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

[Feynman]

@Groundloops

Sorry, I wasn't trying to suggest they were interference (I think the harmonics are from the circuit)... I meant... 1) are the harmonics artifacts from the DSP in the spectrum analyzer/antenna and if not, 2) what are the practical significance of the harmonics?   Are they caused by the harmonic-rich sharp wave?  Why five octaves up? 

I think the 2Mhz harmonic comes from the circuit.  It's an exact multiple of 32, so it's five octaves higher.  (Take 2,000,000hz / 62,500 hz) = 32.  Then 2^x = 32???  Well x = 5, so five octaves.

Anyway Faraday cage is one way to know for sure, maybe measure inside your microwave   .   

@Zon

The higher frequencies don't mean your circuit isn't 25-350khz.... it just means there are higher frequency 'reflections'. I think your circuit is probably base freq 25-350khz. We are still trying to figure out what this means.  Groundloop is just seeing higher frequency echoes.

@ Stefan

I agree with your ideas, I think Bedini style is good idea here.

@all

No mail today. Just opened old tape deck , no toroids, but now I have some nice inductors and caps!  I'll open an old computer PS and see if I can find a toroid to get a replication going. I'm interested in the phenomenon here, what exactly is causing this... is this just a really efficient LED driver or are we dealing with something more?

[Groundloop]

@Feynman,

OK, now I get it. I think there are over harmonics. When I look at the scope there was very sharp spikes.
So the oscillation is spike like. If the pulse is somewhat like an square pulse then there can be all sorts of over harmonic frequencies. I do not have a Faraday cage right now. I'm working on my version 3 right now. Will post later.

Groundloop.

[gyulasun]

...
One funny thing, with my high Ohm resistor and very little current usage from the battery,
the LED will only light when connected ONE way. If I turn the LED around then it goes black.

Any ideas why this happens?

Hi Groundloop,

It happens because the switch-off pulse (I mean the induced voltage due to the collapsing flux just after switch-off) has got a definite polarity (usually it is opposite to the supply voltage) and LEDs need forward voltage polarity to give out light.  So if you turn it around, it will get reverse voltage polarity, hence no light.  You can check pulse polarity with your oscilloscope of course.

Gyula

[amigo]

Perhaps we should also explore why does 2N2222 perform better than 2N3904. I have noticed this over and over again in the past months while experimenting with the Joule Thief and my modification circuit.

2N3904 appears to be a poor performer while 2N2222 excels compared to it, yet when measured, both transistors had similar hFE here, so it must be some other characteristic that contributes to overall LED brightness and longevity.

I'm going to run more tests this weekend just to see what happens and report here...

[supersam]

@all,

if you look at all of the numbers in this thread, in the rodin math type base nine i think you will find  alot of, 1,2,4,8,7,5's, i wonder how the 3,9,6's can be incoporated to make this work.  the rest of the math seems to be in place!  just a thought.

lol
sam