Page 3 of 3

Posted: Sun Sep 24, 2006 10:57 am
by Cinquero
Tibors wrote:If it is solid, then it is not a heat PIPE. Then it is just a copper rod.
Are you sure that vendors care for a strict definition? I'm not. A pipe is a pipe. A copper pipe with solid copper in it is still a pipe. :-)

Posted: Sun Sep 24, 2006 11:40 am
by Devonavar
Yes. Vendors may not care about the strict definition, but they certainly care whether their product works effectively. There is absolutely no point to building a heatsink with solid heatpipes; it would push up cost needlessly without providing any performance gain. Heatpipes are hollow, and use phase change to move heat. That's the definition of a heatpipe. They're all like that.

Posted: Mon Sep 25, 2006 7:17 am
by rjm
We're getting off topic but no a heatpipe isn't a heatpipe unless its hollow and filled with a volatile liquid. Someone once drilled a hole in one to deliberately leak the fluid out (just to check, y'know) and his heatsink promptly became useless. Copper may have good thermal conductivity but its still nothing in comparison with a functional heatpipe.

Back to the ionizer, the negatives keep piling up:

- the "black wall" effect, deposition of charged particles on surfaces
- corrosion from ozone byproducts (its also moderately unhealthy for people too in high concentrations)
- I can also add that corona discharge can be a nasty source of RFI noise

Some advantage would be gained to mounting the ionizer such that it expelled air from the case. That way all the ions and ozone are ejected directly into the room rather than passing over the motherboard and other components.

You could also put a magnet on the end of the tube to trap the charged dirt particles.

But hey, suddenly a 120mm case fan or two isn't doesnt seem such a bad arrangement after all!

Posted: Mon Sep 25, 2006 11:13 am
by 2Lurk
Lets try to keep this thread on topic.

Negatives:
1) Low CFM
2) essentially no ability to resist back pressure
3) grime (black wall effect)
4) need for and noise (whine) of a separate HV power supply
5) oxidation (corrosion)
6) static charges
7) Negative health effects of ozone byproduct
8) RFI noise

Positives:
1) Non mechanical
2) No noise at air movement location (HVPS can be located further away)

The CFM, voltage, and ozone production I believe are positively correlated. I am guessing if I try to design an Electro Hydro Dynamics (EHD) "fan" with decent CFM that I will exceed the 50-ppb industry standard limit of ozone production. Therefore I don't think we will see commercial implementation of a fan substitute, since companies will probably be shy of the potential lawsuits regarding some nut's health.

But I do think the potential is there and would encourage someone to experiment. If the "fan" were used to only expel air from the case as others have suggested the grime and corrosion effects would be reduced.

Posted: Mon Sep 25, 2006 11:32 am
by Cinquero
How about combining the acceleration of ions with heat pipes? I guess that the radiator part of a heat pipe is limited in size because of the passive heat transport. Is that correct?

If yes, could we possibly make heat pipes even more efficient by accelerating the internal gas transport?

Posted: Mon Sep 25, 2006 12:03 pm
by stevea
Cinquero wrote:How about combining the acceleration of ions with heat pipes? I guess that the radiator part of a heat pipe is limited in size because of the passive heat transport. Is that correct?

If yes, could we possibly make heat pipes even more efficient by accelerating the internal gas transport?
Pehaps slighly, but I think you're missing the point. The internal fluid of the heatpipe boils at one end and recondenses at the other "cool end". These phase changes from liquid to gasseous and back will typically require a lot of calories for a very small change in temperature. Increasing the convection in the heat pipe isn't a huge increase in efficiency so long as these are properly mounted.


Oh yeah - to that guy who thinks pentane and hexane are appropriate coolants - what are you smoking ? Hopefully nothing since these are both extremely flammable.

Posted: Mon Sep 25, 2006 12:13 pm
by jaganath
Oh yeah - to that guy who thinks pentane and hexane are appropriate coolants - what are you smoking ? Hopefully nothing since these are both extremely flammable.
The quantity of working fluid in a heatpipe is truly miniscule, you are more in danger of electrocuting yourself on the power supply than setting yourself alight with the gas from the heatpipe.

Posted: Mon Sep 25, 2006 12:16 pm
by Cinquero
stevea wrote:Pehaps slighly, but I think you're missing the point. The internal fluid of the heatpipe boils at one end and recondenses at the other "cool end". These phase changes from liquid to gasseous and back will typically require a lot of calories for a very small change in temperature.
From a theoretical point of view: NO change in temperature. :-)
Increasing the convection in the heat pipe isn't a huge increase in efficiency so long as these are properly mounted.
Are you really sure about that? Hmmm... how does one calculate the mean free path for gas particles again?

