Mineral Oil Cooled Computer

  • Thread starter Thread starter Grinder
  • Start date Start date
G

Grinder

Nothing new, but I thought the execution was sort of tidy. You'll
probably want to turn your speakers DOWN before clicking the link.

 
Don't you hate it when you have your system set at a good
volume and someone decides flash should be at full
volume?!!!!!!!!!!!!!!

The problems seems to be the same as always, that they chose
very low heat hardware (since when is a socket 7 system hard
to cool in audibly?). Granted, they tried newer hardware
too, but when I can do that with stock parts, why would
anyone want to clean oil off to do basic parts swaps? It's
still an interesting video, but I for one will stick with
air-cooling and highest o'c at lowest cost if heat becomes a
factor.

The tank of oil has a large heat capacity, but there wasn't
anything cooling it other than convection from the tank sides.
Whilst it might take a long time to warm up, it would be
interesting to know what the final stable temperature is when
heat loss from convection from the tank sides equals heat
input.

Also, some of the finer heat sinks (such as the thin fin spacing
on CPU heatsinks) will be far too close for something this viscous,
and there's a danger of no flow at all through them. With the
heat capacity of oil being much higher than air, only a relatively
small flow is required, but there's a danger of none at all, which
would turn your chips to fries.

I notice the sequences with the tank bubbles were speeded up
considerably, presumably due to the viscosity of the oil
making the bubbles too slow otherwise.
 
It was light oil. that stuff is not much thicker than water.
If you combine this setup with a radiator plumbed into a small fridge it may
let you do some safe overclocking st speeds that normaly cause trouble.
 
Grinder said:
Nothing new, but I thought the execution was sort of tidy. You'll
probably want to turn your speakers DOWN before clicking the link.


Using oil for this seems to be a bad choice. There are coolants such as
FC-770 with which one can flood a computer with none of the downsides of
oil. Can you imagine the mess involved with an oil-soaked computer when you
need to change a card or perform some other simple maintenance? On the
other hand 3M's Flourinert coolants are completely inert and will not
interfere with contacts. It is also completely non-viscous and will flow
into places where oil cannot go and you can even solder on a circuit
immediately after removing it from the liquid. There _is_ a reason that
Cray chose this cooling method for their hyper-hot ECL circuitry.
 
John said:
Using oil for this seems to be a bad choice. There are coolants such as
FC-770 with which one can flood a computer with none of the downsides of
oil. Can you imagine the mess involved with an oil-soaked computer when
you need to change a card or perform some other simple maintenance? On
the other hand 3M's Flourinert coolants are completely inert and will
not interfere with contacts. It is also completely non-viscous and will
flow into places where oil cannot go and you can even solder on a
circuit immediately after removing it from the liquid. There _is_ a
reason that Cray chose this cooling method for their hyper-hot ECL
circuitry.

The stuff is pretty pricey, isn't it?
 
Grinder said:
The stuff is pretty pricey, isn't it?

Yes. It is. There was a less expensive alternative, < 1/4 the price of
FC-770, a few years ago but I haven't done any looking lately.
 
Dosn't airconditioner gas get water in it?

John McGaw said:
Yes. It is. There was a less expensive alternative, < 1/4 the price of
FC-770, a few years ago but I haven't done any looking lately.
 
Back
Top