3v OU Flashlight

[TinselKoala]

Well, that's good news, I was afraid I'd run out of wire.

I wonder how many other "mistakes" there are on the schematic I'm working from, which definitely does say 27 and 54 mH.

However, I've tried the lower value side with several different inductors from 6 mH to around 900 mH and they all "work" in the sense of making the LEDs light up brightly. Just using the inductor on the "27 mH" side and nothing on the other side, those wires still disconnected. The heavier the inductance the less current is drawn by the device overall, in my preliminary tests. The thing works as a Joule Thief, needing less than 2 volts from the supply to light up the two LEDs in series to extreme brilliance, with an oscillation of flat-topped spikes of about 9 V peak.

[verpies]

I wonder how many other "mistakes" there are on the schematic I'm working from, which definitely does say 27 and 54mH.

Both of these schematics shown here state 27 and 54mH.
Wesley says 5.4mH, common sense agrees with this.

What does Akula say?  Is there a link to that video without Wesley's voiceover ?

...with an oscillation of flat-topped spikes of about 9 V peak.

But not the 80V_P-P shown on Akula's scopeshot.

[Hoppy]

Well, that's good news, I was afraid I'd run out of wire.

I wonder how many other "mistakes" there are on the schematic I'm working from, which definitely does say 27 and 54 mH.

However, I've tried the lower value side with several different inductors from 6 mH to around 900 mH and they all "work" in the sense of making the LEDs light up brightly. Just using the inductor on the "27 mH" side and nothing on the other side, those wires still disconnected. The heavier the inductance the less current is drawn by the device overall, in my preliminary tests. The thing works as a Joule Thief, needing less than 2 volts from the supply to light up the two LEDs in series to extreme brilliance, with an oscillation of flat-topped spikes of about 9 V peak.

This chip is also used as an LED driver as you are probably aware, so its presumably quite efficient at this task.

I'm still considering the possibility of coin-cell or two in a pot, wired internally to the pot solder tabs. Its a pity we cannot see the underside of the 3V OU flashlight board, so that it might be possible to draw up a meaningful schematic.

[scratchrobot]

hi scratchrobot,

It looks like your device was just to about to self run for few seconds longer than the rest then it stops.

Do try to change resistor  to white led with a little higher value.

Do prevent the copper wire from touching the ferrite core.

I am sure someone may ask you where to get the ferrite pot core online base on the one your are using video.

Thanks for the tips, i will try them. The ferrite pots i used are B65701W 30 x 19 i got them from ebay.
I will also try other ferrite pots with the circuit. I don't think the circuit will self run but i like trying different coils to see what happens.


Regards

[TinselKoala]

Both of these schematics shown here state 27 and 54mH.
Wesley says 5.4mH, common sense agrees with this.

What does Akula say?  Is there a link to that video without Wesley's voiceover ?

I don't know, I am just working from the "official" schematic.

But not the 80V_P-P shown on Akula's scopeshot.

That's right.... just almost exactly one tenth of the value shown on that scopeshot. Of course we don't know that the probe attenuation matches the channel setting....

More troubling is the waveform itself. I am not seeing anything like that, yet. In Akula's shot it appears that the frequency of the spikes is modulated by the voltage of the sinusoid from the secondary. But please don't forget that I don't have the complete inductor in my testing yet. However, the transformer inductors I have tried seem to have no effect of the "secondary".

If the actual inductances used are 2.7 and 5.4 mH, I should be able to get those values on my flyback ferrite easily enough. I'll be doing these tests with a full transformator later on this afternoon.

The DC-DC chip starts behaving strangely when the supply voltage is below 2 volts. I'm wondering now if the Akula waveforms are made by underdriving the chip.