_Creww]

you can get that kind of head room from passive cooling

cept yu would be sacrificing space

and liquid metal cooling doesnt use pumps so its passive and if they add the speaker to liquid metal cooling that could be a nice alternative...

but generally speaking, adding my power to the equation is not the most efficient option.

[W1nTry]

I've read liquid metal cooling and it was deemed non-cost effective by sapphiretech. They were developing the tech for their cards but realized that SURE it had great absorbtion properties, BUT it was HORRIBLE at getting RID of the heat transferred to the liquid metal, meaning you pull the heat from teh chip THEN WHAT? they were stuck with the problem of having a HOT liquid which in turn made it WORSE for cooling. That's why there are no production used liquid metal cooling systems in PCs. But think of it this way, if you have a liquid cooling system which uses a pump AND the sound arrangment, you could probably use a SMALLER pump thus LESS energy and the speaker would fit the reduced power envelope, thereby maintaining the same power consumption as the original liquid system but with MUCH better results. THereby overall it would be a win win.

[daniboy79]

hmmm
makes sense , except yu would then have to use more power for the speakers...
not the most energy efficient solution

but a solution none the less

the higher the frequency, the less power needed. (in the case of theirs being 1Khz) there would'nt be a need for much energy for the job.

[Crixx_Creww]

yu said the higher the frequency the less power needed

but 1 khz is like mega LOW frequency O.o inaudible rumble type frequency that is

[daniboy79]

yu said the higher the frequency the less power needed

but 1 khz is like mega LOW frequency O.o inaudible rumble type frequency that is

 

1Khz = 1,000Hz

a typical Sub-woofer starts around 10~15Hz and 'rolls-off' at less than 90Hz. so ...1Khz is about 10 times higher in frequency than a sub. and most of us humans stop hearing frequencies upwards of 20,000Hz whcich is 20 times higher than 1Khz.
now, what's was that about 'inaudible rumble type' frequency?

if only i had the time to look for an online tone generator, i would've posted a link. what i do have is a free download/install for a tone generator i use for room/hall acoustic tests/speaker setup.

http://www.nch.com.au/tonegen/index.html

how many of you are musicians? lol ...sorry i had to ask

[Crixx_Creww]

A more rigorous exploration of the lower limits of audibility determines that the minimum threshold at which a sound can be heard is frequency dependent. By measuring this minimum intensity for testing tones of various frequencies, a frequency dependent Absolute Threshold of Hearing (ATH) curve may be derived. Typically, the ear shows a peak of sensitivity (i.e., its lowest ATH) between 1 kHz and 5 kHz, though the threshold changes with age, with older ears showing decreased sensitivity above 2 kHz.

http://en.wikipedia.org/wiki/Psychoacoustic_model

so basically a low almost inaudible rumble at 1kz hmm


that tone generator is cool though

changing the shape of the waves produces surprising tests of my hearing.

my lower limit on a sine wave seems to be 206hz, and here i thought i had remarkable hearing.

lol playing with it again

i went from 22000 all the way down to 206 hz
and when i hit 420 hz my computer started to rattle lol lol
well the computer in the office not mine per say

[daniboy79]

lol... i found the resonant frequency of my house with that program (43Hz)

...had everyone thinkin there was an earthquake goin on!! muhahahahaa

hey crixx, try using at least 4 tones in sinewave, and make some tones start from high and go low, andsome low to high and some just vary in the middle frequency. talk about Twilight Zone sound effects!!

[W1nTry]

One would find it interesting that effectively the sound waves are adding kinetic energy to the cooling fluid which SHOULD in effect increase its temperature, HOWEVER it dramatically REDUCES the temperature... since men went off ona tangent about sound... put some science to what I just said and give an idea as to what is going on here....

[Arcmanov]

Well, it would be highly inefficient to boil the cooling liquid, which is the opposite of what they are trying to achieve here.
Applying the sound increases the kinetic energy of the particles within the liquid, but it seems as if its also increasing the speed
at which heat is being absorbed, transmitted, and then radiated.  Since the liquid cooling system is closed, then there would be no
chance of bubbles forming, since no outside heat is being applied.

Very interesting phenomenon here, and a whopping 147% increase in cooling performance!!?
Not bad at ALL.

How much for one? 

[W1nTry]

Well I figuring yuh buy yuhself a nice all-in-one water cooling kit like the Corsair Nautilus, then jerry rig a 2" speaker with a small ic outputting 1KHz close to the reservoir/cooling fins and voila!

[daniboy79]

that will be the worlds most annoying computer to work on - but then, ah hear Arc at one time had some ThermalTake fans that woulda make that noise seem insignificant

i'll keep my eye on these developments for sure, i don't know how much i want a computer that sounds like a UPS runnin on batteries right thru.

[Crixx_Creww]

its because of eddy currents created by convection in the water

the sound waves at 1kz probably create a wave motion effect that negates the eddy currents and increases the efficiency of the convection cooling.

[W1nTry]

You won't hear 1Khz dani... not to mention that that 'hum' you hear from electrical pieces of equipment is usually the freq of the current which in trinidad is ~60Hz.... sheesh... Interesting take there crixx... now the question is... anyone capable of trying it out?

[Arcmanov]

ah hear Arc at one time had some ThermalTake fans that woulda make that noise seem insignificant

Yep.  Those were some 80 mm bitches that spun at 5500 rpm.  Cooling was great, but three of them at
that speed sounded like an F-15 on takeoff.  Thank god for 120 mm.

[W1nTry]

I still have one of those fans actually... it's cooling a socket A Barton core! Thermaltake Volcano 9 ftw!!! if I crank dat b1atch it does get quite loud... I can only imagine 3 of them O_o