Little problem called the second law of thermodynamics, there.

Lets see what the article states that's relevant.

It generates power using ambient heat by way of a liquid with a very low boiling point.
It is being built for refrigeration uses.

Therefore, it either releales its own waste heat into the refrigerated area (which still notices a temperature decrease), or back into the heat it gained the initial energy from.
Given those two options, either thermodynamics is wrong, or just your misapplication of it.

That wouldn't work. If someone doesn't believe me, he should try to cool a room by leaving the fridge door open. (Well, it probably would work for a couple of minutes, but not in the long run).

You need to cool the 'liquid' (right now its a gas), so it becomes a liquid again. But you can't transfer thermal energy from the gas to the initial heat (just thermodynamics).

You could install heatsinks, but I don't think that would work to well .

If the outside temperature is cooler than the human body, and
If you don't mind a low-efficiency power source, and
If you can power all the cyberware on tens, maybe a couple hundred watts captured by the system,
Then this would work.

Beware hot environments. That'll shut the system down.

False. It does NOT do that. Read the article again, and visit Matteran's website. The system uses solar energy to heat the working fluid to at least 150-160F.

Quoting the website: "As depicted, the cycle requires a heat source of only 150F to produce refrigeration or air conditioning. Heat sources include solar, geothermal, & cogeneration."

Further, "The heat exchanger is then COOLED by the external ambient temperature source."

The outside air is used to cool the working fluid. Heat is supplied by - in this case - solar concentrators. This is described at length on the Matteran website:

http://matteranenergy.us/Animation.htm

Step through the animation.

That conclusion describes a perpetual motion machine, which should be your first tip off that you made an incorrect assumption earlier.

No, he had the right idea.

What the solar-thermal system in the article does is this:

1) Say the outside environment is 105F, like a steamy patch of Florida.

2) Solar concentrators aim sunlight at tubes carrying the working fluid and heat the fluid to over 105F, perhaps to over 150 or 160F. The hotter the better; the Carnot efficiency will be better. Let's say it gets heated to 160F.

3) Since the solar concentrators heated the fluid to more than 58F (in fact, to 160F in this scenario), it boils. The resulting steam blows through a turbine or piston engine, which spins a shaft. The shaft can drive a conventional air conditioning unit, or there's some alternate air conditioning systems that can be used.

4) The spent gas (now about 140F after having some heat converted to mechanical energy in the turbine) heads to a heat exchanger, i.e., a radiator. Outside air (105F) blowing over the radiator cools the working fluid (since 130F is hotter than 105F) before it's again sent to the solar heating section. Return to step 2 and repeat.

When applied to cybergear, you run into several problems:

1) The heat source, the body, is only 98.6F, give or take a bit. If the outside environment is warmer, the system will pump heat into the body rather than vice versa. Not only will you fail to generate electricity, you'll risk overheating the person.

2) Since the temperature difference between the body and outside is so low (rarely more than 100F), the efficiency will be lousy.

3) The human body only generates several hundred watts of waste heat, maybe up to 2kW for a behemoth bodybuilder working out hard. If you capture 5% of that as electricity from this system - maybe tens of watts - you're doing well.

That article is playing fast and loose with numbers.

Yes, there are memory metals that can be much stronger than natural muscles BUT those aren't the shape-shifting materials that manage larger contractions than natural muscles. The best memory metals can only shift their length by about 6%, far short of muscles. Further, you're lucky to get cycle times better than one second - you need both active cooling and heating to achieve that.

However, the fact that some materials can contract more than natural muscle is encouraging.

My favorite cyber power source was always blood sugar and oxygen, or alcohol and oxygen. Both are proven power sources for humans and machines, and it gives a street samurai and excuse to drink like Bender.

Heat differences are the basis for most power plants, from a car engine to nuclear power and solar-thermal systems.

I can't speak for the electronics (which are actually mostly optical and thus won't benefit - much - from superconductors). I don't know what the power drain of SR electronics are, though I note even beefy SR3 cyberdeck chips could go in the skull without risk of baking brains like a modern CPUs would. So, the power drain wouldn't be high with or without superconductors.

Electric motors are already ~90% efficient, so you wouldn't be saving much there.

Finally, I'm not certain that superconductors would help much with the brute motors of cyberlimbs - it would depend on the type of motor. Some contracting plastic like Battletech's myomers won't benefit from superconductors since most of the waste heat is generated in the plastic. The same can be said for memory metals, which depend on heat to shapechange. Electric motors I mentioned previously. Hydraulics...dunno.

I'm sure superconductors would be of benefit in power storage and you'd save a bit in delivery, but I don't think they'd have radical benefits in limbs either because electrical cybernetics are already pretty efficient or because the superconductors can't be used in an application.

While it has little to do with the actual power source, it does give us more materials to effect with the power source. We're all fans of carbon nano-tubes because the space elevator might use them if they get more advanced, but the idea of using them for muscle (or muscle-support) is kinda interesting.

This research might make it so that the CNT won't be rejected.