opamp thermal errors

Aug 15, 2015 26 Replies
P

The usual heat sink calculation uses just surface area and linear air flow.

If you really crank up the air flow, maybe pin fins are better, because maybe the extra turbulence increases the temperature gradient near the fins and so makes the heat transfer better. However, the heat capacity of air is pretty small, so I sort of expect that the temperature gradient might not get that much sharper. For a given back pressure (which is all you get from a muffin fan), I sort of doubt pin fins are an improvement.

I don't have a slam-dunk first principles calculation of that, though, which is why I was wondering about data.

Cheers

Phil Hobbs

Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Vote
Twitter Facebook LinkedIn
J

Air flow is computationally perverse. Air seldom behaves anything like intuition would suggest.

There is usually a still zone, or even reverse flow, just past the center hub of a fan. My fan above the pin-fin no doubt has strange and chaotic air flow, maybe dead near the center. But it cooled off the ZYNQ chip a lot... it was shuting down at 70C ambient with no cooling. The pin-fin helped some and then the fan helped a lot more.

Mounting the fan a bit off-center could cool better; we didn't try that.

We did have to design the adapter plate; the mounting holes on the microZed match no known fan dimensions.

Pin fin sinks are available in sizes that match various chips nicely. We buy from Cool Innovations usually.

John Larkin Highland Technology, Inc lunatic fringe electronics jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Vote
Twitter Facebook LinkedIn
D

On Thu, 20 Aug 2015 08:49:38 -0700, John Larkin Gave us:

Also adding a size reducing shroud to direct the flow and increase its velocity also helps. Especially on pin fin.

Vote
Twitter Facebook LinkedIn
J

I guess we could have bent down some of the sides of the adapter plate, to force more air into the heat sink. Once we got the box to run at 100C ambient, we declared victory and moved on. I wanted to run the temp up until it failed, but the guy who owns the prototype doesn't share my sense of adventure.

The ZYNQ fpga has an internal temp sensor and apparently has a default shutdown at 125C, which the tools let you change if you really want to. We didn't go there. So I don't imagine we'd get much above 100C ambient anyhow.

John Larkin Highland Technology, Inc lunatic fringe electronics jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Vote
Twitter Facebook LinkedIn
J

Pin fins are used a lot in military equipment, usually driven by 10,000 rpm screaming axial-flow "muffin" fans. The flow is most assuredly turbulent, and advantage because the roiling air scrubs the film of still air away. These fans can cost $500 each, and they do scream. I think some of them are capable of flight.

Joe Gwinn

Vote
Twitter Facebook LinkedIn
L

Den fredag den 21. august 2015 kl. 15.49.26 UTC+2 skrev Joe Gwinn:

the rows of double stacked fans in a server can get pretty loud

formatting link

-Lasse

Vote
Twitter Facebook LinkedIn
J

Yikes, we'd have to furnish each box with earplugs. The tiny fan that we used is inaudible. It also circulates air inside the box, which cools everything off a bit.

John Larkin Highland Technology, Inc lunatic fringe electronics jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Vote
Twitter Facebook LinkedIn

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required