Testing the TK Tar Baby

[Groundloop]

Awesome TK, you got it!

The moral of the story is that when you short one charged capacitor to a discharged capacitor and lose one-half of your energy it's identical to an inelastic collision between a moving mass and a stationary mass.

In both cases you produce heat and that accounts for the 'missing' energy.

MileHigh

MileHigh,

How much energy do we loose when we charge a capacitor from a coil?
(The capacitor is a part of the coil, speaking about the LC nature of a coil here.)

First we apply a voltage (and current) over the coil. We remove the voltage
and the voltage over the coil goes to zero while the ampere in the coil goes to infinite.
This happen because the C part of the coil is fully discharged. Then the voltage
polarity flips and the current goes to zero while the voltage goes to infinite.
(Speaking about an no loss ideal coil here.)
In real life we have some resitive losses in a coil so we get a dampened oscillation.

So the question is, how much energy do we loose at each capacitor charge and discharge
for each cycle of dampened oscillation? And what causes the dampening? The coil resistance
or the capacitor charge or discharge losses? Do we loose half of the energy at each capacitor charge
at each cycle?

GL.

[MileHigh]

TK:

But how do the systems know to lose exactly half their energies to heat, and keep half in KE or capacitance?

You mentioned conservation of momentum.  And when we look at a the cap version we know that there is a conservation of charge.  Lo and behold, charge is equivalent to momentum.

Mass akin to Capacitance,  M same as C
Momentum (Mass x Velocity) akin to Charge,  MV same as Q
Velocity akin to Voltage, V same as V

C = Q/V, so Q = CV

Same as M = MV/V, so MV = MV

Conservation of charge or conservation of momentum dictates that 1/2 of the energy must be lost as heat for energy to be conserved.

MileHigh

[TinselKoala]

Right.. !! And thanks for forcing the clarification. It's been a long time since I've thought about the very basics, and it's always good to review and check one's work and understanding of it.


NOW....


Early tests with the 5 x IRFPG50 in place indicate no major differences from the 830s so far.... except as predicted the frequency of the oscillations has decreased significantly.... significantly in TWO ways -- and in the ability to partially turn on the Q1 transistor.

The first way of course is the magnitude of the decrease, which is as predicted knowing the difference in the mosfet's various capacitances, especially the gate. The second is what it reveals about the NERD team's device.

The delayed timebase of the HP180 is telling me that there are just under 17 periods PER 7 microseconds. Doing the math we get just a tad over 2.4 MHz. Which is exactly twice the 1.2 MHz mentioned on the NERD RATs video.
ETA: Philips says 2156 kHz.

Coincidence? Bizarre unaccounted for exact doubling of the frequency due to more random wire lengths? OR..... another MISTAKE in instrument interpretation by the NERDS?
I don't recall ever seeing an expansion of their oscillation trace so that one could actually determine anything about it.

One more difference I've noticed: It is easier to get the Q1 mosfet to turn on partially by mismanaged offset or frank bipolar pulsing.


So.... Again, I would really like to see _evidence_ in the form of analyzable data, that show the NERDs getting substantial load heating or drain current when the oscillations are occurring and a strict negative-going gate drive pulse is used.


Right now it appears that I can either get realistic load heating by allowing mosfets to turn fully or partially on during the antiphase from the oscillations, OR I can use a strict, non-offset unipolar negative going pulse to insure that only Q2 mosfets are involved and oscillating.

So.... for me to go further I need to know in which mode to operate. The heating mode draws over 1.5 amps on my inline ammeter, the purely oscs mode draws less than 200 mA depending on gate drive amplitude but usually less than 100 mA.

I'm going back over the actual data from the NERDs (scope traces, what dumps I can find) to see if there is real evidence that they are heating their load _without_ drain current flow indicated by a drain trace voltage drop. Any help here would be very much appreciated.

ETA: Heh... when I first stuck all the 5 mosfets in there and turned it on, at first it looked "normal", just like what I'd seen before, until I tried the bipolar pulsing and IT DIDN'T WORK to turn Q1 on !! Frantic scrambles, checking everything... sure enough, I had put Q1 in it socket adapter "backwards" -- so that it in fact WAS in strict parallel with the other four. Insert facepalm here, with lulz.

[TinselKoala]

Scoping across the battery.

Upper A trace is the common mosfet drains, at 100 V/div, baseline indicated.
Lower B trace is across the battery terminals, at 20 V/div, baseline indicated.

Main timebase is 0.2 milliseconds / division , delayed expanded timebase 1 microsecond / division.

Inline DMMs indicate 37.8 V on the battery, 100 mA draw. (Batteries were freshly charged overnight; present no-load voltage 37.9, load temperature (from previous tuning !!) at 110 F.)

[MileHigh]

Groundloop:

The amount of energy you lose per cycle of oscillation is dependent on the resistance in the wires.  Beyond that, you have a lot of misconceptions about inductors.

I suggest that you read the thread linked below about how coils work when they discharge their stored energy.  I urge you to try to understand it completely.  People on the free energy forums experiment with coils for years without actually understanding how they work.  That should change and the more people that understand the more peer pressure there will be on others to understand.

http://www.overunityresearch.com/index.php?topic=1312.0

MileHigh