How it's made: heat sinks

afair copper is almost twice as good a aluminium but expensive.

most of the benefits of copper is probably already had by just using a copper heat spreader to get the from the small cpu to the larger heatsink and then the rest of the way aluminium work just as good

I don't think copper is not strong enough for a heat sink, but it's freaking heavy compared to aluminum. You can only bolt so much weight onto these boards before it causes problems. The copper base is great for spreading the heat out to the fins, but the fins don't need to be so highly conductive to carry their portion of the heat and couple it to the air.

snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:

That depends on what it is made from and what surface treatment it got.

I made IR calibration ovens decades ago, and now they make calibration surfaces where the temp has to be homogenous across an entire surface, which is very difficult to achieve due to local air currents and other factors.

Not sure if you have a full grasp of just what a black body radiator is. Maybe you only have a fool grasp.

"Commander Kinsey" snipped-for-privacy@nospam.com wrote in news: snipped-for-privacy@ryzen.lan:

Copper "work hardens" and such a bend would create micro-fractures right at the bend joint.

To my knowledge Aluminum does not, but that may not be true of all of the alloys in that metal's family.

However, these thoings are done by engineers for customers, so I am sure that efficay was examined, determined and compared to other means and it would not exist at all were it not able to produce an effective heat sink.

I would be interesting to be able to take measurements of the temperature profile across a fin in a heatsink. You can do finite element analysis once you make your assumptions, but to get past the unknowns, measurements would be useful. I suppose all that really matters is the base temperature given an air source and air temperature. That delta T is all that is really important. But knowing the temperature distribution on the fin could help design better fins, perhaps. With the pressed in fins, it could be practical to alter the width of the fin from top to bottom. But I suppose there's not much point really. Just make the fins fat enough to carry as much heat as needed. I wonder if they have an equation for fin width/spacing vs. the number of fins, to identify the optimum?

Or just go with water cooling. I've always liked that idea. I was going to build a water cooling system from scratch for a desktop. I ended up with a laptop before I got it ready for test. They already use heat pipes. It's hard to beat that. They have really tiny radiators with very, very thin fins, but lots of them. I wonder how they get any air to flow through them??? It must be magic.

torsdag den 21. april 2022 kl. 00.12.14 UTC+2 skrev snipped-for-privacy@decadence.org:

only if you try to bend it back, annealed copper is dead soft only gets hard once you have bend it

"Commander Kinsey" snipped-for-privacy@nospam.com wrote in news: snipped-for-privacy@ryzen.lan:

Let me correct the f*ck out of your utter stupidity.

Water cooling takes internal case air and passes it through the radiator OUTSIDE the case, and no, the pumps do not fail and they do not have leaking problems. Maybe you are just so goddamned stupid and dismissive that you failed to note that the problems with the early water based "systems" were overcome over a decade ago.

Clifford Heath snipped-for-privacy@please.net wrote in news:16e723e88710fdef$7$1100196$ snipped-for-privacy@news.thecubenet.com:

I agree. They are likely considering the air flow available as being the defining (restricting) element. A wavy design would likely need more flow to gain motion at the boundary layer, but should be far more efficient once achieved. I would even like to see the numbers on a system that "injected" air at the very base of the sink fin array with small jets in front of each one, and also, of course, normal air flow systems to carry it all off.

Clifford Heath snipped-for-privacy@please.net wrote in news:16e7243ae6ab4d09$1$1100196$ snipped-for-privacy@news.thecubenet.com:

And then they banned freon. I am sure the refrigerants now used would be good too though.

The best thing is that transistor size reduction (node size) has made power consumption drop drastically as transistor count and fucntion increases.

This is why I love electronics so much compared to the rest of the world's induxtries. It gets better and beter as it gets less and less expensive.

The post I made about that science girl is one where she goes over the new NVidia units and they have BILLIONS of elements each.

It is literally amazing to me how many logic toggles get made these days without errors. We used to worry about how many vias a PCB layout had as points of failure. Now even PCB manufacturing is amazing, as is the level of quality.

Even the quick turn short run guys are pretty darn good.

torsdag den 21. april 2022 kl. 01.27.18 UTC+2 skrev snipped-for-privacy@decadence.org:

Freon is a brandname, there is many different types

Lasse Langwadt Christensen snipped-for-privacy@fonz.dk wrote in news: snipped-for-privacy@googlegroups.com:

I guess you are not very observant. The bend happens at the end of each skiving, and the speed of the bending is what causes the IMMEDIATE 'work hardening' and no, "bending it back" is not required. The solid matrix of copper atoms shears and micro-fractures occur immediately. Regardless of the fact that you are too inexperienced to understand the process. Your precious "full annealing" disappears right at the bend and does so immediately. Copper is one of the worst mediums for it too.

The military are fully versed in these properties. They also know about things like gold inter-metallic embrittlement.

Copper bends are properly made in a very slow motion. I have used large copper bars to make grounding system, etc. And slow bending is a priority. Very slow bending.

torsdag den 21. april 2022 kl. 01.48.35 UTC+2 skrev snipped-for-privacy@decadence.org:

good thing we don't use copper pipes or wires then ...

Lasse Langwadt Christensen snipped-for-privacy@fonz.dk wrote in news: snipped-for-privacy@googlegroups.com:

I used to make environmental chambers. The company dumped a hundred pounds a day before the ban. I know what a Chemours brand name is and what the industry uses as a blanket term as well. We made table top dual stage chambers that could go to -150C and chambers big enough to place an entire missile into and chambers that had both hot and cold sides and a transition door between to thermal shock test PCB assemblies under power. I really don't need a primer on refrigerants or the transition to the new ones. Which they refer to as "Puron" these days. But thanks.

They do copper sinks this way also:

Here's a solid copper server cooler shown next to a light-weight desktop bi-metallic:

Danke,

Lasse Langwadt Christensen snipped-for-privacy@fonz.dk wrote in news: snipped-for-privacy@googlegroups.com:

They too fracture and that is why copper pipe runs for plumbing are done with straight sections and then "elbows" and such are used to make turns. Soft copper comes in rolls but get straightened for installation and many runs And does not lend to being bent much as the diameter collapses and flow suffers.

Wiring is similar inasmuch as nobody makes hard turns in wiring runs, and the little loops made at the ends for termination should be made in a slow fashion to minimize the effect of 'work hardening'. It is in fact taught in a proper electrical wiring / electrician's class.

Sarcasm can be funny, but not so bright sarcasm, not so much.

copper is a much better heat conductor than Al. It also costs more and weighs more. In the processor world, you have to transfer the heat from the relatively small surface of the procesor to the rest of the heatsink in addition to transferring the heat to the air. So you'll see heatsinks made of multiple metals. Everything is always a trade-off, somewhere.

All copper heatsinks are surpringly heavy, and still don't conduct heat as well as good heat pipe, so you'll see heatpipes in some heatsinks even where there is no need to move the heat somewhere else due to space constraints, like in a laptop.