Electromagnetically propelled cold plasma as a means of cooling

E

eagle_orion

n00b
Joined
Mar 9, 2008
Messages
18
What are the technical difficulties / impracticalities / "ineffectivenesses" of integrating a "micro" heat pump on a motherboard? What I'm referring to here is, basically, a refridgerator engine.

In the case of the motherboard, you could use a very small pump device to push a gas into a network of very small tubes, maybe even filament like, on the back of the motherboard. This would pressurize the gas, heating it up, and allowing it to release heat as it moved through the tubes. The tube-network would lead to chipsets, cpus, and other hot spots. Shortly before reaching these points, the tubes would converge into bottlenecks that keep the gas before these points pressurized and the gas after depressurized. Once the gas passes through a bottleneck, it enters a compartment directly under a processor heatsink, where it depressurizes and cools considerably below room temperature, thus allowing the processor to dissipate heat without using heavy duty cooling devices that are compromised by the necessity to work at room temperature.

In fact, really, this would be a re-distribution of heat across the motherboard, instead of direct dissipation to the environment from specific points (like the cpu.)

Now as for what you'd use as a pump that works on such a small level like that without producing noise, I'm not sure. I had the idea of using a changing magnetic field and one-way valves to propel an ionized gas. Not sure if that's feasible though.
 
This just sounds like you want to use a pc wide phase changer, and this it would be possible, but it would be impractical since you would need to have it be air tight all of the time to prevent condensation. More realistically you should look at multi-point phase changers maybe a 3 stage one for something similar to this.
 
So someone invented it before I did? Dang.

Well, that was predictable. But when I look at some of these, they seem to mind-bogglingly extreme performance coolers. Does anyone REALLY NEED to implement this idea only in a form that cools the cpu so much, your surprised it doesn't liquify oxygen? The idea I had was something on a par with high grade air cooling performance, but with NO MOVING PARTS, in other words, absolutely dead silent. Isn't it safe to say that if you WAY cut down on the performance of a phase cooler, you can reduce the size of a phase changer considerably and reduce the necessary elements of the device, allowing you to easily integrate it into the chassis, and buy it for a whole lot less? Why has this not been attempted? Am I missing something? I can see why an EXACT version of the scheme I presented would have issues (condensation, of course,) but that can be prevented with the appropriate thermal insulation, and separation from the motherboard, as I'm sure current phase changers would require.
 
There are better ones out there, much better but they are custom builds most of the time. -150c and colder, but they are multi-phased.
 
yep.

Technically , you could make a case that was in itself , a phase evaporator.

Sheesh I'd hate to see the compressor for that.

Almost a 'half-submersion' kinda setup.....closer to the 'spraycooling' cray toyed with...
 
Don't have a clue who cray is, but at least now I know one more person to stay away from.

I'm not sure you'd need a big compressor. Like I said, I'm going for extremely LOW performance (at least low for a phase changer.) I'm thinking the coolant doesn't even NEED to change phase to release enough heat to cool a heatsink to, say, 40 C ABOVE the freezing point. Not much pressurization may be involved, which is why I suggested electromagnetically propelling a plasma of some sorts. Or heck, just use a fan. I tiny one, insulated in a tube, so you can barely hear it. Any gas will do, whatever works, don't think traditional refridgerants. Those are designed for phase changing, which like I said, may not be necessary.

Traditional air and water coolers attempt to expose a CPU heatsink to a 20 degree Celsius environment as much as possible. Why not expose it to a 5 degree celsius environment? Less exposure necessary. And THAT'S the point.

I am dealing with this in very non-scientific fashion, so this all could be wishful thinking.
 
wow

I've heard of phase changing but never really knew what it was.

What kind of liquid to you put in there/refill? Are those hard to maintain? How come I don't see more people with them and only use water cooling? That looks a lot more effective and easier...

Do they make ones that cool the CPU, NB, and Video card?
 
colinstu said:
wow

I've heard of phase changing but never really knew what it was.

