PhysicsProf Steven E. Jones circuit shows 8x overunity ?

[xee2]

Great!  thanks, Xee2.  Hope you don't mind my showing it to people...?

No problem. Use as you like. If you want changes, let me know.

[nul-points]

hello Steven

i've taken up a couple of your suggestions for related investigation:-

a) increased base feedback Cap & transformer inductance for reduced repetition rate;

b) arranged circuit to suit feedback of o/p energy to supply.


a) has been achieved with the use of a 0.1uF ceramic capacitor and a transformer constructed from approx 2x 150 turns of 3x0.2mm Litz wound on 50x10mm OD ferrite rod, enclosed in a ferrite tube (approx 35x30mm OD)

the waveform modifies to a 12.5us pulse (+ similarly dimensioned coil-field collapse, immediately following in anti-phase), with an approx 345Hz repetition rate
(see trace below for AC waveform at emitter)


b) has been achieved by inverting your generic common-collector oscillator, using a PNP transistor, to enable easier re-direction of the o/p current path into the required energy-storage components

i realise that i'm now investigating a circuit which is different to your circuit 'specifics', but i believe that it still retains the essence of your design 'generics' and therefore it should provide a relevant test-bed for observing variations of harvesting and recycling the o/p

(see below for schematic of inverted, looped, srj1-family circuit)


the circuit is powered by two well-depleted AAA NiMH cells; these produced a total of 2.05V off-load, which has dropped to approx 1.5V in-circuit

this battery is connected to the positive emitter supply (which i'll label Vee), via an inductor of a few mH

the broad o/p recycling strategy has been to replace the emitter LED with a schottky diode and direct the current path away from the emitter into a buffer capacitor

this capacitor is charging up to approx 2.9V

the buffer capacitor is connected to Vee via an inductor of a few mH and a red LED

the LED is not bright, but it is easily visible

the current draw from the battery is approx. 50uA  (largely due to the very small mark-space ratio)


if time permits, i intend to monitor the battery terminal voltage trend for different configurations of the o/p recycling arrangement

i hope to post occasional progress reports; if any results suggest a further mod, please call it out and i'll try to include that in the 'schedule'

PS  i believe it's 'safe' to use CoP as an alternative to your 'n' - i understand it's an accepted measure of system performance, used for example in heat-pump technology (where CoP = 4, say, is not an unusual value)

all the best
np

[Apologies for rather large sized images - the 'scope trace loses resolution badly, when resized any smaller!]


http://docsfreelunch.blogspot.com

[JouleSeeker]

NP -- I am studying your circuit with interest, and learning.

i realise that i'm now investigating a circuit which is different to your circuit 'specifics', but i believe that it still retains the essence of your design 'generics' and therefore it should provide a relevant test-bed for observing variations of harvesting and recycling the o/p

(see below for schematic of inverted, looped, srj1-family circuit)

I agree with what you are doing here and look forward to your results.  Thanks for taking a close look at this, and for innovations in "harvesting and recycling the [output]" which is the most difficult part of the evaluation of the circuit. 

Best wishes for your success,
Steven

[TinselKoala]

I will remind you again: Power is not energy. Power multiplication is easy. If you measure, for example, my TinselKoil using the same techniques you are using here, you will find that it draws an "average" of 7 amps at 120 VAC from the wall, and the secondary arc is 8-10 amps at over 30,000 volts "average". Put that in your COP and smoke it.

I note that 5 or more entire pages have gone by in this thread and only a single person has said anything about measuring ENERGY in and out in your circuit.... besides me, that is.

I have put up a couple videos showing how POWER measurements are very susceptible to artifacts like stray inductances and measuring points, in a Joule Thief essentially equivalent to the circuit here under test. I also show how an ENERGY INTEGRAL is obtained and how that integral, when properly computed, is less sensitive to these artifacts.

Mean power during a time period.... can be considered an energy value ONLY if properly measured and computed. Have you learned nothing from the Ainslie affair? The proper way to compute energy out is to do it with an oscilloscope that can handle the math, OR.... like I said: trace it out and count up the area under the instantaneous power curve.

As long as you are talking about power in and power out and mean or average power.... you are clearly barking up the wrong tree. To try to show COP ratios using power you need much more sophisticated apparatus than you are using, something like the Clarke-Hess power meters, or calorimetry.

Please... just for fun.... do an actual energy balance measurement on your Joule Thief.

[hyiq]

Hi All,

@TinselKoala

I agree, its important we get this right to insure we know what we are dealing with. Like I said the only real way is to make itself run.

This is why I think when we post a Circuit Schematic we should get in the habbit of showing measuring points and component values better.

My current circuit is attached with all points clearly marked. Please can everyone give me your feed back. If I have something wrong let me know and I will correct and re-post. Also Probe polarity settings. I have used DC polarity on the Voltage side, but have used both DC and AC for the Cuttent Probe settings. My Scope is set to "Mean" measurement over time.

I am not measuring RMS or PK-PK. So far, I think a sort of Impedance matching seems to be going on. John Bedini said this some time back and I did not understand what he meant.

So lets set some standards to follow, constructively help in moving forward to trying to understand this and start scaling this up when we do.

all the best

  Chris