heat sink for video camera in vacuum

I've had CHIPS MELT ;-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     | 
| Analog Innovations, Inc.                         |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| Phoenix, Arizona  85048    Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Corn, or potato? ;-)

There's no battery, this runs from 12 V DC, and dissipates 792 mW.

Jon

We run some gear that has up to 32 chips dissipating about half a Watt each in a shoebox that runs in high vacuum. we have a cold plate for the hot chips to a water cooling loop. This works, and the cold plates carry off additional heat from the rest of the electronics via conduction through the PC board. Some of these systems have run for months at a run in high vacuum (5 x 10^-7 Torr).

In fact, the camera will be on the other side of the chamber, but in this experiment we have most of the electronics outside the vacuum chamber. (It's a lot faster to swap a board fiddle with a cable or whatever when you don't have to vent, lift the lid off the chamber and then pump back down.)

Jon

Yes, for a lot of fixed installations, they do that. A PMT won't work, they need an image to adjust the shape and position of the beam. But, for use inside an experimental chamber, the target will likely be at a different position each time, and it may be hard to put a window where you need it. So, a camera inside that can be mounted for the right view is the most logical approach except for the cooling issue. We still don't know about thermal coupling inside the camera body. If this one fries, then we will have to buy another and open it up and provide proper cooling of the chips. But, the way they do it at MSU is to just burn them up and replace, which sounds like the wrong way to do it.

Jon

It IS black anodized, but the size is too small. So, we are going to try increasing the effective surface area about a factor of 20 or so, but then because the plates have two sides, it actually is

40 X more area. Rough calculations show it will remove 2.4 W heat at 15 C rise, so if there is good thermal conductivity in the camera in vacuum (not known, yet) then it should work indefinitely.

Jon

Adjustable mirror?

I used to go to cocktail parties where I'd be asked what I did for a living. I'd reply, "Chips", and watch their eyes glaze over :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     | 
| Analog Innovations, Inc.                         |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| Phoenix, Arizona  85048    Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

How fast will the camera heat up?

One thing to think about (maybe you already have) is that you have two surfaces that need to emit/absorb the radiation... the first is your camera and the second is the walls of the vacuum chamber. If the walls have low emmisivity then they are not going to absorb much of the radiation. (Maybe there is enough area to make up for this.) (I've got an equation for two parallel plates with the same area.)

Could you stick a thick piece of copper braid onto the camera and chamber wall to use as a heat conductor?

George H.

This is a big chamber, about 2 meter diameter and almost a meter tall. All unfinished aluminum. That's a fair amount of surface area.

Some people do stuff like this, I don't think the thermal conductivity of a foot or so of heavy braid actually carries much heat. If the braid was very short and well-coupled at each end, it might work.

Jon

"Michael A. Terrell" wrote in news:4K2dnfBmLcNobYbMnZ2dnUVZ snipped-for-privacy@earthlink.com:

Why not put the camera in a glass tube with melted in pins to wire it, and filling the tube with dry nitrogen, before sealing it up? No failing chips, and enough glass surface radiating away the heat. Or even glue two glass windows in an alu pipe, filling with dry nitrogen and use melted in pins in the backside glass bottom. For that, you could even use the bottom part of a big audio endstage tube. About 8 pins included for free. OP said some sealing epoxies did nor seem to out-gas to much.

It's very easy to calculate, even more so around room temperature.

I have to use integrated thermal conductivities because it changes a lot at low temperatures.

Hey, maybe you could try an underwater camera housing? Test it to see if it will hold 15 psi to ambient.

tm

So at a guess you get something like the area scaled by the emmisivity of the Al.

d

You should pretty much be able to calculate the copper braid number. And get something close to that. I'm guessing it would be better than the radiator, but you'd have to do the numbers. Can you keep it shorter than a foot? It'd be fun to measure the radiator too.

Say here's a question for the science types. Can I measure the black body radiation inside a room temperature low emmisivity vacuum chamber? And if I made then made the walls high emmisivity, would I measure less radiation?

George H.

d

I'm with George.

When we were running a chunk of lidless Siemens ECL memory in vacuum - back around 1984 - so that we could use our stroboscopic electron beam tester/ voltage contrast electron microscope to measure voltages and timing at points on the surface of the working chip, we clamped one end of a shortish length of copper braid to the bottom of the board carrying the chip, and clamped the other end to the aluminium wall of the vacuum chamber. We didn't do any heat transfer calculations, but it worked and made the guys from Siemens (now Infineon) pretty happy. ECL static RAM tended to run hot.

--
Bill Sloman, Sydney

Fill the cam with epoxy or vacuum wax. Either is a better thermal conductor than air.

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation

The ROT is 150,000 k/w per ohm.

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation

id

What the 'bleep' do you do Spehro? (to be so wise in the ways of copper.) And I don't suppose you're looking to hire a physics/design type? (I've just sent out a application for a 'lab guy' job at a local college, I'm not sure I'll either get or accept the job, but I feel liberated.)

George H.

- Hide quoted text -