What kind of liquid to you put in there/refill? Are those hard to maintain? How come I don't see more people with them and only use water cooling? That looks a lot more effective and easier...

Do they make ones that cool the CPU, NB, and Video card?
Click to expand...

Uh they are really expensive, 90% of people do not need to go to -30C and colder. You really cant run the 24/7 there are a lot of risks involved. the use carious liquids from R22 and so on any gas that can be compressed to create a refrigeration effect. I know there are people on here who can explain it better but there are a few places on the web to find out more about them. As for the cooling of the cpu, NB, and GPU, not practical, maybe a CPU and GPU, but that would be a large system.
 
azuros said:
Uh they are really expensive, 90% of people do not need to go to -30C and colder. You really cant run the 24/7 there are a lot of risks involved. the use carious liquids from R22 and so on any gas that can be compressed to create a refrigeration effect. I know there are people on here who can explain it better but there are a few places on the web to find out more about them. As for the cooling of the cpu, NB, and GPU, not practical, maybe a CPU and GPU, but that would be a large system.
Click to expand...

I run a 24/7 water cooled / phase change setup. It is a high maintenance setup. Most of the time, I run at -34c on my QX6700 @ 3.905ghz. My NB, SB and both GPU's are watercooled. I use the DC4 compound for moisture control.

DSCN8979.jpg


DSCN8820.jpg
 
officermartinez said:
I run a 24/7 water cooled / phase change setup. It is a high maintenance setup. Most of the time, I run at -34c on my QX6700 @ 3.905ghz. My NB, SB and both GPU's are watercooled. I use the DC4 compound for moisture control.
This is more or less what i am going to be doing as well, my phase is in preproduction now.
Click to expand...
Click to expand...
 
eagle_orion said:
Don't have a clue who cray is, but at least now I know one more person to stay away from.

I'm not sure you'd need a big compressor. Like I said, I'm going for extremely LOW performance (at least low for a phase changer.) I'm thinking the coolant doesn't even NEED to change phase to release enough heat to cool a heatsink to, say, 40 C ABOVE the freezing point. Not much pressurization may be involved, which is why I suggested electromagnetically propelling a plasma of some sorts. Or heck, just use a fan. I tiny one, insulated in a tube, so you can barely hear it. Any gas will do, whatever works, don't think traditional refridgerants. Those are designed for phase changing, which like I said, may not be necessary.

Traditional air and water coolers attempt to expose a CPU heatsink to a 20 degree Celsius environment as much as possible. Why not expose it to a 5 degree celsius environment? Less exposure necessary. And THAT'S the point.

I am dealing with this in very non-scientific fashion, so this all could be wishful thinking.
Click to expand...

You do realize that a plasma is a 4th state of matter due to it's extremely high energy content.
http://en.wikipedia.org/wiki/Plasma_(physics)

That article lists a lot of reasons that you wouldn't want a plasma flowing through a CPU.
You are basically asking a florescent light to cool a CPU and I don't think that's going to happen.
 
rodsfree said:
You do realize that a plasma is a 4th state of matter due to it's extremely high energy content.
http://en.wikipedia.org/wiki/Plasma_(physics)

That article lists a lot of reasons that you wouldn't want a plasma flowing through a CPU.
You are basically asking a florescent light to cool a CPU and I don't think that's going to happen.
Click to expand...

I think this has kinda moved away from true "plasma," that term is just being used loosely at this point.
 
That's why I said "cold plasma," (if such a thing is possible, which I know it is at least in part.) Besides, (if you'll tolerate the ridiculous idea just a little bit longer,) you could ionize the gas at the heat pump with just a voltage. Then, as before with a changing magnetic field or just an electric charge of sorts, you could force the gas molecules into the radiating coils (or just plain tube network) before they lose their excited state. When they do de-ionize, all that extra energy should easily be released as light. Anything that still contributes to the heat of the gas from this process should not matter, as long as the gas is given enough time to cool to room temperature before it is depressurized.

