backtop


Print 91 comment(s) - last by SoCalBoomer.. on Mar 7 at 1:02 PM


MSI ECOlution Chipset Cooler  (Source: TweakTown)
MSI ECOlution chipset cooler operates on the Stirling Engine Theory

MSI has designed a new chipset cooling fan that is able to operate without electricity. MSI’s new chipset cooler, which is accordingly dubbed the “Air Power Cooler,” offers all of the benefits of a cooler with a fan without drawing any power.   

Energy efficiency of fans can make a large difference, especially in enterprise environments where hundreds of PCs are running at once.  Although passive cooling is always an option, it doesn’t offer the cooling capability of a fan.

The new MSI cooler isn’t a passive cooler but actually uses a fan to cool the chipset without using any electricity. Ironically, the fan gets its power from the very thing it’s trying to remove — thermal energy.

The system is based on a beta Stirling engine. As hot air expands in the system, it applies pressure to the central piston in the heatsink pushing it up. The piston's movement upwards rotates gears which in turn spin the fan. Thermal energy generated by the chipset is converted into kinetic energy.

The fan blows through a common looking finned radiator to disperse the Northbridge’s heat production.

MSI tells DailyTech that the system is able to convert 70% of heat power to kinetic energy. It is important to note that enough heat must be supplied to spin the fan blades. If the chipset isn’t hot enough, the entire system will not run.

MSI is working on the cooler with Taiwanese company Polo-Tech. The powerless fan is expected to make its debut on MSI’s ECOlution during CeBIT 2008.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Efficient and automatic?
By ninjit on 2/29/2008 8:22:51 PM , Rating: 4
70% efficiency is pretty good.

And seeing how this thing will depend on the heat generated by the chip, it seems like it could be fully automated in it's operation as well, without the need for a thermistor and control logic.

i.e.
The hotter the chip is, the faster the fan spins, n'est-ce pas?
Just how we would want it.




RE: Efficient and automatic?
By ImSpartacus on 2/29/2008 8:25:37 PM , Rating: 2
I'm a little skeptical of the 70% efficiency, but I can't complain if its true, that would be sweet.


RE: Efficient and automatic?
By lukasbradley on 2/29/2008 9:01:10 PM , Rating: 2
I agree completely. 70% of what? The ambient air temperature?


RE: Efficient and automatic?
By Comdrpopnfresh on 2/29/2008 9:17:39 PM , Rating: 5
70% of the thermal energy absorbed by the unit. This isn't anything new. Stirling-engine fans are used when set on top of oil-heating units on rural areas. There is even room to increase the efficiency of this process, by making the heat-absorbing parts a better "blackbody"


RE: Efficient and automatic?
By AnnihilatorX on 2/29/2008 9:25:14 PM , Rating: 2
Doesn't law of thermal dynamics give a maximum theoretical efficiency of a heat engine?

Efficiency = 1- Tcold/Thot something like that. The absolute maximum efficiency you can extract from a temperature gradient is related by that formula.

A chipset may generate a temperature difference of say 50 Kelvins. Assuming chassis temperature is 300 Kelvins. 1- 300/350 = 14% efficiency maximum.


RE: Efficient and automatic?
By Duwelon on 2/29/2008 10:35:01 PM , Rating: 2
In laymens terms, your saying that the efficeny is an equation between the ambient temperature and the temperature of the heat source, ie the chipset?


RE: Efficient and automatic?
By masher2 (blog) on 2/29/2008 11:41:19 PM , Rating: 2
Basic Carnot efficiency, yes. That's why ultra-high temperature power plants, for instance, are so much more efficient than a car's engine.


RE: Efficient and automatic?
By drinkmorejava on 2/29/2008 11:16:12 PM , Rating: 2
I believe you are correct in using the carnot cycle efficiency, however, you used the efficiency given by a heat producing engine (1-Qdotl/Qdoth)˜1-Tl/Th. In this case though Tl and Th are not readily apparent. If you use the ratio of (Wdot/Qdot)=(Power)/(Heat Transferred) it becomes apparent that it is possible to achieve high efficiencies. For your reference, power is equal to the work done by the fan per unit of time.


RE: Efficient and automatic?
By dare2savefreedom on 3/1/08, Rating: -1
By marsbound2024 on 3/1/2008 4:22:12 PM , Rating: 5
You know how you spend hours a day watching American football or baseball or any other sport? Not you, you say? How about spending hours working on your car? Not you, you say? How about spending hours playing paintball? What about reading books? Listening to music? Photography? Stargazing? Gaming?

The simple fact of the matter is that we all have our hobbies and love our hobbies when they only get better and better? Also, I am sure plenty of us have girlfriends and many of them actually have the same hobbies as we do. And no, they are not all ugly.

Maybe sir, you need to think more and respect the fact that everyone likes different things and that is what makes being a part of civilization worth it.


By drinkmorejava on 3/2/2008 3:28:15 AM , Rating: 3
haha, thank you. For your information, I do have a girlfriend. Furthermore, I would like to know how being an aerospace engineer precludes me from having a life.


RE: Efficient and automatic?
By ViperV990 on 2/29/2008 8:30:02 PM , Rating: 2
I think the fan speed depends on the temperature difference between the hot end and cold end of the engine, i.e. the temperature of the chip vs. the ambient temperature.


RE: Efficient and automatic?
By leexgx on 2/29/2008 10:46:19 PM , Rating: 2
the chipset needs to be at 60c before it starts working but the hotter it gets the faster it spins


RE: Efficient and automatic?
By ImSpartacus on 3/1/2008 3:01:58 PM , Rating: 2
I'm not sure everyone would even want their CPU's hitting 60c to begin with. If that's the only time the fan starts up then there will be some problems. I'm sure it still runs when it's below 60c.


RE: Efficient and automatic?
By ImSpartacus on 3/1/2008 3:02:41 PM , Rating: 2
Nvm, chipset cooling not CPU. For chipset I guess it would be ok.


RE: Efficient and automatic?
By Anonymous Freak on 3/1/2008 11:24:44 PM , Rating: 2
Not only is it chipset, but 60 °C isn't that big a deal. My CPU runs at 66 °C (Overclocked 4.0 GHz Pentium Extreme Edition at 100% usage all the time due to distributed computing,) pretty much constantly. And I have a reasonably high-end cooler.

With a stock Intel cooler, a Core 2 Duo at stock speeds can hit 50 °C under load. And that's with good case cooling. In a bad case, I've seen stock CPUs hit 70+, easy. Older Pentium 4 notebooks under load can hit nearly 100 °C.


RE: Efficient and automatic?
By ImSpartacus on 3/2/2008 9:14:48 AM , Rating: 2
I disagree. There is no set temperature that good for a CPU (or chipset), but I have met people on both sdes of the spectrum. Some don't want their entire PC getting above 50c, others run 65c 24/7.

I personally would not run my CPU at stock. My chipset may or may not be slightly oc'ed. I would not be running on stock cooling. I would not like my CPU to run 24/7 over 50c, maybe 60c at unusually high load. My other components may get warmer, but I would try to preserve one of the most expensive parts on my build.

It comes down to opinion, people differ. Nothing works for everyone.


"We basically took a look at this situation and said, this is bullshit." -- Newegg Chief Legal Officer Lee Cheng's take on patent troll Soverain











botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki