"Smoking Gun" - finally!

[PaulLowrance]

The following post shows a circuit of what I had in mind -->
http://greenselfreliantenergy.com/forum/index.php?topic=19.msg53#msg53
The above circuit would not require a resistor. So if a diode is an open-circuit, then it only affects that individual line of diodes.  If a diode shorts, then it's no bid deal, just a loss of one single diode.

Of course, adding a resistor across each diode would help if a diode becomes an open-circuit, but IMO that's a bit much. If by chance a diode is open, then it's just one line out of 10 billion lines. So 1 / 10 billionth is not going to be noticeable. A 10cm x 10cm diode array chip consisting of 1 trillion diodes would have about 5 to 10 billion diode *lines*, where each line consist of 100 to 200 diodes in-series.

PL

[tinu]

...
That brings up the other topic. People see that number, "trillion," and freak out.
...
PL

Hi PL,

I’m not sure this is the case. At least not for the people that really matter for business.
I suppose they freak out for another reason(s):

There are good studies available that show, as predictable, that any noise theory (1/f  included) that does not obey thermodynamics will fail. I regret for not being able to nominate at least one such study but I remember learning about them some time ago and I recall they are freely available on the internet.
Simply put:
1. one can not extract energy from an isothermal reservoir alone. Noise-extraction method (as any other conceivable attempt) is included. I have little doubts that there is room for debates here but I’m nonetheless open to discuss the subject.
2. If we speak about a temperature gradient across a junction, then it is indeed theoretically possible to extract useful work, but:
a) Carnot is the king anyway and for infinitesimal thermal variations (as for a diode resting in environment) the maximum attainable efficiency is that given by the known formula and consequently the extracted work is minuscule. Is it worth trying?
b) if one is to create and mantain a significant thermal gradient across the junction(s) to overcome 1) and 2a) above, the wafer is no longer of an appropriate design for that purpose and possibly the whole device shall move into carrier-diffusion area hence into thermo-electricity, which is already known technology.

Any thoughts about the above concerns?

Cheers,
Tinu

[PaulLowrance]

There are good studies available that show, as predictable, that any noise theory (1/f  included) that does not obey thermodynamics will fail. I regret for not being able to nominate at least one such study but I remember learning about them some time ago and I recall they are freely available on the internet.
Simply put:
1. one can not extract energy from an isothermal reservoir alone. Noise-extraction method (as any other conceivable attempt) is included. I have little doubts that there is room for debates here but I’m nonetheless open to discuss the subject.
2. If we speak about a temperature gradient across a junction, then it is indeed theoretically possible to extract useful work, but:
a) Carnot is the king anyway and for infinitesimal thermal variations (as for a diode resting in environment) the maximum attainable efficiency is that given by the known formula and consequently the extracted work is minuscule. Is it worth trying?
b) if one is to create and mantain a significant thermal gradient across the junction(s) to overcome 1) and 2a) above, the wafer is no longer of an appropriate design for that purpose and possibly the whole device shall move into carrier-diffusion area hence into thermo-electricity, which is already known technology.

Any thoughts about the above concerns?

Cheers,
Tinu

Tinu,

1. Thermodynamics is a theory of macroscopic systems at equilibrium, and therefore the second law applies only to macroscopic systems with well-defined temperatures. On scales of a few atoms, the second law does not apply; for example, in a system of two molecules, it is possible for the slower-moving ("cold") molecule to transfer energy to the faster-moving ("hot") molecule. Such tiny systems are outside the domain of classical thermodynamics. For any isolated system with a mass of more than a few picograms, the second law is true to within a few parts in a million-- Reference: Landau, L.D.; Lifshitz, E.M. (1996). Statistical Physics Part 1. Butterworth Heinemann. ISBN 0-7506-3372-7.

2. The mathematical concept of equilibrium is an impossible state in real life; i.e., it would require infinite thermal insulation to achieve perfect equilibrium.

3. Standard Gaussian thermal noise equation used in modern nonlinear physics violates the laws of thermodynamics. Reference: Thermodynamically valid noise models for nonlinear devices. So either there's a problem with thermal noise with Gaussian distribution or it is possible to rectify thermal noise.

4. You do not use thermodynamics to solve diode modeling problems. To solve diode modeling problems you use diode modeling mathematics that is based on semiconductor physics, which is based on quantum physics. The most accurate small signal semiconductor mathematics *clearly* predicts that diodes *must* rectify natural ambient thermal energy.

5. I wrote a trapdoor software simulation program that clearly shows natural ambient thermal energy can be rectified. When time permits, I will release the source code, but IMO it won't matter because the evidence that my diode arrays are producing a DC voltage from natural ambient thermal energy is becoming overwhelming.

6. For 12 months now I have been measuring a DC voltage produce by a passive diode array contained in at least two layers of metal shielding, far in the various rural areas.


If you're interested in some mathematics them please help yourself -->

http://greenselfreliantenergy.com/physics/dirtydetails/

If you have any questions regarding the testing procedure, then you'll find a lot of the answers -->

http://greenselfreliantenergy.com/forum/index.php?board=9.0


I would be more than happy to have a scientific discuss about this topic, but it must be mathematically based. In terms of scientific discussions, I'm uninterested in claims and handwaving, no offense intended. Thermodynamics mathematics does not provide barrier height, depletion width, dynamics resistance, rectified voltage and current, which is why the best physicists use semiconductor mathematics for diodes. I can assure you that the best small signal semiconductor mathematics clearly predicts that diodes must rectify natural ambient thermal energy.


