Joule Thief

[Groundloop]

@lasersaber..... I'm not sure just what I've got! I hope you don't mind, I've named your circuit the "GhostLight" because of the way the LED will stay on dimly for so long with stuff disconnected, and how the circuit seems "haunted" sometimes. Spooky noises and everything, weird vibrations, tingling spines
.
I am using this term to refer to the schematic Stefan posted yesterday, with the bridge and the extra wire....

I still can't seem to find the HV but I did get all the other effects, as I noted before.

And also, my transformer is different so that might be why no HV for me... yet.

After I saw scratchrobo's video and thought about Hoppy's post a bit, I went ahead and removed the bridge from the GhostLight circuit and made another change or two... and got the thing working well as a low-voltage JT... but with a twist.

I've blown 2 x 2n2222 transistors and a white LED since starting with this thing, but not the 3055 yet. Have had some hot transistors when running on the 5v supply, but now  I think I've got this version stable.

Next will be to try to get a transformer more like the spec, and keep searching for the HV that is in there somewhere.

Here's a video I just shot..... a gang of thieves!
Still uploading, will be done in a few minutes.
http://youtu.be/LQe0In73pbY

TK,

Looking forward to your video. The video is still being possessed by YouTube. :-)

I'm currently working on a scaled up version of the circuit. Did find a 70 Watt laminated
soft iron toroid core. It is a factory made transformer where I took of the low voltage windings
and made my own 29 turn new primary. I will still use the factory made 230VAC winding as
a lamp driver output. I will use the NTE2354 as a power transistor with a MJE340 as a driver
transistor. Sorry that I have no photos but my digital camera did choose to go bust.

GL.

[hartiberlin]

Lasersaber just posted a new circuit diagramm at his site.

Here is a copy for backup reasonsn.

Have a look also at his site:

www.laserhacker.com


Many thanks Lasersaber for your great work !!

Regards, Stefan.

[hartiberlin]

Yes.... there is something very strange going on. This circuit caused me to check the voltage between my good Earth ground and the house wiring. Using the Fluke 83 DMM on AC voltage setting, I get a rock-steady 57.4 VAC between the Earth ground and the ground leads of my test equipment!! There is no "current" available though, it will just barely light a green LED to a slightly visible glow. Scoping the "earth ground" as a signal source I get a beautiful sine wave at 60 Hz, p-p around 150V.
....

Just a quick note, before I go to bed...

Well my Cable TV connector bush ground has also such a potential voltage difference to my
other 230 Volts AC house ground line...or versus the heating pipe metal...

This is also why so many TV tuner PC cards blew, when connecting them to the cable
connector bushes.
There you can always see a little blue spark jumping, cause these cable TV ground potentials
are different than the PC power supply ground line potential....

So it seems it could be some kind of voltage potentail equalisationcurrents flowing there...

Have to check into this further and have to see, if I could run a Joule thief from this to charge batteries...

Regards, Stefan.

[NickZ]

  All that you are noticing is the stray AC capacitance in the house wiring, going to the outside ground connection. Turn off your house's breakers, and that stray voltage will dissapeare as well.  Just try it...
  If Lasersaber is finding something other than this stray AC capacitance from the house wiring, then this needs to be looked into.

[Groundloop]

Hi,

I have now built and is starting to test my new larger circuit. The oscillator self starts because
of the 100K bias resistor. Oscillator is running fine on many different input voltages. This version
has higher losses and uses more current just to idle. But I will try to light some larger bulbs with
this circuit. My initial calculations shows me that the circuit will waste approx. 0,5 Watt on idle.

[EDIT1] I have added one more BY255 and got the idle current down even more. This circuit has
now been tested on a 230VAC 3W LED lamp and that worked. I did also test a 8W LED lamp but
the lamp did only blink on and off. I tested on a 11W CFL and the lamp did not light. My conclusion
is that this circuit version is wasting too much power and is not good at all. The soft iron toroid core
is wasting too much energy and is probably way to large for a practical led lamp driver.

GL.