[rescue] This Just In: HP to buy Compaq

[BSD Bob the old greybeard BSD freak]

> > "Linear Feet Per Minute"
> >
> > I really don't quite understand how to apply that to air flow though,
> > unless perhaps they mean "at the depth of the heat sink"....
> >
> Actually it's more appropriate than you might think, since calorie exchange
> from the heatsink is determined largely by the surface area of the object
> and the difference in temperature between it and the surrounding air, more
> air exchanges per minute mean better cooling (of course...).  Exchanges per
> minute is a function of air velocity and distance across the object...  Make
> a little more sense?  Ultimately, the more familiar CFM divided by enclosed
> volume still boils down to air velocity, (i.e. how much time does some
> discrete volume of air remain in contact with the surface before being
> replaced by cooler air)

My rule of thumb.....(I run a lot of things in electronics, e.g., radio and
computers by rulse de thumbio).... is that if it sounds like a jet engine,
then it probably has sufficient cooling to be reasonably effective.  Another
is that if you can lay your hand on it for 10 consecutive seconds, it most
likely has sufficient cooling.  The exceptions are running jet engines and
lit power vacuum tubes.

It is a function of both air velocity and dissipative cooling capacity.
In confined spaces, you need cooling air velocity.  In open spaces you
need sufficient area to dissipate the heat away from the device.  In,
for example blower cooled vacuum tube radio transmitters, the rule of
thumb is that sufficient cooling air is present if the tube will float
in the air stream above its socket.  The same tube, in open air will
run fine at up to 1/4 power with no cooling air at all, except for the
dissipative heat exchange.  Computers are a little different in that
they are mostly more confined.  So, air movement is very important.

I gotta admit, though, my alphas tend to run hotter than I would have
designed them to run, especially the little power regulator off to the
side of the cpu.  That thing runs normally hot enough to cook an egg.
The alpha manual says not to run the thing (3000/300) with the lid off
or the cpu will burn up.

Bob (ex KC-135 crew chief/GE turbofan rebuilder/radio operator.... a long
     time ago)