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

[amigo]

@sanmankl

You really need to measure the hFE of the transistors you use because it makes a difference (bigger the better). I've noticed 2N2222 works better than 2N3904 though your current consumption is way too high, even 30mA is more than a standard Joule Thief circuit and not even my modified version which runs at about 12-13mA.

Actually in my latest incarnation of the circuit, there's no need for 1K resistor because I am using a 2N7000 N-FET which seems to be giving much better amplification than standard NPNs. Circuit uses standard 20 turns of bi-filar 30-34 AWG on a miniature ferrite bead/toroid core. Also, the current consumption has fallen down below 10mA (around 8-9mA depending on the oscillating frequency).

I believe our goal here is to get the LEDs to be the brightest and connect as many as possible, with the lowest current consumption thus prolonging the battery life.

[turbo]

Marco,

Thank you for the break-out of the methods, I just wonder why some people still prefer spark gap over pulsing the coil...there's lots of talk about hydrogen spark gaps, etc...

Also, if we were to encase the micro-gap into a glass ampule and (partially) evacuate the air wouldn't that prevent atmospherics from interfering with it?

Well i have also used NE2 bulbs as THE spark gap.
I just discharged a full capacitor into the bulb.
This creates a violent blue flash rather then an orange glow, but the bulb didn't like it...it's now in the damaged components collection.
Anyway at high voltage it's a dielectric thing, you can pick up a NE2 bulb and it will even glow brighter as the energy flows from your finger thrue the glass and into the bulb, or the other way around.
Much like those plasma globes with all the sparks inside.
Those are glass too.
But, i think indeed it is best to use open air spark gapping as i also think it needs to be in contact with the environment.
However if we go into the Radiant energy area, it will just penetrate about anything.

M.

[turbo]

Hi,

This circuit is driving 9 5mm white LEDs. No matter how many LEDs I put in (I only have 9), the amp remains the same. I would have thought that the more LEDs I hook it up, the dimmer it will be. It doesn't. Is this normal?

Hi

Do you think that is normal? i think it's magic.
I have seen the same thing many times....one time my whole room was filled with UV solar tubes all in series and it did not matter how much i connected they all lite up..

It looks as if the power feed only acts as a catalyst to establish the connection.
My next step is to focus all that light onto a little solar cell to see what it put's out, as you know solar cells are only about 15% efficient.

M.

[broli]

Hi

Do you think that is normal? i think it's magic.
I have seen the same thing many times....one time my whole room was filled with UV solar tubes all in series and it did not matter how much i connected they all lite up..

It looks as if the power feed only acts as a catalyst to establish the connection.
My next step is to focus all that light onto a little solar cell to see what it put's out, as you know solar cells are only about 15% efficient.

M.

This is quite a coincidence since I was trying to get the following post through, but it was giving me a hard time...

This might sound funny but maybe you could try and hang a small sollar cell, like those found in calculators,above the leds (works probably best with white leds). If this led is bright for a very longer time than it should be then maybe just maybe you can extract an intresting sized amperage from the solar cell.

Good to hear someone is actually going to give it a go.

[Groundloop]

Hi,

I build the circuit following Groundloop's schematic (trifilar) and after testing it with various npn transistors, this is what I found out.

This circuit is driving 9 5mm white LEDs. No matter how many LEDs I put in (I only have 9), the amp remains the same. I would have thought that the more LEDs I hook it up, the dimmer it will be. It doesn't. Is this normal?

- 2N3904, lowest current comsumption at 30mA. LEDs. Dimmest but still bright enough to see "white" and not green tint when it's not enough power.
- 2222A, higher comsumption at 65mA (tried another 2222A, it goes up to 75mA)
- BC548, 2nd lowest at around 60mA
- BC337, highest at 80mA

Frequency as suspected, changes with various transistors.lower frequency, lower current consumption.

Right now, I'm putting the circuit (2N3904) to on test-run for time until battery runs flat (it's a NmH 1V2 batt). So far, it's running for 4 hours. Too early to tell anything.

Cheers, cp

@sanmankl,

No, that is NOT normal. Can you be more specific? When you say current did not go up did you mean the input current from the battery to the circuit? Also can you give some values to your measurements?
Measure the input current in mA and then add some more LEDs to the output. Did all your LEDs light at the same level? Etc.

Groundloop.