I was thinking of a very SLOW moving circulation. That way the gas can be cooled passively. This may be a problem if a processor heatsink isn't exposed to enough coolant circulation. However, a secondary way to solve this problem other than increase circulation is to increase pressurization and cooling of the gas. A non-phase-changing setup may not work.
 
Forget about plasma.....that's hot not cold.

This does not mean ionic/metallic electromagnetic pumping won't work.

Liquid metal (Gallium-indium alloys) EM pumped cooling loops have been done .There are some issues , however they do work , and they are silent.The major drawback here is that the liquid metal coolant tends to dissolve every other metal part in the cooling loop.

As far as an EM pumped phase change unit ? I'm going to say not feasable (yet).

You would need a conductive or magnetic liquid , that was similar to current refridgerants as far as physical specs goes.No such cat exists to my knowledge.

Invent it.You'll be a rich man.

:D
 
Oh come on! That is way over simplified.

True visionaries aren't hampered by defeatist comments like, "plasma is hot," or, "there's this thing called gravity," or, "if it weren't illegal," or, "not if it explodes."

As a self-proclaimed entrepreneur of my own inspirations, I object to your opinions. Besides, it looks like the magnetic liquid research is already covered. Although it does make me wonder, how does one go about discovering these marvels? Our current knowledge of the chemistry and physics behind these phenomena is a little too limited to produce a true strategic approach to the issue. I'm more interested in discovering the nature of low temperature superconductivity among otherwise nonconductive materials. Maybe then we'll really begin to figure out what's going on at those sneaky microscopic levels.
 
OP, it is painfully obvious you don't have the background to even begin to understand what you're proposing. Go get a doctorate in Physics from a respected university (and maybe another in mechanical engineering while you're at it, or at least spend a few years working for a HVAC company) then come back and explain to us why this will never work. Or at least why it won't work currently. At the moment, you're just frustrating.
 
eagle_orion said:
Don't have a clue who cray is, but at least now I know one more person to stay away from.
Click to expand...

If you don't even know who Cray is, it makes me doubt you have even the general knowledge to know of what you are talking about or trying to implement. It's like saying you want to build a supercar but you don't know who Enzo Ferrari is. But then I guess you could be a chemical/mechanical engineer or just a really bad physics student and not know anything about computer history...

Hey you know what? I'll just hook-up my computer to the highest wattage TEC unit in the world but I'll do something innovative like inventing a cold fusion "(if such a thing is possible, which I know it is at least in part.)" powerplant to run it! In all seriousnous, if you invent this cold plasma thing, I'm certain you would make a lot more money implementing it in things other than cooling off-the-shelf computer parts...How is this even a sensible discussion? Applying fringe theoretical technology to current low-end mass market consumer goods is just kinda...ridiculous? It's like saying you have this brilliant idea to invent adult-diapers made of carbon nanotubes.

I don't mean to disparage you but as of a week and a half ago, you still haven't built your first computer. http://www.hardforum.com/showthread.php?t=1283238#post1032187355 From your post, it's obvious that you have zero experience with even putting together a home computer with off-the-shelf parts. Making a thread postulating ideas about how to cool it with theoretical technology just makes you look a little overzealously foolish.

It doesn't appear that you are either experienced in nor aware of the simple concepts of computer cooling (for example, you seem to be markedly worried about your CPU heatsink falling off which is almost a zero-probability event) or of the many great simple technologies (water cooling, vapour phase change, TECs, total immersion, even simple heatpipes) that already exist to cool computers and that big science solutions really aren't neccessary. If anything, for a consumer product, the simpler, more reliable, and more economical the better. Build a mechanical machine that pours dry ice onto the CPU and recycles the evaporating CO2 into fresh dry ice and recirculating it. That would probably cool the computer much better than your solution.
 
eagle_orion said:
Oh come on! That is way over simplified.

