Efficient Boost Oscillator Led Lamp

[nul-points]

hi skywatcher

ok, so you've completely removed everything from the internal circuitry except the LED plate, which can now be shown on a schematic as having a polarisation (ie. +/- leads) and probably some series/parallel connection of the LEDs to give some idea of the likely operating voltage

in which case, a buffer cap is likely still to be necessary, a non-polarised unlikely to be better, but there might be some scope for trying a different value buffer cap.  a Schottky diode might give a small benefit over a 1N4XXX type.  a FW bridge of Schottkys might give improvement over the single 1N4XXX type

in future experiments, its worth winding 2 same-length secondaries (ie. trifilar with the primary) cause then you have the choice of connecting the 2 'secondaries' in parallel (and used like your trafo here) or in series if you want to use a centre-tapped 2-diode FW bridge


hope this helps
np

[SkyWatcher123]

Hi nul-points, thanks for all of your helpful sharing, I will keep those things in mind as I experiment.
I'm finding some interesting observations as I experiment with different lighting devices.
I decided to make a test of direct current with 555 timer pwm, powering a homemade led bulb, using 4 sets in parallel of 5 in series surface mount leds, ones salvaged from those T8 led flourescent replacement tubes.
I set the voltage so they were under voltaged, so as to boost efficiency.

Though they were bright, somehow they just did not have the same area or surface lighting, that I can achieve with the oscillator powered devices.
Plus, with the oscillators, I can seemingly use less power to achieve a better visual perception of lighted surface area.
I'm not sure why, though my tests seem to show this.

Maybe it's the fact that the 555 timer, at full open, gives full duty cycle, where as with the oscillator, it is always at a certain lower duty cycle, creating good surface area lighting while minimizing input power.
And I changed circuits at the moment, I'm using that large bifilar coil with ferrite c-core in the joule thief configuration.

Though not common grounded, the gutted led bulb across the capacitor, with diode off collector into 470uF electrolytic capacitor and other end of cap and coil to positive of primary coil, bedini style really.
At around 700 milliwatts, I'm getting the about the same or better area surface lighting level, by my eyes perception, that I was getting with the 555 timer, direct current pwm mode, though that was using around 1.4 watts input.
Definitely interesting results.

peace love light

[nul-points]

sounds good SW, enjoy your investigating - and enjoy the holidays, too!    all the best, np