On July 29th, 2009, a theory flashed into my mind. Moments later after analyzing the theory, it felt like a lightning bolt struck me, literally! My entire body was tingling with excitement unlike I've never experienced before. IMO, this theory is on the correct path, further details to come, but here's the outline -->
The eventual goal is to compete the mathematics to prove the theory, which I am working on, but a few areas such as diodes in-parallel is posing some difficulties, mathematically speaking. Diodes in-parallel has always been my nightmare. :-|
Experimental data is showing that electrical resistance of materials, probably all materials, varies to some degree depending on current and settling time. At extremely low current levels, the resistance of insulators can take weeks to appreciably settling down, that is, reach equilibrium. Lets say the diode was reset, IOW disturbed, and the resistance is 10 Gohm, then 1 pA of DC current is forced through the diode. Over time the diodes resistance will slowly increase after being reset. In this example lets say that in 10 hours it eventually settles at 100 Gohm. Now lets decrease the DC current to 10 fA, in which case it settles at 10 Tohm in 5 days.
Resetting the diode is achieved with sufficient energy. For example, 1 uA for 10 seconds might be sufficient to reset one particular diode, or rapidly heating the diode from 70 F to 120 F. When the diode is reset, it is at its minimum zero bias resistance. Once the diode is left unconnected and undisturbed, the resistance begins to slowly increase. It is ambient thermal energy that performs this task. As this slow process occurs, charged carriers become trapped over time jumping from the conduction band to the valence band as the charges drift across the junction.
The moment a load is placed across the LED, the diode is placed out of balance due to noise *current*, and ambient thermal energy causes the resistance to change, which releases charged carriers. While the load stays on the diode, the stored energy from trapped carriers slowly releases, which shrinks the depletion width. As the diodes stored energy depletes, the DC voltage and current will decrease, and will continue to decrease over time. An overly simplified equation would appear as E = I^2 * dR * t, where E is energy, t is time, I is effective noise current, and dR is the change in resistance, as it is the change in resistance that releases the energy stored in the diode.
In order to continually draw energy, on average of course, the load must switch back and forth between normal resistance and insulator. Lets use an example where the switching rate is 10 Hz. For a diode array chip, the load could be 100 ohms for 0.1 seconds, then the next 0.1 seconds it would 10 Tohm, and the next 0.1 seconds it would 100 ohms, etc.
What this means is that expensive THz diodes are not required. Even a cheap low bandwidth amorphous diode would do the job. At such low current levels, the diode reacts at an incredibly slow rate, which of course would rectify Johnson noise, but the DC voltages would be so low that it's useless here. According to the theory, ambient thermal energy is striving for equilibrium by adapting to the diodes resistance, which varies according to diode load. So if there's no load connected to the diode, then ambient thermal energy slowly increases diode resistance to maximum. When a load is placed on the diode, more noise current flows through the diode, and it is ambient thermal energy that seeks a new equilibrium state, which releases charged carriers resulting in less resistance.
If the theory is true, it appears that the TED effect is after all a helpful effect, as it's responsible for the release of charged carriers. To be clear, this theory states that the main DC voltage is not due to the rectification of Johnson noise, but due to ambient thermal energy slowly trapping charged carriers, and then Johnson noise energy releasing such trapped carriers. So it's not a rectification process per say. This is good news in that leading edge THz diodes are not required to produce a diode array chip that produces usable amounts of power. In fact, inexpensive *amorphous* Silicon diodes would do the job. The chip would consist of at least 10 in-series diodes, and then paralleling such *groups* to obtain sufficient DC current. No two diodes are connected in parallel.
Created on 2009-07-31 20:18:18 by EnergyMover
FE diodes, FE Misc devices, Free energy, Free energy devices, Science, Scientific hypothesises, Diode, Free energy, Scientific hypothesises