The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: "Bill Sloman"
Subject: Re: modeling a water cooling system
Date: Wed, 23 Oct 2002 00:25:51 +0200
Organization: Planet Internet
NNTP-Posting-Date: 22 Oct 2002 22:25:49 GMT
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
"Gary" wrote in message
> Mark Hathaway wrote:
> > Howdy people,
> > I'm in need of some help in modeling a water cooling system I'm mid
> > way through designing. The system will be used to cool four audio
> > amplifiers.
> > see here: http://www.passdiy.com/pdf/sonofzen.pdf
> > Each amp will have about 700 watts of heat to be dealt with. Water
> > cooling is the only way, ambient temperatures get too hot in
> > Australia for 1) comfortable room temperatures with ~3kw of heat and
> > 2)heatsinking to keep mosfets and resistors in their SOA assuming an
> > ambient temperature of 30-40 deg c is just out of the question, so
> > water cooling it is.
> > Where I'm stuck is data for how much heat a small car intercooler or
> > motor bike radiator could move assuming fan forced air with the
> > usual large computer power supply style fans.
> > Looking at a commercial heat exchanger example
> > http://www.thermacore.com/heatex_lahx.htm shows watt/deg c numbers
> > that I'd never imagined before. Where the data falls down is how
> > much air flow and water flow is required for such high watt/deg c
> > numbers.
> > Ideally I'd like to use a decommissioned car intercooler or two (or
> > motor bike radiator, would like to keep it aluminum). Anyone able to
> > point me to data or examples of achievable deg/c numbers for such a
> > heat exchanger?
> > Thanks for your effort,
> > Mark Hathaway
> Mark, water blocks for your system would have to bee customed made.
> Cold plates with peilters is another route you could test to cool the
> water blocks coolant other than just a radiator.
This is not helpful - Peltier junctions do transfer heat against a
temperature gradient, but at high cost - you seem to end up having to
dissipate about ten times as much heat at the exhaust side of the Peltier
as you suck out of the thing you are cooling, and in this application they
really have nothing to offer.
As for the heat exchanger, Farnell list some big aluminium extrusions that
are designed to be used with a fan, and you end up with thermal resistances
of a about 0,25C/W. The local thermal resistance for a single transistor on
the assembly would be about 1C/W, but if you can spread the heat around you
can get down to 0,25C/W and lower.
It isn't a cheap solution - think about $100 worth of extrusion, plus fan.
In order to get useful cooling out of a Pelteir junction we ended up dumping
the waste heat into a 150mm cube of heat-sink extrusions, coupled to a 150mm
fan (for the old-fashioned, 150mm is six inches) via circulating water.
After six months the water always got airlocks, so we went over to
transfering the heat via a heat-pipe.
Bill Sloman, Nijmegen
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup