Liquid cooling heatsink design

Well, yeah, you couuld model a bucket of ice as a Zener regulator, and an 85W CPU as a current source, and a pumped water flow as a thermal resistor. It wouldn't be a great model, though.

I like JL's idea of two jugs and gravity flow (just swap low-for-high on a stepped jug stand to recycle). But, it takes tending.

Alternately, use a cooler chest of ice with a pump, for long runs. At 85W, in eight hours you'd melt 7.3 kg of ice. If a pump is quiet enough, a cooler chest is more than adequate.

Why not put the whole computer PCB into deionized water ?

i water is used e.g. in HVDC stacks and it seems that Ruskies are running some of their supercomputers submerged into such water.

Laptops that run a performance-oriented processor like an i5 or i7 sound like jet engines when the active-thermal system spools up to cool the processor, as it exits idle/web browsing load and starts doing real work. Whoooooossshhhhh if the house is quiet I can hear mine blowing from two rooms away

Correction: DI water is used ...

Nobody pays for vats of water

Have you looked at

Nonsense.

f a vapor.

moving heat as latent heat of vapourisation/condensation, you move a lot m ore heat per unit mass.

member being impressed by it whistling through my vacuum line when I was do ing vacuum distillation.

t transfer fluid went in - heat pipes had a reputation for being a bit anem ic at low temperature differences, but if you test them at goods inward you can send back the one's with any uncondensable gas (as was done with the h eat-pipes referred to in my 1996 paper).

lant extracting the heat, you have to move more volume.

tion, so you need to move a lot less mass to shift the same amount of heat.

s) per gram. If you are shifting heat with circulating water, and can affor d 10 degrees Kelvin difference between the device being cooled and the fins radiating the heat, you move 10 calories per gram - rather less.

And the viscosity about two orders of magnitude less - water is 8.9x10^-4 P a.s and steam 12cx10^-6 Pa.s.

Steam moves a lot faster in a heat pipe than water does. The condensed wate r wicks back relatively slowly, but heat pipe set-ups don't seem to have bi gger pipes than water-cooled systems - it isn't as if the pipe-work takes u p much of the space in either.

You are right. I had a major brain malfunction.

Dan

rculating water, after about six months of continuous operation. There may be a well-known solution to this, but the heat pipe was the easiest we foun d, and it let us get rid of the water pump. Convection driven systems can w ork, but you need enough vertical height for the mass of the column of thin ner, warmer water to generate enough pressure to drive enough circulation. By m

Just a guess, but I think the pump had a seal on the shaft of the motor whi ch wore and let air into the system.

Dan

It's even worse when the intake/exhaust ports get dirty or the fan bearings start to wear:

Then stop buying shit!

circulating water, after about six months of continuous operation. There ma y be a well-known solution to this, but the heat pipe was the easiest we fo und, and it let us get rid of the water pump. Convection driven systems can work, but you need enough vertical height for the mass of the column of th inner, warmer water to generate enough pressure to drive enough circulation .

hich wore and let air into the system.

I have no idea of the technology used in the circulating pump.

You can certainly use a rotating magnetic field to drive the impeller in a fully sealed - glandless - system. European central heating systems that re ly on circulating water have relied on them for as long as I've been expose d to them (which is since about 1980).

Fisons Applied Sensor Technology - which had become Affinity Sensors by the time the paper got published - had a tame mechanical engineer, and the dev elopment boss - Jim Molloy, now sadly deceased - was a real expert on flui d handling systems. In other circumstances I might have taken an interest, but in that situation I would have been trying to teach my grandmothers to suck eggs.

Not your first, or last.

The MSI GS40 in the second link is a favorably-reviewed mid-range performance/gaming laptop that costs well over a grand, even Obama wouldn't buy me one

That's kind of vague there, but I can't say how efficient their X86 processors actually are.

That is to say more specifically the low-power variant processors of the x86 architecture line don't meet the minimum system requirements as specified by the manufacturer of the software they wanna run.

Which has as much relevance as anything else you've had to say. ...about anything.

Krw has a touching faith in his own capacity to decide on "relevance".

In practice, he can only judge whether a post agree with his own opinions, or fails to, and while all of us can agree that krw is insane, we are less interested in whether other people's insanities line up with krw's.