Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: john larkin <JL@gct.com>
Newsgroups: sci.electronics.design
Subject: Re: dumping a lot of heat
Date: Fri, 06 Dec 2024 20:53:17 -0800
Organization: A noiseless patient Spider
Lines: 114
Message-ID: <16i7lj5p3ssu6l065fccotcrp1oatk5err@4ax.com>
References: <lo51lj1da7a9ar0r9iavrcckuk00njsuoa@4ax.com> <1r434eh.9mfcivcsdztaN%liz@poppyrecords.invalid.invalid> <0ib3ljpfinh9tgso5j12ppagu9mkgiku7k@4ax.com> <1r43ukh.jujesq2i55aoN%liz@poppyrecords.invalid.invalid> <vitbvn$1ltub$1@dont-email.me> <1r44xn3.ng9dg1xoaiq2N%liz@poppyrecords.invalid.invalid> <sg16lj5atqibsg0b51pdgph71u2ha4ck6i@4ax.com> <1r45hb8.ryp44ae976g0N%liz@poppyrecords.invalid.invalid>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Injection-Date: Sat, 07 Dec 2024 05:53:21 +0100 (CET)
Injection-Info: dont-email.me; posting-host="d622e0c36653c930386b3c6397586c08";
	logging-data="3037896"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX19J60A4n+dJ2OU7BEPdcnKp"
User-Agent: ForteAgent/8.00.32.1272
Cancel-Lock: sha1:2XyOAFAIwEAHugnPsoVF0y+/PQ4=
Bytes: 6709

On Fri, 6 Dec 2024 17:40:38 +0000, liz@poppyrecords.invalid.invalid
(Liz Tuddenham) wrote:

>john larkin <JL@gct.com> wrote:
>
>> On Fri, 6 Dec 2024 08:29:33 +0000, liz@poppyrecords.invalid.invalid
>> (Liz Tuddenham) wrote:
>> 
>> >John R Walliker <jrwalliker@gmail.com> wrote:
>> >
>> >> On 05/12/2024 22:03, Liz Tuddenham wrote:
>> >> > john larkin <JL@gct.com> wrote:
>> >> > 
>> >> >> On Thu, 5 Dec 2024 08:55:32 +0000, liz@poppyrecords.invalid.invalid
>> >> >> (Liz Tuddenham) wrote:
>> >> >>
>> >> >>> john larkin <jl@glen--canyon.com> wrote:
>> >> >>>
>> >> >>>> I'm thinking about building a biggish rackmount dummy load box. It
>> >> >>>> would simulate series resistance and inductance. Part of the problem
>> >> >>>> is that it will need to dump a lot of heat.
>> >> >>>>
>> >> >>>> We are using copper CPU coolers on PC boards, which are great up to a
>> >> >>>> couple of hundred watts, but I'd like to do a kilowatt or two.
>> >> >>>>
>> >> >>>> https://highlandtechnology.com/Product/P945
>> >> >>>>
>> >> >>>> It would take a heap of expensive extruded heat sinks and fans to get
>> >> >>>> rid of a kilowatt. At 1 K/W, a pretty good heat sink, that's 1000 degC
>> >> >>>> temp rise.
>> >> >>>>
>> >> >>>> A small hair dryer can dump a kilowatt. So some sort of red-hot
>> >> >>>> nichrome coils and a vicious fan might work.
>> >> >>>>
>> >> >>>> I'd prefer to not use water.
>> >> >>>>
>> >> >>>> I wonder if there is some sort of runs-red-hot power resistor.
>> >> >>>
>> >> >>> If you are using elements at near red heat, remember you need to
>> >> >>> keep the radiant heat away from the outer walls of the cabinet.
>> >> >>> Reflectors just throw the problem elsewhere and eventually will
>> >> >>> tarnish, the best system is several spaced blackened steel baffle
>> >> >>> plates with vertical air passages between them (visual black is not
>> >> >>> always IR black).
>> >> >>
>> >> >> Seems to me that black baffles will absorb IR and get hot, so devolve
>> >> >> to air-cooled heat sinks.
>> >> > 
>> >> > Yes, all the energy is eventually going to finish up heating the air,
>> >> > the only question is the pathway it takes.  One way to avoid that would
>> >> > be to construct a massive infra-red searchlight beaming the energy away
>> >> > from the earth - or a broadcast transmitter beaming upwards.
>> >> > 
>> >> > To dump heat into the air, you either have to have something very
>> >> > conductive with a large surface area or you need another way of
>> >> > spreading the energy across a suface, such as heat radiation.  A big
>> >> > sheet of thin, blackened steel plate for heating by radiation is a lot
>> >> > cheaper than a thick die-cast aluminium lump with fins for heating by
>> >> > conduction.
>> >> > 
>> >> > The economics of mechanically-forced air cooling are better than
>> >> > convection unless you are able to use a tall 'chimney', so that the
>> >> > energy of the waste heat is used to generate the draught.
>> >> 
>> >> A full sized rack cabinet could be remarkably similar to a chimney
>> >> if it had baffles in the right places.
>> >
>> >Yes, as long as the ceiling of the room isn't lined with polystyrene
>> >tiles.  It might not be very good for the air conditioning system in the
>> >building where this is installed, regardless of which dissipation method
>> >is used.
>> >
>> >In the days of germainium transistors, one firm used to make cabinets
>> >with a 'clerestory' roof, like an upturned tray suppoted on spacing
>> >pillars above the ventilation holes in the roof proper.  This allowed
>> >the heated air to flow out under the lip in case some idiot put the
>> >instruction manual on the top of the cabinet.
>> >
>> >Another possibility, especially if there aren't going to be many of
>> >these on sale and the installation will be done by the firm that makes
>> >them, is to make a hole in the wall and stick some stainless-steel
>> >boiler flues up the outside of the building.  Even better, in an old
>> >building. use a redundant fireplace and put the contol box in a
>> >decorative housing on the mantlepiece.
>> 
>> We have in mind some commercial rackmount products, 1U to maybe 5U
>> dummy loads that people would buy and bolt into their 19" racks like
>> any other instrument. They would blow hot air out the back, like most
>> other gear.
>
>
>You are going to need an awful lot of air if the exit temperature isn't
>to be too high.  Ordinary domestic fan heaters seem to be the best
>starting point and some models would fit comfortably as a side-by-side
>pair into a 5U 19" rack dissipating about 3kW each.  Any more than about
>1kW per Unit height is going to produce a dangerous temperature rise or
>need a noisy high-speed blower to cool it.

Yeah, about a kilowatt per U sounds about right.

I was just running a Dynatron copper CPU cooler at about 300 watts,
and the exit air was painful. I might spec 250 on that one, which
would be uncomfortable but not dangerous.

https://www.dropbox.com/scl/fi/bbotoirgcm3g5bsgo4b9i/20241206_145219.jpg?rlkey=yvvs6y09ggzz8h78qv980iitt&raw=1

Incidentally, cooler specs seem to assume specific CPU power
dissipation which correlates weakly to characterizing the cooler as a
general heat sink.

One could put a bunch of kilowatt-class mosfets on a couple of those
coolers and make a monster class-D amp.