Quote from: avalon on May 09, 2014, 01:53:00 PM
I am speechless over here.
I do no know for sure what I have found, but my 3v flashlight V2.0 went crazy.

The power supply is at 2.59V, but I now have over 1 kV on the secondary! (540V in the picture)
It is, obviously, a resonance, but at this stage I am inclined to think it is a magnetostrictive resonance excited by an external (primary) magnetic field.

Anyway, at 230V ~ RMS it looks like I have enough power to light up a relatively big CFL light bulb while having the circuit self-sustain.

~A

Good job
I got smoke from my scope

Quote from: avalon on May 09, 2014, 02:43:36 PM
I know, right?
Anyway, I am re-building the circuit again. At some stage the voltage increased even further and the coil started 'dancing' around the table. Then (with a big spark) both LEDs, MC34063, and a couple of capacitors gone up in smoke.
BTW, the power supply is limiting the current to 100 mA max so all that power didn't come from the PSU.

~A

Well that is more interesting if the current was limited, but what was the value and voltage rating of your storage capacitor, they can store a lot of energy over time by charging up despite the current limiting of your supply, especially if you lost voltage feedback to pin 5, the duty cycle would increase and voltage on the rail would soar on that capacitor. Can you reference the schematic you are using.

Also a good forensic analysis of your coil winding technique would be an important post mortem so that it can be replicated.

Quote from: Vortex1 on May 09, 2014, 03:05:59 PM
Can you reference the schematic you are using.

The schematic is on the page 5 of the Onsemi datasheet. The only difference is that I use a variable resistor on the feedback (22K) and there is the secondary coil.
The coils are 37mH primary and 57 mH secondary.

~A

Hi all,
beware of the probe setting 1:1 or 1:10: i saw plenty of volts just because probe was set to 1:1 and scope to 1:10.
For the moment no positive result or my own clone of this LED OU circuit: it works fine as a standard step_up circuit.
When the output 1000µF is removed, the 34063 is always working at its maximum frequency: why removing it ?
Regards.

Quote from: avalon on May 09, 2014, 02:43:36 PM
I know, right?
Anyway, I am re-building the circuit again. At some stage the voltage increased even further and the coil started 'dancing' around the table. Then (with a big spark) both LEDs, MC34063, and a couple of capacitors gone up in smoke.
BTW, the power supply is limiting the current to 100 mA max so all that power didn't come from the PSU.

~A

Perhaps a bigger/larger load? Sounds promising though. keep up the good work!