To be deleted...

[itsu]

Hi null-points,

thanks for the nice little diagram, its indeed a nice weekend project.

I put it together with what i had laying around and it worked straight away.

Toroid is a little yoke, coils 40:120, 0.5mm wire, 2n2222 transistor, 2x 1N5819 diodes, 10K potmeter, 2200uF cap, 2x 2000mAH rechargeables, bright 10mm led.
Yoke has one side isolated with a mylar sheet so there is a gap.

Running for about 12hours now, still going strong (well, the rechargeables are powerfull), i will see how long it goes on.
Screenshot shows the voltage over (yellow), current through (green) and the power (red) in the LED.


Video here: https://www.youtube.com/watch?v=hbWvCAH1aFc&feature=youtu.be


Thanks,  regards Itsu

[nul-points]

...,
thanks for the nice little diagram, its indeed a nice weekend project.

I put it together with what i had laying around and it worked straight away.

Toroid is a little yoke, coils 40:120, 0.5mm wire, 2n2222 transistor, 2x 1N5819 diodes, 10K potmeter, 2200uF cap, 2x 2000mAH rechargeables, bright 10mm led.
Yoke has one side isolated with a mylar sheet so there is a gap.
...
Thanks,  regards Itsu

Hej Itsu!

That's a verrrrry nice setup you have there - and i see from your videos that you take your experimentation very seriously!

Thank you for taking the time to make & post a video about my little circuit - i hope that other experimenters, such as youself, find it as interesting (and hopefully useful) as i have

There have been a couple of occasions now, starting with well-rested cells/batteries, where it appeared that after 1 discharge/recharge cycle the 2 terminal voltages had both increased slightly!   I was in the middle of other tests so i just made a mental note to revisit that scenario and be a little more careful in my study of it

I realise that you constructed the circuit with parts to hand - if you decide to spend any time with the circuit, you might be able to make a couple of changes which may improve efficiency slightly

Firstly, some of the 1N58xx Schottky diodes have a fairly significant reverse-leakage current (a characteristic which i've used in some very low-power oscillator circuits) - when it comes to reducing any possible leakage paths around charging circuits, i try to use diodes like these BAT42s

Secondly, the 2N2222 is a dependable workhorse for many circuits, but to maximise gain & switching speed (again, for improved efficiency) i try to use high-gain, high-frequency, higher breakdown voltage devices such as BC547 (i had to settle for BC337 here - i must have eaten all the 547s!)

I'm interested that you picked up on the air-gap - if i remember correctly, Harold claimed that the energy stored increased with gap size, so i've folded some insulating tape to create a gap of approx 2mm  - of course, this creates an angle of contact at the opposite side of the toroid halves, too

My tests, so far, appear to show that the circuit can extend the useful operation for the input charge on the cells/batteries, but it takes a long time to turn results around for each particular test - hence my initial trials using a single 750mAh AAA cell for i/p and for o/p

I've moved on to use a slight variant of the circuit, combining an idea from some earlier work, using an 8 LED head from a commercial flashlight

This latest setup uses a 3 cell battery of NiMH AAAs for i/p (& also for o/p obviously) - current draw for this setup is 65 mA, which will help me get more tests completed per week

I'll be interested to hear your observations, but i hope that this doesn't detract from your other investigations

All the best with your experiments

hej
np

(Edit)  PS  my variant circuit is running at approx 40kHz but my intial circuit, which you show, runs at 500kHz - a smaller yoke may give you higher frequency operation, if this can improve efficiency

[itsu]

Hi NP,

thanks for the tips, i have most of the components you mentioned, so will change them after i know how long the present setup will last.

It does not take much efforts to set it up, the toroid is the most time consuming part to build.

This morning (so after again 8 hours) the LED was still on, but the supply battery was at 0.5V  , so i switched again and now its at 1.2087V, mostly
recovered by its natural chemical process i guess.   The now run battery is at 1.278V

By the way, the both rechargeables where almost drained when i started, so i am already surprised they last this long.


Regards Itsu

[conradelektro]

@ nul-point and itsu:

May be you have two super caps instead of the two rechargeable batteries. The tests would be much shorter.

I attach the photo of a suitable super cap, but any similar type would work, more or less capacitance, even the 5 Volt types would be good.

The 1 F type super caps are low cost.

With a laboratory power supply or with a 1.5 V battery the super caps should be charged to 1.5 Volt before the test.

Greetings, Conrad

[nul-points]

@ nul-point and itsu:

May be you have two super caps instead of the two rechargeable batteries. The tests would be much shorter.

I attach the photo of a suitable super cap, but any similar type would work, more or less capacitance, even the 5 Volt types would be good.

The 1 F type super caps are low cost.

With a laboratory power supply or with a 1.5 V battery the super caps should be charged to 1.5 Volt before the test.

Greetings, Conrad

hi Conrad

thanks for thinking around the project!

actually, i made my first tests of the circuit using 2 supercap stacks, and last night i started testing with a 1F capacitor as 'B2' - you're correct, the run times are considerably shorter!

thanks for reading & for the suggestions

all the best
np