Akula eternal lantern 4

[MenofFather]

I saw your schematic with extra stuff removed. ... L2 is connected backwards (38 winding) and the LEDs are backwards, they are driven from negative potential to ground

If i good understand, L2 conected not in backwards, two coils wound in same direction. LEDs is feeded from back EMF. C28 capasitor can be leaved.

[d3x0r]

If i good understand, L2 conected not in backwards, two coils wound in same direction. LEDs is feeded from back EMF. C28 capasitor can be leaved.

I see; I thought the diode symbol was so if currnt ran backwards (against the arrow) then it lit up...

Fasltad sim

Hmmm... reversing the coil still sort of works; you sure it's the back emf and not the forward emf that's supposed to be used?



I mean for the ground-deform to happen to erly switch the transistor, the first action of L2 must be to conduct away from the emitter... and therefore into the LEDs which is backwards for what you have by common notations.

[d3x0r]

Extended simulation, carried on using BEMF... it breaks the sim and I cannot identify why; or how to fix the error.


T0 is 0 of reset,
T1 is all power/ground switches closed, drive switch open
T2 is opening all power/ground switches and let it settle; small capacitances balance
T3 is with SW1 closed, building current (yellow traces)  in the primary coils
T4 is well, really just the very last instant... the voltage of the low rail goes to positive, there's no KA that transpire to actually fill the low side cap to that voltage..  the low rail should still be grey, but it is 8.7V which is higher than the input voltage...


I guess you can add capacitance to a gate to get the double pulsing after turnon also... which is the variable cap... I keep blaming inductance because that's what my situation was


when the simulation time is really really low ( it takes a long long time to charge the caps, so there's extra power/ground switches to fill to initial condition faster....


The sim... Edit: fixed initial state of export

[avalon]

Hi Avalon,

diagram of your setup ?

Thanks

Look it up in 3V OU flashlight thread

~A

[MenofFather]

I see; I thought the diode symbol was so if currnt ran backwards (against the arrow) then it lit up...

Fasltad sim

Hmmm... reversing the coil still sort of works; you sure it's the back emf and not the forward emf that's supposed to be used?



I mean for the ground-deform to happen to erly switch the transistor, the first action of L2 must be to conduct away from the emitter... and therefore into the LEDs which is backwards for what you have by common notations.

I now have problem wiev schematic, but I shore, that LEDs is feeded from back EMF. And self running is feeded from normal transformation.
Here acording to T1000 citation LED is feeded from Back EMF

"The idea behind is - after first pulse you let BEMF to reach maximum and hit with current from transistor again on its peak time so the magnetic domains when coming back wil be supported by current and voltage from BEMF WILL ADD with current from transistor so C28 cap will charge from much higher voltage on positive half-cycle. The LEDs are powered from BEMF."
http://www.overunityresearch.com/index.php?topic=2358.msg39158#msg39158
Only important part can be, that capasitors must be 50 or more volts that 1000 uF.