Posted: Mon Sep 25, 2006 6:58 pm
by stevea
Cinquero wrote:From a theoretical point of view: NO change in temperature. :-)
Yeah a slight change in temp for a non-ideal gas, but it's mostly entropy.
Increasing the convection in the heat pipe isn't a huge increase in efficiency so long as these are properly mounted.
Are you really sure about that? Hmmm... how does one calculate the mean free path for gas particles again? [
Yes - I an extremely sure about that. The mean-free-path(MFP) has little to do with gas conductivity. Instead the only factors that you can realistically tweak are to choose gas molecules with less mass, smaller radii and operate the gas at a higher temp (which defeats the whole purpose).

Now I know someone will note that gas thermal conductivity is proportional to (n * <v> * MFP), where n is the molecular density and v the mean velocity, *BUT* if you factor out the 'n' and the 'MFP' you just get a relation where conductivity varies in relation to sqrt(T) [square root of the absolute temp]. This fact rather surprised me when I was desigining some equipment for very high altitude (65k feet+) use 2 years ago.

I see that modern heat pipes use a "wick" or scintered material to carry the fluid phase, while older models used gravity to drive the liquid to the hot end. Either way - the improvment in heat-pipe conductivity does not help in cooling a CPU.

I have read that cheap heat pipes have ~80 times the conductivity of a similar copper volume, and very well built ones perhaps 1000 times. This really doesn't help us. It's like saying the 0.5 watts lost in a circuit breaker could be halved if we use two in parallel - it's true but it is not a useful gain.

Heat pipes can help us transport the heat far more (say 80 times more) effectively than copper, which is already pretty good. We ultimately must transfer the heat to the environment, primarily air and that's the real problem here. (more to follow).

Posted: Tue Sep 26, 2006 7:39 am
by aPackOfWankers
The solution is perfectly obvious.

What you need to do is remove all of the sidewalls from the PC and suspend it in a pendulum arrangement. As it silently swings backwards and forwards, it moves through fresh air, thus cooling it. A small linear motor under the pendulum gives the PC a small boost each time it passes the nadir.

A more advanced version would have the PC silently swinging in a circle.

Requires a fair amount of space, but is basically silent.

Posted: Tue Sep 26, 2006 7:43 am
by qviri
I'm not sure I want my hard drives and optical drives spinning around... That is an interesting idea, though.

Posted: Tue Sep 26, 2006 7:45 am
by Cinquero
Yes. And I'm charging entry fees to my new and personal modern arts museum :-))).

Let me think what that work should express... hmmm... hacker insanity?

Problem: I'd guess that that solution will create A LOT OF turbulences :-(.

Posted: Tue Sep 26, 2006 8:43 am
by Rusty075
aPackOfWankers wrote:The solution is perfectly obvious.

What you need to do is remove all of the sidewalls from the PC and suspend it in a pendulum arrangement. As it silently swings backwards and forwards, it moves through fresh air, thus cooling it.
:lol:

Grandfather clock + mini-itx mobo = brilliant!

Posted: Tue Sep 26, 2006 1:52 pm
by ijgs
aPackOfWankers wrote:The solution is perfectly obvious.

What you need to do is remove all of the sidewalls from the PC and suspend it in a pendulum arrangement. As it silently swings backwards and forwards, it moves through fresh air, thus cooling it.
hahaha! sweet!

Another alternative (whilst we're on the subject of destroying perfectly functioning appliances for non suitable applications) is to remove the lid from a hard drive and carefully mount a 12cm fan blade onto the hub - thus moving air around the case without introducing additional moving parts, cooling the hard drive, and subscribing to "extreme modders-r-us" all in one hit. Here's a picture of one I made earlier...

Image

Ionic cooling..? tchah!

on a separate note - my word this forum's got some technical know how to it! it's a pleasure to see!!

all the best,

Ian

Posted: Tue Sep 26, 2006 2:00 pm
by Cinquero
Maybe we should "just" find the optimal fan geometry that produces no audible turbulences. Would be a nifty task for a distributed computing project: let the geometry evolve by using some sort of GA (genetic algorithm) and do finite element calculations to determine the noise (=fitness)...

