To be deleted...

[nul-points]

Hi guys, first of all, thanks as always for your input

I think its time to address a few possible misconceptions....

Conrad,  i do see a flaw in the capacitor to capacitor tests ...but it may not be the one you imagined - my circuit has been intentionally arranged to ensure that the LED only conducts with flyback energy

The clue is in the naming: 'flyback flashlight'

Your proposed replacement of the LED will not work for any supply voltage below the effective forward voltage of the LED + circuit voltage drops (eg. 'On' voltage of Q1, IxR drop of L1, T1, etc)

In order to try and present a complete test, using the shorter runtimes which a supercap would provide, i was diverted away from my original requirement for LED biasing which ensured it received only flyback drive

I could repeat the test with the output capacitor pre-charged to match the input, as only this will be a true representation of a regular operating condition for my circuit but your proposed re-arrangement of the circuit is unsuitable as the white LEDs won't light at all for supply voltages below and away from their forward voltage (say below 2.5V approx?)


Carroll, thanks for the update, i'm pleased to hear that the David Bowling thread is actually producing some useful end-results

Maybe i need to repeat a statement near the beginning of my first post in this thread "Ok, nothing novel, there - except...."

I'm not claiming any originality for this concept here - these ideas go back a LONG way, well before Mr Bowling or i turned up on the scene!  :-)

The "except..." there refers to the fact that i've been measuring real gains with this circuit, whereas i've tried many times to get anything useful from the classic 'charge 2 parallel batteries from 2 in series" and also 'charge 1 battery from 2 in series' type approaches

...for example, in Prof Jones thread, back in 2011:
  http://overunity.com/10773/physicsprof-steven-e-jones-circuit-shows-8x-overunity/msg301808/#msg301808

That's great that you're finding real gains with that approach, i'd given up on it  :-)


This particular circuit, here and now, is my own development - a slightly ironic response to apparently over-hyped claims of OU flashlights for sale and horrendously over-complex (and under-unity?)  'self-running lanterns'

I wanted to show that you can achieve real gains with a fairly simple circuit, and as Itsu said it can be built and running over a weekend!

I try to provide a slightly different approach in my threads - low on hype, heavy on data, full-disclosure of circuit details

I've started to investigate the circuit performance in more detail, and on the basis of what i'm seeing, i plan various developments, not just flashlight/lantern applications

I'll just carry on - doing what i'm doing - and sharing what i'm finding with anybody who is interested in the same things


I wish you all you guys good success with your individual endeavours
np

[itsu]

Ok,  i see i am falling behind in this thread, but my bat42 just arrived.   I will catch up later today.

I made a pcb with the 2x bat42's, BC547, 10x 10F supercaps, 100K pot, 10mm led, 100uF buffercap.
The 10x 10F supercaps a divided 2x 5 in series making 2 stacks of 2F caps, backed/loaded by 2x 750mAH 1.2V rechargeables for now.

I made a video of this, kind of in a hurry and unprepared, so it looks clumsy, sorry about that.

First 3 minutes describe/shows this new setup.
Then i show/measure something i don't understand (led pulls 5mW, while neither battery/supercap seems to supply this)
From 13 minutes into the video on, i show the rundown of the circuit with the 2x 2F supercaps only.

No free lunch here as expected.

Video here:  https://www.youtube.com/watch?v=VAa-AAEWXv8&feature=youtu.be

I have some screenshots of the power calculations on the led, and both batteries if needed.
Where does the power come from powering the led?  Is it because its kind of floating compared to ground, so a measuring error?

Regards Itsu

[conradelektro]

No we don't have OU yet but any system that can extend your run time from the batteries by several times normal is worth seeing how far we can push it.

The problem is the comparison circuit. You say, that you can push the run time. That may well be, but you push it in comparison to what circuit? You push run time in comparison to what alternative?

If the goal is run time mainly disregarding LED brightness, longer run times can be achieved.

If the goal is "most light for the least input",  a circuit (although it pushes run time) might be a looser.

Like in all endeavours, success depends on the definition of the goal.


@Nul-Points:

You are of course right, my placement of the LED was not good for lower supply Voltages! It was late and I was not thinking straight. Sorry!

First argument: I do not know whether "grounding the LED" is the right comparison circuit in respect to your "full circuit". I suspect it is not, but I can not point out a good comparison circuit for supply Voltages from 2.2 Volt down to 0.6 Volt. May be I get an idea later on.

Second argument: For supply Voltages from 5 Volt down to 2.2 V I think I have found a good comparison circuit, which I attached. The important point is that the variable resistor is set to a value which makes the LED shine at 5 V as bright as in your "full circuit" at 5 Volt. From your graphs I take the Value of "0,4" for light brightness at 5 Volt (supply Voltage) for your "full circuit".

From your graphs we already have the data for your "full circuit", which is "LED brightness curve from 5 Volt to 2.2 Volt" and "run time from 5 Volt to 2.2 Volt". Now it would be nice to have this data also for the very simple comparison circuit.

The very simple comparison circuit would be the "normal way" of driving a LED which then can be compared (concerning LED brightness and run time) with your "full circuit". Unfortunately only from 5 Volt down to 2.2 Volt.

@Itsu and all thinkers:

Let's compare Nul-Point`s two alternatives, his "full circuit" and his "LED grounded circuit". My question: where does the energy go if the LED is grounded in comparison to the energy which goes into the storage battery (or cap)?

