To be deleted...

[nul-points]

Hi Itsu

Here's an example of the comparative stability of illumination, taken from my current photometric test run using a solar cell and a 24 LED head with nominal drive level around 50mA (+/- 10mA say) from 3x 750mAh NiMHs

Over a 6 hour period in the centre of the runtime, the supply volts decreased from 3.9V to 3.77V, a 3.3% drop

During the same period the illumination intensity reading decreased from 3.64V to 3.55V, a 2.4% drop

The NiMH cells spend the largest proportion of their time in this 'flattened' section of their discharge profile, between approx 1.4 - 1.2V, with 1.3V indicating approx 50% charge.  This region can be considered reasonably linear (for all practical puposes)

This all goes to show that the flashlight illumination from this circuit will be perceived to be reasonably constant, for the duration of its effective working voltage range

I've been investigating LED drive circuit technology as a result of this project, and i find that it is quite common to use pulse drive circuits based on boost converter chips, eg. to supply LED backlighting in portable equipment

You can understand that the designers will have 2 important design goals on their list:
- good efficiency for battery operation;
- stable illumination levels

Most of these application notes for the different devices show the LEDs driven in grounded mode (as we saw in my comparison tests above), so i believe we were performing a valid comparison with our circuit in this project

If i understand him correctly, Conrad believes that there is no 'free lunch" - ie. we can only extend runtime at the expense of illumination, in which case we would just have to make a tradeoff as to what was most important for us at a certain occasion - a brighter illumjnation for less time - or dimmer illumination for more time

But - if we can achieve more work (because work is not conserved) from the same input energy, then we can extend the runtime by a proportion (losing a little voltage each time, of course, due to losse), and as we've just seen in these preliminary results from my current test, the illumination decreases proportionally less than the voltage

Since our eyes are very non-linear with regard to illumination intensity, we can extend the effective voltage discharge slope and our eyes/brain will interpret the illumination as remaining relatively constant

This is of course *extra* to any effective gains made by pulsing light with an appropriate duty cycle - i am saying that we can use a circuit like this to provide additional work to that other gain, using the same original amount of energy

Tests continue to provide real data in order to make informed decisions, rather than just offer unsupported opinions

Keep up the good work guys
np

[nul-points]

Results are in for comparative 'Illumination vs Runtime' tests between a commercial 24 LED head with straight DC drive, and the same LED head driven by a 'flyback flash-lite' circuit being tested in this project thread

Both setups used the same 3-cell 750mAh NiMH battery, fully-charged to the same initial voltage, with circuit settings adjusted to provide the same initial intensity of illumination

This is to test the suggestion that the flyback flashlight only extends the illumination runtime of the device by decreasing the illumination intensity (ie. actual, not perceived, illumination ) - so this is not merely extension due to duty-cycle

The red traces record the battery terminal voltage, the blue traces record the illumination intensity profile; please note that the 2 graphs have different time scales - the DC run lasts 5 hours, the flyback flashlight run lasts 10 hours


For readers who aren't interested in actual data, i'll include a little story which i heard many years ago, to keep you amused while the rest of us consider the implications of these results...

A famous fast-bowler, who was a member of the English cricket team, was once invited to join a village cricket match, by his friend whom he was visiting

Let's call this bowler 'Henry' (after the unit of inductance)

The visiting team provided the umpires for the game

Henry, not wishing to take unfair advantage of his skills, bowled modestly and soon realised that he was not only playing against 11 members of the opposing team, but also against the umpire

When one ball, bowled by Henry, hit the batsman's leg-pad and was deflected away from the wicket, Henry's team cried out "Howzat!" to claim the dismissal of the batsman

"Not Out!", replied the umpire

Shortly afterwards, another delivery by Henry was heard to just 'snick' the bat as the ball flew past the batsman, into the wicket-keeper's gloves

"Howzat!" shouted Henry's team

"Not Out!", replied the umpire

By this time, Henry realised that the umpire was not interested in fair play, he was only interested in making sure that his team won

So, with some reluctance (see what i did there?), Henry took his celebrated long run up to the delivery end and sent an express-train of a bowl, rocketing down towards the batsman

The middle stump was knocked clean out of the ground and went flying past the wicket-keeper, who had to dive for cover

Henry turned to the umpire, winked, and said in his gentle Lancashire accent, "We nearly 'ad 'im THAT time, eh?"

...meanwhile, back at the bench...

The data not only disproves the idea that the approach being used in this thread is only making a gain in runtime at the expense of illumination intensity, it also shows that whilst the runtime has been doubled by the circuit, the illumination remains at a more consistently high proportion of the same initial level than the commercial DC drive arrangement, using NiMH cells


"We nearly 'ad 'im that time,eh?"  ...he was a card, old Henry!
np

[itsu]

Nul-points,

...meanwhile, back at the bench...

The data not only disproves the idea that the approach being used in this thread is only making a gain in runtime at the expense of illumination intensity, it also shows that whilst the runtime has been doubled by the circuit, the illumination remains at a more consistently high proportion of the same initial level than the commercial DC drive arrangement, using NiMH cells

Those are amazing results indeed, its a win win situation looking at the doubled time and stable / flatter light output.

I will be collecting some parts to be able to build a 'flyback flash-lite' circuit myself as i have a project in mind which is using leds and for which this circuit would be perfect.

Thanks,   Itsu

[nul-points]

 












...So - it's true....





.....the drugs really *don't* work.....





 

[nul-points]

Lol