Posted: Tue Sep 26, 2006 8:25 pm
by Ben C.
My sister's Ionic Breeze was making a continuous crackling sound of about 20dB at one meter. You couldn't hear it from far off. But this could be an issue for someone looking for total silence in a computer right next to them.

Posted: Thu Sep 28, 2006 11:19 am
by colm
I read 2.5 pages of threads and.... I don't quite understand. :?
The goal worked for materials given. That is all that mattered. the numbers would be different, if there is no actual CFM of air flow. To make them up like the builder did was ridiculous however.
Big heatsink - little fan.
Decent natural flow and dissipation doesn't need much to meet the goal. I liked it, except, a fan at 1000 rpm keeping a hot cpu cool is very very good (my current pc for over 6 months now). Especially at 12v, low amp, non-dust collecting, and keeps moving through all air environments. I am most certainly quiet enough. Zero noise accepts external noise, which is interference. Good idea , but a bit of overthough and not enough facts for long run.

Posted: Sun Oct 01, 2006 3:20 am
by shoebox9
My sister's Ionic Breeze was making a continuous crackling sound of about 20dB at one meter.
20db @ 1m. Has it been cleaned recently?

I really like the sound of a pipe based ion wind generator.

A couple of pipes in parallel could run vertically out of a case (ie exhausting heat), and would (could be designed to) also benefit slightly from the chimney effect.

Air would be gently pulled into the case via suction, be "pumped" out the charged chimney pipes, and charged particles would grime up your metal desk legs instead of your PC internals. Ozone would exit the chimney, but not freely circulate inside the case.

I think there is a lot of potential here, but it sounds like the best design would to be based around exhaust air rather than intake air.

Posted: Sun Oct 01, 2006 1:29 pm
by loz
I agree.
It seems a good solution for a static box, with a chemney.
I would add a grounded piece of metal (easy to clean) at the end of the chemney to attract ionized dust.

Posted: Sun Oct 01, 2006 4:22 pm
by shoebox9
Yes, maybe the last 3" (or whatever) could be a grounded chimney extension (ie use an insulating rubber gasket)- both increasing the chimney "draw" effect due to extra length, and collecting the worst of the grime.

Posted: Sun Oct 01, 2006 10:27 pm
by Ben C.
Your right. My sister's Ionic Breeze probably needs cleaning.

20dB at one meter was just a mental guess based on my regular use of a Bruel & Kjaer type 2203 sound meter. I purchased it just for sound testing computer components.

Apple iMacs show that you can cool with convection alone. But I like the idea of the chimney effect speeding that up. I'm thinking of routing heat from the CPU to three large verticle copper pipes. Between the CPU and the three copper pipes are multiple heat pipes which spread the heat evenly along the large verticle copper pipes. This would have good surface area and little restriction to the convection air flow. The question is how many watts or BTUs will this system move?

Posted: Mon Oct 02, 2006 12:37 am
by jaganath
Apple iMacs show that you can cool with convection alone.
Interesting; which iMacs don't use any kind of forced cooling (ie no fans, no pumps)?
The question is how many watts or BTUs will this system move?
This thread has the relevant equations.

PS. all ionisers make a hissing/crackling sound, it is caused by the corona discharge.

Posted: Mon Oct 02, 2006 4:18 am
by qviri
jaganath wrote:
Apple iMacs show that you can cool with convection alone.
Interesting; which iMacs don't use any kind of forced cooling (ie no fans, no pumps)?
The G4 Cube did...

Posted: Tue Oct 03, 2006 6:07 pm
by Ben C.
Thanks for the thread jaganath.

I just bought my sister's old iMac because it has no fans. It's a iMac (Summer 2001) G3, 600-MHz. Although I have seen other model iMacs with fans in them.

This G3 looks like a television set. Besides a CRT monitor buzz which you would only hear if you put your ear up to it, the only noise coming from it is a hard drive.

Posted: Wed Sep 16, 2009 6:09 am
by princesnicole
Very absorbing DIY project, but I alarm BS on the CFM claim. Firstly he never tells you how he abstinent it (or whether he artlessly apprehend it off the packaging of the ion generator), secondly even automated admirers active off 24V don't aftermath that affectionate of airflow. Thirdly if there absolutely was 325CFM traveling through his rig there would be lots of turbulence babble etc, so not 0dB. I accept no agnosticism that the technology works, so there's no charge to amplify the CFM.

_________________
Aprilaire