PL

[tinu]

Hi PL,

I’m not unfamiliar with semiconductor physics, although it’s been a while since working with it closely. Recall that most equations are based on simplifying assumptions and that most solutions describe only a limited operational interval, out of which their validity is not even assumed to hold true.  We may go through the whole set of equations if you wish but it won’t help much imho.

Instead, let’s keep on basics for a while and then move further, ok?
Here it is, large cracks that apparently went unnoticed/undiscussed:

1. You said that the chip will cool down when generating power. The chip (as well as any single junction within) is still a macroscopic body that comprises a huge number of components so any statistical deviation from thermodynamics is no larger than the odds my dog will play Bach on my piano. Please reconsider the self-cooling issue and either retract or state what will go hot in the setup and based on what macroscopic energy source.

2. You also estimated quite huge and unbelievable power densities (18kW/sqm) given the circumstances. Assume for a moment you have the chip done, 1sqm, placed into full sun (1.2kW/sqm) and you output +10x more of what sun provides. Where is the source that gives you that extra power? Will the technology stop Global Warming? No offense but see the point and huge chain of ironies? It’s hardly believable even by the most optimistic dreamers. Not to mention what would be the explanation for 18kW when one moves the entire setup into shadow and let it reaches thermal (quasi-)equilibrium. Please detail.

3. I can’t see the path for logically deducing that even if the Gaussian model for noise is known of not being accurate that means noise must be rectifiable by a diode. Instead, let’s recall a diode will rectify any signal which is larger than known values, thus crossing its Fermi barrier. But the signal is *outside” of the depletion area within a diode, hence outside the ‘diode’ itself. I therefore fully agree an appropriate diode will rectify the noise from a coil (or from a resistor etc) or even from an identical diode as long as the rectifying one is kept cooler.  But do you have any good reason to believe that a diode will rectify its own thermal noise? Why is that? Any references available? I ask because I can’t see how it could possibly do that self-rectification of its own noise. If it could do that, according to the same thermodynamics again, it will cool itself down to 0K. Please comment.

Cheers,
Tinu

[PaulLowrance]

Hi,

Where's your math! I don't even see references. This is my final request. I think this will be a short lived discussion.

Hi PL,

I’m not unfamiliar with semiconductor physics, although it’s been a while since working with it closely. Recall that most equations are based on simplifying assumptions and that most solutions describe only a limited operational interval, out of which their validity is not even assumed to hold true.

"Simplifying *assumptions*"? I don't think you'll find that many physicists who specialize in quantum physics telling you that semiconductor physics is based on *assumptions*.  Please provide a reference.  Perhaps you meant to say "approximations."  All mathematics, including thermodynamics, are approximations.

1. You said that the chip will cool down when generating power. The chip (as well as any single junction within) is still a macroscopic body that comprises a huge number of components so any statistical deviation from thermodynamics is no larger than the odds my dog will play Bach on my piano. Please reconsider the self-cooling issue and either retract or state what will go hot in the setup and based on what macroscopic energy source.

Each diode is considered a separate system. For most diodes (up to microwave diodes), the thermal noise across the depletion region is equal to sqrt(kT/C), where C is the junction capacitance. Each diode converts thermal noise into DC. Multiple diodes aggregate the DC energy.

2. You also estimated quite huge and unbelievable power densities (18kW/sqm) given the circumstances. Assume for a moment you have the chip done, 1sqm, placed into full sun (1.2kW/sqm) and you output +10x more of what sun provides. Where is the source that gives you that extra power? Will the technology stop Global Warming? No offense but see the point and huge chain of ironies? It’s hardly believable even by the most optimistic dreamers. Not to mention what would be the explanation for 18kW when one moves the entire setup into shadow and let it reaches thermal (quasi-)equilibrium. Please detail.

By means of thermal conductivity; e.g., air or liquid flowing over the chip. 

3. I can’t see the path for logically deducing that even if the Gaussian model for noise is known of not being accurate that means noise must be rectifiable by a diode. Instead, let’s recall a diode will rectify any signal which is larger than known values, thus crossing its Fermi barrier. But the signal is *outside” of the depletion area within a diode, hence outside the ‘diode’ itself. I therefore fully agree an appropriate diode will rectify the noise from a coil (or from a resistor etc) or even from an identical diode as long as the rectifying one is kept cooler.  But do you have any good reason to believe that a diode will rectify its own thermal noise? Why is that? Any references available? I ask because I can’t see how it could possibly do that self-rectification of its own noise. If it could do that, according to the same thermodynamics again, it will cool itself down to 0K. Please comment.

It would not cool to 0K -->
http://en.wikipedia.org/wiki/Thermal_conductivity

Yes, I have good reason. The best known mathematics that is design to predict semiconductor behavior clearly predicts diodes must rectify natural ambient thermal energy. Furthermore, 12 months of meticulous measurements on ZBD's (zero bias diodes) has always shown a DC voltage. Surely you know how to calculate the noise *current* that flows through the diode depletion region that is solely due to thermal noise, correct?  On my website you can find the mathematics to calculate IS.  Next, you can calculate the diode dynamic resistance based on such noise current with a Gaussian distribution.
Some math for you -->
http://greenselfreliantenergy.com/physics/dirtydetails/
Are you aware of the diode square law. It is well known that the rectified DC voltage produced by a diode is relative to the square of the AC voltage signal across the diode. You can place 1pV rms across a diode and it will still rectify. I've placed less than one microvolt AC across the SMS7630 diode and it still rectified.

BTW, natural temperature gradients is the cause of Johnson noise. You can read just about any good physics textbook to learn that molecules, atoms, charged particles are randomly moving at RT, which could be calculated as temperature gradients.
http://en.wikipedia.org/wiki/Temperature

Math, please. Otherwise it's a pointless discussion based on claims and hand waving -->

PL