I think I've found Maxwell's Demon, however the demon is quite large...

[DreamThinkBuild]

Yes that was exactly my thought, this is why you could put the entire "power plant" into a "box", isolate it from the environment and have a wire coming out, giving you free electricity. In the heat pump scenario you would have two wires, one input and one output, with the out wire giving more than you put in because of the heat pump's extreme efficiency of separating heat into temperature potential.

All that is needed is two extremes. I found this interesting motor made of muscle wire on youtube.

http://www.youtube.com/watch?v=CR2f_NE6j4I

One side contracts when hitting the cold water while the other expands when hitting the warm water.

[Nabo00o]

All that is needed is two extremes. I found this interesting motor made of muscle wire on youtube.

http://www.youtube.com/watch?v=CR2f_NE6j4I

One side contracts when hitting the cold water while the other expands when hitting the warm water.

Hmm I've never seen that one before, and I like the rpm it gets even when it is loosely put together 
But like a sterling engine it is possible to use at a not too extreme temperature, and this might be less complicated to make also. Do you know how it works? I thought muscle tissue was exited by signals of electricity in the body?

[DreamThinkBuild]

Muscle wire is wire that expands when a current passes through or heat is applied. Neat stuff for small robotic projects. It contracts when it gets cold, although the drawback is the slow return. With the motor in the video using ice quickly contracts the metal. I'm assuming in the video that if the other tank is boiling water then as the wire expands on one side it pushes up on the wheel which feeds the wire into the cold tank which pulls the wheel. Pushes on the warm side pulls on the cold side then loops back.

Here's some more info on the wire.
http://en.wikipedia.org/wiki/Shape_memory_alloy

[Nabo00o]

Seems activity dropped down somewhat.....

Anyway, I just wanted to add that if a thin pipe is used, like the one we just have discussed, the inner wall should actually as well be isolated from the working fluid. I think it matters a whole lot if we use water, and a little less if we use air, this is because air doesn't transfer its heat as fast as the water does to the inner wall material (plastic, metal or anything).  The problem is that even though the fluid is isolated from its outside environment, a relatively good heat conductor such as PVC would constantly work against the creation of a temperature gradient inside the pipe by transmitting heat from the top to the bottom, thereby severely limiting the efficiency.

I believe that this is a real problem with the design, and does probably make it necessary to increase the width slightly, just to compensate for the volume taken by isolation.
Do any of you know of an insulator that would fit this task, to be thin, water proof (or at least capable to withstand water) and at the same time be a good insulator?

Naboo

[Nabo00o]

I know its been awhile since anything happened on this tread, partly because I have focused on other projects for the last months. Still I just want to inform you that the phenomena which cause the temperature to change is a called a temperature inversion.

You can check out the wiki page on it, at least they have some information regarding the process:
http://en.wikipedia.org/wiki/Inversion_(meteorology)