It's a presumed semiconductor implementation of the thought experiment:
https://en.wikipedia.org/wiki/Brownian_ratchet
We like to think that, by some intuition, the ratchet mechanism has zero forward force and infinite reverse force required to turn it; but that's clearly not true, as the pawl is pushed up a ramp as the crank turns, and for the pawl to remain in position, it must be spring-loaded, and this exerts a lever force on the crank. For thermal motion to be able to turn it, the spring must be so slight that, not only will the crank turn by thermal motion, but the pawl will jitter about as well, allowing the crank to skip backwards after all. The combined effect is no net movement.
Tim
Beautiful! I especially like the paragraph specifically on attempting to rectify Johnson noise. I'm not surprised that someone (Leon Brillouin) had thought of that idea before, and way back in 1950s. Discussions of this sort are not total nonsense, noting that Brillouin wrote a paper on this in a respected science journal :-
Brillouin, L. (1950). "Can the Rectifier Become a Thermodynamical Demon?". Physical Review 78 (5): 627
The demon, of course, is Maxwell's Demon, a thought experiment by the great scientist James Maxwell. As I have online access to all the science journals, I might even read his paper.
![ThumbsUp :-+](https://www.eevblog.com/forum/Smileys/default/icon_smile_thumbsup.gif)
But hey, we can make that diode work OK. We place our rectifier diodes and series/parallel wiring in the fridge, but the 'power-station-warm-resistor-array' at room temperature outside the fridge, and then just have a pair of high-voltage-high-current wires leaving the fridge, connected to our TV, or through an inverter to the power grid or whatever.
Tell you what Tim. Lets split equally on the Nobel Prize. Deal?
![Smiley :)](https://www.eevblog.com/forum/Smileys/default/smiley.gif)