Single AA battery to light WHITE LED for long-long time

[Feynman]

@ Zon

Can you please detail the winding of your toroid  (bi-filar?)   

Thanks

[amigo]

@ Feynman

You can use a small toroid core or a ferrite bead. I actually did not have a ferrite bead but had small ferrite rings, several stacked together to makeup a bead. On a ferrite bead I do about 20 turns bi-filar wound using 30-34 AWG. For a toroid core in the MAKE video they used plain insulated wire, two lengths of about 10 inches, bi-filar wound to make 7-10 turns.

[Feynman]

Thanks amigo

I'm wondering how the initial permeability of the toroid effects the circuit.  (What type ferrite?)    I'm also wondering if there is a way to make this self-running (no battery only caps) like the micro-TPU. . .  except lasting longer than an hour or two.

Also in Zon's circuit, it would be nice to see a scope trace, if for nothing else to figure out the output frequency of the conventional positive EM energy.  Also, perhaps increase the electrolytic capacitor value and see if that allows more LEDs to be added either in series or parallel.  . . and whether the additional load lowers the run time, by how much, etc.
test

[Feynman]

Okay, I did some research on this tonight.

LEDs will light up with frequency input anywhere from 0 - 250khz pulsed DC.   The highest I tried was 250khz and it probably works into Mhz.  Watching the light, you can't tell the difference;  the LED looks like it's continuously lit.

The current draw for a standard white LED using continuous DC at 6V through a 1k resistor was approx 15-20mA.   Standard 1.5V NiMH battery is about 2000 mAh, so that should run for approx 100 hours in a conventional setup (sound correct?). 

[amigo]

In my "bedinified" circuit I was running 25 parallel LEDs at a time, but as I said I was not happy with the light illumination. I was using super bright super wide angle white LEDs (these have flat tops) and oscillating frequency depended on the size of the ferrite bead I'd use.