Thursday, 20 July 2017

The Green Lantern

Purpose of Latest Tests

This is an experiment with looping energy back to the source, using a blocking oscillator as a self-driving switch

the approach used here builds on two previous areas of test

a) my first investigation goes back about 9 years - it looked at an anomaly in conventional charge accounting - the test circuit was a simple switched-capacitor setup, transferring charge from one large capacitor (or battery) to another, using a much smaller capacitor, and an inductor, as a charge transfer method

b) in my second set of tests, just over a year ago, i looked at the use of blocking oscillators as a building-block for LED flashlight and battery-charging applications

my interest in thread (a) was revived recently, when i finally found someone who could give a practical explanation for the phenomenon seen when switching large proportions of 'charge' between capacitors (which many of us had documented as an approx 50% gain in total 'charge', stored in circuit capacitors, in stand-alone circuits, when transferring charge from input to output and using many small 'packets' of charge)

John Decker, a physicist, explains on his website:

the deep-rooted misuse of the word 'charge' when applied to the storage of energy in capacitors and batteries - he suggests the use of the word 'gorge' instead - charge is very definitely conserved - gorge'(as we've discovered, practically) is very definitely not conserved

if gorge is not conserved, this raises the question; "is there some way we can we re-cycle gorge to do more work?"  (my question, not Decker's!)

my latest investigation draws together all these threads, as building-blocks ...attempting to use 'synergy' to get more use from 'energy' 

my original circuit arrangement transferred gorge from input storage to a separate output storage, increasing total gorge as a result

my new circuit configuration returns what was the original output back to be part of the input - to see if gorge can be re-cycled (looped-back) and re-used

Approach used

these are the steps involved:-

first create a series-connected stack with 2 parts A & B - use a battery for store A, above, and either another battery or large capacitance for store B, below

this stack will be used to transfer gorge to a smaller store, Csw (eg. a lower-value capacitance) via a series inductor L1

gorge will then be switched in short pulses from Csw to store B (B2 or C1)

using a blocking oscillator as the switch means that we can arrange for some gorge to return to store B, as we charge the main inductor in the oscillator,

and then we can arrange for its coil-field collapse energy to be directed to return some gorge to store A

so, as an overview of the cycle, we allow a small amount of gorge to transfer, with low loss, from store (A+B) to Csw, then we transfer another small amount of gorge from Csw back to store (A & B) and repeat

if we comply with the transfer requirements, as shown by Decker, then we should be able to increase the total gorge in the system

Circuit operation

a switching capacitor C2 (the Csw), is filter-charged, via an inductor L1, by two similar batteries in series

the action of the blocking oscillator causes the collector winding of T1 to charge as current flows from C2 into B2, and the resultant voltage across the base winding of T1 re-inforces the turn-on of the transistor switch Q1 via C1

when the oscillator reaches saturation point the base voltage drops and Q1 switches off; the collapsing field in the collector winding is able to find a current path via the base winding and the LEDs, this current spike is preferentially applied to B1 (in a lower impedance path than L1, C2)

so B1 & B2 supply energy to the circuit, charging C2, and a proportion of this this energy is then distributed back to B1 & B2, with some being dissipated in the LEDs (and a small proportion of the energy being dissipated as losses in the circuit components, as heat)

Initial Results

The main and flyback windings are each 20 turns of 0.45mm magnet wire on a 20mm diam. x 10mm high ferrite toroid

with a Lantern/Flashlight type setup, using 4x 5mm Hibrite LEDs (green, on this occasion) as a parallel load to the blocking oscillator, and 2x 6V 10AH gel lead-acid batteries as B1 & B2,  the current draw of the active circuit, measured between L1 and C2, is approx 15mA

the base bias resistor VR1 has been adjusted to maximum resistance for this test, but still produces a bright output from the LEDs - the photo (showing the view from above) gives a good indication of the somewhat obscuring effect on the eyes, caused by the circuit in operation

in my ongoing test, both B1 & B2 started with low 'charge', having a combined on-load voltage of approx. 11.7V (both batteries are 'on load' since they are connected in series across L1 and C2, which then supplies current to the blocking oscillator)

initial results are very promising:  IF the battery experienced a current draw of 15mA, each Battery terminal voltage should decrease by approximately 1.5mV per hour - so the expected voltage drop for the combined stack of 2 batteries, over a period of 10 hours, should be in the region of 2 x 1.5 x 10 = 30mV

the combined starting voltage (Vb1+Vb2) for this initial test is 11.73V, therefore we would expect that voltage to decrease to approximately 11.70V after 10 hours continuous operation

the graph below shows the datalog results from the initial test:  the 10hr voltage drop is certainly nowhere near 30mV, it is not even 10% of the expected drop

in fact, at the moment, it is difficult to determine visually from the datalog whether the average combined voltage has dropped at all !   

[Update:  the average rate of combined voltage drop over the last 18 hours is 1mV/hr approx. - this represents a current drain from the batteries which is about 1/3 of that which is being supplied to the blocking oscillator.  If correct, this suggests that the circuit is able to perform at least 50% more work for the same energy  ie. 60mW being supplied --> 90mW being used**]

so it appears that the looping-back of the gorge has had a significant effect on the net current draw from the batteries!   i will continue to monitor the combined terminal voltage of B1 + B2 so that a more accurate value of the actual load on the batteries can be determined

Input power: (Vb1+Vb2) x battery drain current  = 12V x 5mA approx.
Power converted: ((Vb1+Vb2) - Vb2 ) x  circuit supply current = 6V x 15mA approx.]