Overclocking Using an Element of Peltier

[Grall]

...Which is powering the steam turbine electric generator needed to drive all of those peltiers!

 

[swaaye]

Yes, out of your case and into your furnace or outside the building...

Hmmm it is winter right now....

 

[aaronspink]

Peltiers can be a super-cool (or hot) solution to a problem, IF you have the needed electronics to make it work like you want it to.

The primary use for Peltiers today is chilling/super chilling CCDs/CMOS imagers used for astronomy. They work fairly well in this application, because the CCDs generate very little heat themselves and you are really just interested in extracting as much ambient heat from the CCD in order to increase S/N and/or sensitivity.

 

[Bouncing Zabaglione Bros.]

unfortunately you usually just change the problem from remove the heat from the cpu to remove the heat from the hot side of the peltier

More than that, because you've put more energy into the system, so you have to get rid of the heat of the peltier in addition to the CPU.

 

[Grall]

The primary use for Peltiers today is chilling/super chilling CCDs/CMOS imagers used for astronomy.

That makes sense, although I figured they'd use some kind of phase-change cooling, to reach much lower temps than what I figure a pelt would be able to reach. At least in larger telescopes, where the added complexities wouldn't really be an issue.

Then again, maybe you don't want to freeze CCDs too much either? Repeated deep thermal cycling can do bad stuff to semiconductors, so perhaps there are limits...

 

[aaronspink]

That makes sense, although I figured they'd use some kind of phase-change cooling, to reach much lower temps than what I figure a pelt would be able to reach. At least in larger telescopes, where the added complexities wouldn't really be an issue.

Then again, maybe you don't want to freeze CCDs too much either? Repeated deep thermal cycling can do bad stuff to semiconductors, so perhaps there are limits...

In large telescopes, they can use pretty exotic cooling systems (after all when you are spending 1+ mil on the CCD, you can afford a plumber, but for the prosumer/university level in the $5-500K, it is generally peltiers for both maintenance, mobility, and cost reasons.

 

[Grall]

That does make sense, yes, although as shown by PC enthusiasts, a vapor phase change cooling setup - even a cascading one - doesn't have to be particulary complicated or expensive, all things considered.

You know what temps you can reach with CCD by using pelts? Maybe the power output of those sensors is so low that a bunch of pelts can match non-exotic phase change cooling...

 

[MfA]

I'm surprised no enthusiast has used a cryopump yet ... for a mere 30K or so you could have a rig cooled to LN2 temperatures in continuous operation.

PS. cascade seems a poor man's solution, AFAICS watt for watt single stage cryopumps are a lot more compact at least for LN2 temperatures (hell I don't think anyone has DIY'd a cascade which could pull 100s of Watts down to those temperatures).

 

[nutball]

You know what temps you can reach with CCD by using pelts? Maybe the power output of those sensors is so low that a bunch of pelts can match non-exotic phase change cooling...

I have one on the desk next to me that cools to -80C, it's also available in a version that'll go to -100C.

 

[Grall]

That IS pretty darned impressive...! What kind of wattage do you burn for that performance? (Pictures/linkage please! )

 

[nutball]

Here you go:

http://www.andor.com/scientific_cameras/ikon-l/

Never measured the power requirements (just plug it into the mains!), the spec sheet says 150W.

 

[MfA]

Of course that's probably 149.x Watt cooling for less than 1 Watt load.

 

[nutball]

Probably. Inefficient as it may be it's less hassle than messing about with liquid gasses at >2km above sea-level (which is where these cameras spend most of their lives).

 

[aaronspink]

That does make sense, yes, although as shown by PC enthusiasts, a vapor phase change cooling setup - even a cascading one - doesn't have to be particulary complicated or expensive, all things considered.

You know what temps you can reach with CCD by using pelts? Maybe the power output of those sensors is so low that a bunch of pelts can match non-exotic phase change cooling...

-50 to -150 depending. Large hoses on the biz end of a moving telescope aren't a good idea.

 

[Grall]

Large hoses on the biz end of a moving telescope aren't a good idea.

Those things move incredibly slowly. If we could go to the moon with 1960s era tech I think we could design a hose attached to a rotating telescope.

Besides, the hose wouldn't need to be particulary large. As already mentioned, the power that needs to be cooled off is very low.