The TPU uncovered? (A PROBABLE technique.)

[aleks]

Im working with high voltage sine waves.

How's progress going? Well, I do not believe you can achieve anything overunity with just sinewaves, or just square waves be them high voltage or low voltage without high spikes - if that wasn't the case almost every DC motor with a driver would exhibit overunity - if you are unaware, rotation speed of DC motors can be regulated by means of square waves with varying duty cycle. I believe abrupt spiking is a very important thing in this whole business - and I've shown the reason for this (Dirac delta function is a model of such spike, and when applied to acoustics it means much more).

[otto]

Hello all,

@Aleks

I agree. Not with ONLY sines or squares!!

I really didnt want to post this but......hmmmmm a few days ago I had HV sines with kicks on top of the sines. From my power supply 12V/ miliamperes. I have to say that I used my profi made PS. I mixed the frequencies and in just a short moment I saw that my voltmeter on the PS showed me the maximal voltage and then I saw a lot of smoke from my power supply. Nothing touched, just mixed frequencies. I had a runaway but this time only a voltage runaway, to say so. My PS is burned. My MOSFETS and everything OK. At the runaway I didnt see that the current rised. Just the voltage. All fuses OK.

I know that the output signal from a TPU is sine waves + kicks sitting on them. Such signals I saw before my "success".

Why always in my little garage such "successes" Better I would make something else.

Otto

[aleks]

I know that you all know the Tesla patent 381 970. I also was "familiar" with this patent.

Well, you should have noted that Tesla uses brushes. During motor rotation they produce spikes (sparkles) - much like from your MOSFETs. These spikes are pulsing thinner coils that are arranged in two-phase manner. Collectors (thick coils) are also arranged in two-phase manner. Due to nature of this motor/brush system the spikes produced on the thinner coils are mostly RANDOM with a frequency bump that depends on the frequency of the motor.

So, one of the possible arrangements would be producing spikes by means of pulse density modulation with probability distribution leaning toward the required frequency (60 or 50 Hz). This means that the "driver" should not be a simple square wave generator or 3 of them working in logical AND manner, but a random probability-distributed (frequency-wise) 1bit noise amplified to 400V - well, if MOSFET delivers avalanche performance you won't need to amplify the control signal itself that much. Simply MOSFET will be discharging at random times in concord with that probability-distributed 1 bit noise. This will create a lot of RE potential wavelets that - when summed - will replicate the frequency embedded into the probability distribution.

On another thought, motor/brush system may exhibit some frequency bump, but it is minimal. However, random spiking is still possible. When you integrate the voltage over time, you won't see them, but it does not mean they are not present physically. If that's the case it means each control coil should be driven by noise generator - well, actually both IN and OUT of the control coil should be modulated by noise. It's just a possible variant of this TPU device. I wonder if anybody tried to run this TPU device with uncorrelated HV noise in each control coil. Well, this is of course crazy idea - I'm not sure it may even work.

Noise gate is a transistor whose gate is modulated by random signal so that voltage across the noise gate changes between 12V and 0V. Since there are 2 noise gates driving a coil, the voltage will go randomly from one side to the other. What this brings I do not know, but from simple calculus the RMS voltage across the coil will be 3 dB higher than across noise gates. Of course, I do not even know if the noise signal can be efficiently utilized to do work. (noise gate is not audio noise gate which gates noise, by noise gate I mean gate modulated by broadband noise signal)

[aleks]

Why always in my little garage such "successes" Better I would make something else.

Maybe try to completely decouple collector and control? Just imagine you've connected your control coil to wall outlet. It will burn together with your PS. I believe the same happens when you "tune" your circuitry. Decoupling means control circuitry should not share ground with collector, it should be ungrounded or the surplus energy will flow both paths, through the control circuitry as well. Maybe also try to use "antenna" approach I've mentioned meaning you should not connect collector with the control circuitry via wires. Note that Tesla's power loads on that patent are also not wire-connected with the pulsing coils - induction is in use.

[buzz-ard]

@Otto: I hate that you've lost so much equipment, but in a way that's actually a good thing - you can't burn out testers if you aren't experimenting. All the gear you've fried speaks volumes about how much time and thought you are putting into this! You are a huge inspiration to me.

I am really glad to see that folks here are considering the Tesla patent again. I'm pretty well convinced that this is what SM actually built, taking all of his clues together. Almost all of the rest of his story supports this conclusion, if you look at it objectively. Like tuning his "delay coil" - this is something that would have been needed only if he were using a single input source and was trying to split it evenly between two outputs (thus eliminating the commutator). Or his statement about "100 years before anyone found the frequency" - it had been right at a hundred years since Tesla's patent when SM built his units. There's plenty of other parallels if one cares to step back and think with an open mind. Too many parallels to ignore, IMHO.

The Tesla device used AC sine wave input (from the generator) timed and split using a commutator (as close as he could get to square waves). The benefit of square waves is the fast rise time, which "kicks" the flow around the core. The system would run with gain when "ringing", and the output would be AC sine with a possible DC component. When properly tuned, the opposing vortexes created by toroidal influences would also be present. Much of this is first-year EE stuff, applied in a unique and practical fashion. Think outside the box!

I did my first build similar to the Tesla patent with no core and got interesting but unremarkable results. My current build will be a much more faithful reproduction and should produce better results. As I promised before, I will post a video when I have something to show. Everyone here owes it to themselves to print off copies of patents 381,970 and 382,282 and study them for a few days. The two are almost identical, but the differences also tell something too.