I have noticed that you people still have a problem understanding the jest of this circuit. That is to be expected as it operates the opposite of most. First it is important to see that the LED’s are only lit when the coil is off so we build a magnetic charge for the coil at as low of a pulse as can be used. Second all of the LED’s do not draw forward current as they are not using that scope of the wave form only the collapse . The LED’s are driven by a bifilar coil which is far more efficient than regular coils but more difficult to use as they do not take well to multiple layers but the extra layers are not needed for it to operate. I will say that it is possible achieve great outputs with little input as per normal operation. So you can take a .1 – 1 uS pulse and feed it to the circuit at resonance of the coil ( made up in the off time ) where the greater the magnetic charge the greater the BEMF that will ensue. The trick is to have the field pump as many off cycles as possible that is where the next part of this device I will use ATTINY85 to generate pulses controlled by the following parameters:
- Pulse width (coil charge)
- Coil discharge (Lite LED’s)
- Read voltage on the solar cell
- Wait till cell voltage drops slightly and store that time in memory and then goto 1. ( use that time to loop to 1. – pulse width ( LED’s are off during coil charge )
The result will be 80 – 100% rated output on the solar cell and as a side note the attiny85 uses 1 uA current draw during it’s normal operation. The reason I say cheating at math goes back to the load where the current draw of the LED is not important but the supply voltage is what it looks at so 200 white LED’s should take 12.5 – 13 Watts but they only use .25 Watts That is how I cheated math.
P.S. The computer chip is only needed for the solar cell output gain not to run the LED’s.