Just and idea - external cooling.

[Alan Walker]

Hi,

Just an idea, but if I were to put a PC in a small fridge or beer cooler to
improve cooling can anybody see any obvious problems ?

The PC would be warmer then the surrounding air so I'd not expect
condensation.

Would I burn out a standard domestic fridge motor trying to move this much
heat ?

Problems sealing the door round the leads I can deal with.


Alan Walker

 

[J.B.]

Hi,

Just an idea, but if I were to put a PC in a small fridge or beer cooler to
improve cooling can anybody see any obvious problems ?

The PC would be warmer then the surrounding air so I'd not expect
condensation.

Would I burn out a standard domestic fridge motor trying to move this much
heat ?

Problems sealing the door round the leads I can deal with.


Alan Walker

Your problem would be condence in your pc.

 

[Alan Walker]

In

Your problem would be condence in your pc.

I thought about that, surely as the hardware would be warmer than the
surrounding air there shouldn't be any condensation ?

I suppose I could always put some dessicant in with it, as a sealed space
that should take out any water.

Alan Walker

 

[Replicant]

I thought about that, surely as the hardware would be warmer than the
surrounding air there shouldn't be any condensation ?

Isn't it the warmer air warming the cooler air that causes
condensation? You will get condesnation on the outside of the case but
maybe not on the inside. It's not something I would risk though. They
do make refigeration coolers for cpu's so maybe it is doable.

 

[D.P.]

Isn't it the warmer air warming the cooler air that causes
condensation? You will get condesnation on the outside of the case but
maybe not on the inside. It's not something I would risk though. They
do make refigeration coolers for cpu's so maybe it is doable.

Cold sufaces in warmer air cause liquid condesation.

 

[Alan Walker]

In

Cold sufaces in warmer air cause liquid condesation.

That's what I thought, so as my machine will be generating heat and the
surrounding air will be cooler it should be ok.

I'm starting to think maybe I should run an old 486 in there for a couple of
weeks before trusting it to anything expensive.

 

[DaveW]

Serious? Well, read the spec sheets for the various components in your
computer, starting with the harddrive, to see what ranges of 'ambient air
temperature' they were designed to work with.

Not!

 

[GTD]

I'm starting to think maybe I should run an old 486 in there for a couple of
weeks before trusting it to anything expensive.

I'd give it a try if I were you, especially if you dont care too much
about the 486. Might want to check the operating temp range for the
mechanical stuff: harddrives, cdrom, ect.

 

[1bigjay]

I think condensation would still be an issue. Try putting something hot in
the fridge and see what happens.

 

[Todd Howard]

It has been done....like 6 years ago according to the first link
http://www.cnn.com/TECH/computing/9809/30/japan.internet.fridge/
http://www.ohardt.com/computer/pc.mods/project.hot.fridge/

TH

 

[Gary Tait]

Whereas On Fri, 10 Oct 2003 23:50:50 +0100, "Alan Walker"

In

That's what I thought, so as my machine will be generating heat and the
surrounding air will be cooler it should be ok.

I'm starting to think maybe I should run an old 486 in there for a couple of
weeks before trusting it to anything expensive.

but when you turn youe computer off, it will eventually condense on
it.

 

[Alan Walker]

In

Whereas On Fri, 10 Oct 2003 23:50:50 +0100, "Alan Walker"


but when you turn youe computer off, it will eventually condense on
it.

I'm planning on running a machine 24/7

 

[Replicant]

Cold sufaces in warmer air cause liquid condesation.

OK, thx for the correction.

 

[patrickp]

It has been done....like 6 years ago according to the first link
http://www.cnn.com/TECH/computing/9809/30/japan.internet.fridge/
http://www.ohardt.com/computer/pc.mods/project.hot.fridge/

TH

And in more entertaining fashion http://totl.net/Eunuch/index.html.

patrickp

 

[Dorothy Bradbury]

You are missing the problem here:
o You can run a PC at very low temperatures
---- the critical factor is lubrication on motor bearings
---- on quality fans & motors (hard drives) you are ok
---- on cheapy fans you may find the bearings run at <6 months
o A PC emits 200-220W of heat
---- most fridge/freezer compressors are a similar rating
---- so you are not magically getting the freezer to stay at -18 to -22oC

Specifically that hints at the real problem with this:
o Your freezer/fridge is designed to cool it down once
o Then cycle on-&-off to keep it cool re imperfect insulation

It is not designed for a continuous thermal input inside it, such
that the compressor cycles far more often & even runs continuously.
This will drastically reduce the life of the compressor - months.

The proper refridgerated enclosure systems have to be:
1) sized for the application re thermal load -- so cycling occurs & in spec
2) designed for heavier duty cooling -- re heat-source not a frozen turkey

It's better to use a Peltier cooler for this, but ideally they should be run
off a non-PC PSU - plenty of 12V & 24V industrial SMPS on Ebay.
They have 1,000,000 & even 5,000,000 hr MTBF, very reliable items.

The issue using a Peltier cooler is condensation on the cold plate, ie, the
insulation of the cold plate from the environment such that no part of the
insulation is below the dew point (ie, condensation occurs on it).

Water is the enemy of epoxy boards & processors, so you know.
The voltage of PCs is not too bothered about running water on it.

Industrial PC applications (eg, SBCs in traffic light boxes) actually use
heaters in extreme conditions to keep the air inside above the dew point.
All major industrial/telco/server fan farms similarly actually shut the fans
off at 0oC and can be set to turn on heaters at -5oC re condensation.

Most epoxy's used in the PC industry re PCBs are not waterproof.
They will degrade slowly, but degrade they definately will over time.
This is true even of deionised water (eg, out of AC/dehumidifier).

Cheaper to do it by Pelter than a fridge/freezer, altho less easy

Thanks.

 

[Kevin Lawton]

<snip>
| It's better to use a Peltier cooler for this, but ideally they should
| be run off a non-PC PSU - plenty of 12V & 24V industrial SMPS on Ebay.
| They have 1,000,000 & even 5,000,000 hr MTBF, very reliable items.
|
| The issue using a Peltier cooler is condensation on the cold plate,
| ie, the insulation of the cold plate from the environment such that
| no part of the insulation is below the dew point (ie, condensation
| occurs on it).

True, true.
Anyone got a good retail source of Peltier coolers ?
Kevin.