True visionaries aren't hampered by defeatist comments like, "plasma is hot," or, "there's this thing called gravity," or, "if it weren't illegal," or, "not if it explodes."

As a self-proclaimed entrepreneur of my own inspirations, I object to your opinions. Besides, it looks like the magnetic liquid research is already covered. Although it does make me wonder, how does one go about discovering these marvels? Our current knowledge of the chemistry and physics behind these phenomena is a little too limited to produce a true strategic approach to the issue. I'm more interested in discovering the nature of low temperature superconductivity among otherwise nonconductive materials. Maybe then we'll really begin to figure out what's going on at those sneaky microscopic levels.
Click to expand...

How's this for an opinion: at this time, it's pointless. It would be overly expensive, complicated and inefficient. It would require too much maintenance and for the user to be incredibly careful with the chemicals and materials involved.

"One" doesn't go about discovering it as it is incredibly expensive and complicated. Teams of research scientists and engineers working with grant money for major corporations figure this stuff out.
 
krameriffic said:
How's this for an opinion: at this time, it's pointless. It would be overly expensive, complicated and inefficient. It would require too much maintenance and for the user to be incredibly careful with the chemicals and materials involved.

"One" doesn't go about discovering it as it is incredibly expensive and complicated. Teams of research scientists and engineers working with grant money for major corporations figure this stuff out.
Click to expand...

Look at his posting history, he hasn't even built a computer before, heck he hasn't even installed a heatsink on a CPU before. He might know a thing or two of basic physics or mechanical engineering but he sure doesn't know how to apply it to a home PC :) Heck, he's worried (from his posting history) that if he buys a CPU cooler (his first heatsink/fan) that is too heavy that it might fall off!
 
Why is it that whenever I say something self-deprecating, people seem compelled to join in? A lot? A whole lot? A WHOLE, WHOLE lot? I should leave that tactic alone for now.

And FOR THE RECORD, I have heard of heavy heatsinks falling off under shock during transportation.

Yes, the idea is overcomplicated and fully impractical. I still find the discussion worth the edification... until I get crushed like a bug.
 
The problem with this, will always be heat dump.

Yes, it's possible that you could create a system that might cool magnetic fluids/gasses down to a certain temperature, but you would need to keep the fluid at said temperature, while it deals with absorbing the heat of a powerful component.

Additionally, moving heat requires energy, which means, exactly like a TEC, the amount of energy required to create this effect would be greater than the amount of heat it moves... Making it inherently inefficient... At least, to some degree.
 
eagle_orion said:
Why is it that whenever I say something self-deprecating, people seem compelled to join in? A lot? A whole lot? A WHOLE, WHOLE lot? I should leave that tactic alone for now.

And FOR THE RECORD, I have heard of heavy heatsinks falling off under shock during transportation.

Yes, the idea is overcomplicated and fully impractical. I still find the discussion worth the edification... until I get crushed like a bug.
Click to expand...


Any sign of weakness will be exploited! :p

Why don't you look into thermoacoustic refrigeration?
That's using sound to refrigerate a gas and to move heat. They've created systems that get down to like -30 or lower that basically are using a couple of speaker drivers.
 
rodsfree said:
Any sign of weakness will be exploited! :p
Click to expand...

Haha....Yep.

And no good deed goes unpunished.

:D

-------------------

Thermoacoustics is very interesting I will admit.

As far as I know , the cooling effect from this actually operates into the microwave range.....

This brings the conversation full circle.

You probably could build gas (non plasma) filled channels (read : waveguides ) into a motherboard , then a microwave backplane....hmmm

:D


-----

You are very correct to worry about heavy heatsinks.They can and do get knocked off.I run a very heavy copper heatsink , and I take my rigs to lans allatime , so I am very careful with this.I've seen several pics of sockets torn off mobo's because of a heavy heatsink and a bad pothole.

This didn't used to be a problem , as most heatsink makers stayed within weight spec limits.Nowadays this isnt true......
 
You must log in or register to reply here.
Back
Top