The run time in the "LED grounded circuit" is shorter and LED brightness is comparable for the initial duration. Where does the energy god if not into the LED? But the LED is not brighter?


@Nul-Points:

My ramblings are not intended as criticism. You have such nice measuring possibilities (data logger and LED brightness), therefore I see a chance that you can clarify many questions (and some of my questions might be stupid).

It is always difficult to follow the "energy". And like in all ways of life, what is compared to what? One should not compare apples and oranges.


Most efficient way of driving a LED in terms of most light output for the smallest energy input:

Adjust the power supply to a value (around 3 Volt) so that the LED in series with a 5 Ohm resistor gets exactly its nominal current (depending on the size of the LED between 20 mA and 200 mA, for some very small LEDs it is only 2 mA to 5 mA). That is the "ultimate comparison circuit" for any LED driver circuit.

There is of course the practical wish to use a battery which Voltage decreases over time. There might be the wish to use only one 1.5 V battery, which makes an oscillating circuit necessary (to lift the Voltage to at least 3 V for the LED).

I would say that the oscillating circuit costs energy in comparison to the "ultimate comparison circuit". Some energy will be lost in the oscillation components (coils, transistor) and I suspect that the LED suffers (will break sooner) if driven by spikes much higher than 3 Volt.

I also would say that a battery stays about 90% of its run time between 1.5 and 1.0 Volt. Then it runs down quickly to near 0 Volt. Is it worth while to design a circuit for less than 1.0 Volt (it would only last 10% longer than a simpler circuit which works between 1.5 V and 1.0 V).

The simplest way of driving a LED pretty efficiently: use three AA batteries in series (they will start out at 4.5 Volt and run down slowly to 3 Volt in 90% of their run time), choose a resistor in series with the LED, so that the LED has its nominal current at 4.5 Volt. It is hard to beat that very simple circuit in terms of run time and light output per energy input.

Greetings, Conrad

[conradelektro]

Where does the power come from powering the led?  Is it because its kind of floating compared to ground, so a measuring error?

I watched your latest video https://www.youtube.com/watch?v=VAa-AAEWXv8&feature=youtu.be.

Very nice setup!

I am sure you will find the right setting for your scope to measure the input power.

It may be, that the Current probe and the Voltage probe do not have the same ground while measuring the input Wattage?

The oscillations of the circuit might influence the ground of one of the probes and I suspect it is the Current probe, because it is an active one (while the Voltage probe is comparatively simple in terms of electronics).

Whenever I do scope measurements I find a new way of making errors. An electronics professor once told me that "measuring technique" is the high art in electronics and separates the masters from the apprentices. I am definitely a beginner apprentice. The better the scope the more errors one can make.

Greetings, Conrad

[nul-points]

I feel that it's necessary to make the following statement below, BEFORE it becomes a necessity - i am grateful to ALL the members who have contributed so far - each have made some positive contribution in one way or another.  I have seen on previous threads (of my own and other members) how the main focus of the thread can get diverted and obscured by 'straw men' arguments, irrelevant comments and a high-jacking of the thread-owner's intended test 'plans', therefore...

****************
PLEASE NOTE:-

SINCE THIS THREAD DOESN'T HAVE THE LUXURY OF MODERATION, I ASK YOU POLITELY TO KEEP YOUR POSTS TERSE AND TECHNICAL, RELATING DIRECTLY TO DATA AND CIRCUIT INFORMATION PRESENTED BY MEMBERS WHO ARE TESTING REPLICATIONS OF THIS CIRCUIT

IF YOU WISH TO INTRODUCE AN ISSUE WHICH ISN'T EITHER A COMMENT ON DATA SHOWN OR AN ANSWER TO A QUESTION POSED BY A CIRCUIT BUILDER, THEN PLEASE  P.M.  ME TO DISCUSS IT FIRST

REMEMBER THAT THE PEOPLE AT THE BENCH WILL HAVE THEIR OWN 'SCHEDULE' OF TESTS PLANNED - PLEASE RESPECT THEIR JUDGEMENT AND PLANS BEFORE INSISTING THAT CERTAIN TESTS BE CARRIED OUT

EVERYONE IS WELCOME TO THEIR OPINIONS, BUT THEY REMAIN OPINIONS UNTIL PROVED OR DISPROVED BY DATA - THIS IS NOT THE THREAD TO AIR OPINIONS WHICH AREN'T A DIRECT INTERPRETATION OF THE DATA SHOWN

IF YOU BELIEVE STRONGLY ENOUGH THAT THE TESTS REPORTED HERE ARE MISTAKEN, MISDIRECTED OR MISINFORMED THEN GO AND START YOUR OWN THREAD, CREATE AN EXPERIMENT AND REPORT YOUR DATA

ANYONE WHO APPEARS TO ME TO BE IGNORING THIS POLITE REQUEST WILL FIND THAT I NOT ONLY REPAY THEIR EVIDENT RESPECT BY IGNORING THEIR POST(S) BUT ALSO WILL LIKELY TREAT THEM TO SOME OF MY ASCERBIC WIT INTO THE BARGAIN  ...(I AM OLD AND UGLY AND I NO LONGER GIVE A D@MN !)

YOUR 'VISIBILITY' HERE WILL BE DEEMED TO BE ACCEPTANCE OF THE ABOVE 'GROUND RULES'

OK - LET'S CONTINUE
**********************