One would think that would be the logical thing to do, but apparently the design is “barnacleware” - it gets glommed onto an existing mechanical structure, and no aircraft power is available. The power comes from (are you ready for this) a small propeller-powered generator, once the device is deployed into the airstream.
Right. I must refer you to my first comment. 
Yes there are: Preconditioning: Dual Cooling Of Northbridge And CPU - 5 GHz Project: CPU Cooling With Liquid Nitrogen | Tom's Hardware
This tells us that at least an IC won’t instantly die at low temperatures and appears to function correctly, or at least a digital IC does. So probably no material phase-changes to worry about in the IC itself anyway. Connections, potting etc. are another matter!
You have an interesting job.
I meant condensation in any air contained in the device, if the air was equilibrated with an indoor environment and the device was moved outdoors. The air can have a higher dew point than the outside temperature, and that water would condense inside the device before the absolute humidity equilibrates in gas phases.
There’s a separate condensation issue when a cold device is returned to a warm humid environment, which happens on the exterior.
If it’s potted, though, there won’t be any condensation of the first kind and any condensation of the second kind won’t matter.
It sounds like it will be at a more reasonable temperature while the plane is on the ground, and will have power available from its propeller once the plane is in motion, before it reaches the extreme temperatures. Even though the outside of the pod will be in the extreme environment, is there really any need for the electronics to ever get that low?
This is what I’m assuming.
I’ll see what the customer thinks.
Disclaimer - I’m a pilot and I’ve read up on these systems, but I’ve never flown anything big enough to have one.
It sounds like you are working on something that will be mounted to a Ram Air Turbine (RAT). These are installed primarily on jet aircraft as a source of emergency power. During normal operations, the RAT is stowed inside the fuselage, but not in a pressurized or climate-controlled environment. When all normal sources of electrical (or sometimes hydraulic) power have failed in flight, it is supposed to automatically deploy into the slipstream to provide enough power to run emergency systems. When stowed, the only thing separating it from the outside environment is a thin aluminum or fiberglass door, so it will likely be exposed to temperatures below -50*C for extended periods of time. I can’t imagine it would be difficult to find power for your component to run some kind of heating coil, but in reality that is way beyond the scope of my knowledge. Perhaps LSLGuy or 1920s Style “Death Ray” might have more insight.