Joule Thief

[NickZ]

  MileHigh and all:
  When using the 3.0 Ringer inverter circuits I've burnt out several of the small ac caps that are part of the internal circuitry of the ungutted (standard) Cfl, or Led 110v AC bulbs. Their long term lighting duration is not long at all when using these circuits on 12v. However, once the AC bulbs are gutted they will continue working with even less draw than seen previously, and there are then no caps to burn out. Although these internal green Ac caps crack and blow, even before the leds bulbs are affected. It makes these home made inverters for standard led or cfl use, not all that practical, sometimes, especially at higher light outputs and long term use.
  At $10 or more for the led bulbs, I don't want to see how long it takes, or how many of them will blow. Good thing that they can be gutted, and still reused.                 
                                                                                                      NickZ

[gyulasun]

....
So my first thought would be that when you run the circuit the capacitor voltage increases to a point where you reach a balance point.  The balance point happens when the secondary coil pushes current clockwise through the capacitor (at some high DC voltage) and through the LEDs - that lights the bulb - and then the charged capacitor discharges counter-clockwise through the coil and backwards though the LEDs - that lights the bulb also.

Hi MileHigh,

considering your first thought description on the clockwise current direction through the capacitor and via the LEDs causing the LEDs lighting: this would mean a forward current direction for the LED as per the drawing symbol in Tyson's drawing: this is okay for me.
However, (in the moment the secondary coil does not drive any current) and when the charged capacitor is able to discharge counter-clockwise via the coil and backwards through the LEDs then the LED cannot give any light because LEDs do not light up in the reverse direction. 
It would be good to know Tyson's gutted LEDs are in series with all in a uniform forward direction.
Here is a link to a (not gutted) LED lamp type what Tyson gutted (I asked him earlier what he meant on a gutted LED, waiting for answer):
http://www.cree.com/Lighting/Products/Indoor/Consumer/40W-Replacement-Warm-White-Type-A-LED-Bulb      and I wonder how many individual LEDs are in such bulb.

rgds,  Gyula

[MileHigh]

It's the Wild Wild West out there boys!  Everybody is selling LED light bulbs and changing lighting infrastructure!  Over the next ten years all of the lighting will be done with LEDs.  Trillions of dollars are going to be spent to convert everything over!  Sign up for my $250 seminar to learn how to cash in!  lol

Yes we will see if Tyson can check the diode directions.

You are probably right that the reversed-biased LEDs won't light.  In the back of my mind I thought that I once heard that they do light but I have no real knowledge.  That's a good test right there for fun:  Torture an LED and slowly adjust the reverse bias current and check for light output.  That would call for a slightly sophisticated test setup but nothing too crazy.

[SkyWatcher123]

Hi folks, thanks for all the replies to my questions, very happy you folks are giving it some thought.
Hi nick, yes i did think that already, that maybe my coil configuration is not so good, even if that is true, with this capacitor, it is putting out a heck of alot of light for the 1.1 watt-1.4 watt depending on base resistance used.
And i still am interested to learn why the circuit is operating the way it is.
It is a Cree gutted 6 watt (40 watt equivalent) led bulb, just the leds, which i think are royal blue with small phosphor lenses over each led to create the warm white color and all are in series.
Hi milehigh, as far as I am aware, leds cannot light in reverse direction, though maybe they can when in the splitting the positive type position, tesla switch style.
That is one of the ideas i was throwing around, though at this point, the ringing and each quick reversal of polarity, may be causing forward conduction of led diodes upon secondary coil collapse since it is observable, that the led bulb fades out slower once in awhile, or rings down.
Here is a pic of the gutted bulb, maybe i should make a video showing how this circuit behaves lighting the bulb using manual pulses, it may help to comprehend more of what is going on.
peace love light
tyson

[Pirate88179]

For what it is worth...all of my gutted Cree 60 watt equiv bulbs will light with reversed polarity.  They are very dim but still light.  (maybe 10% of correct hook-up)  Mostly, I am still using the flash board type JT for these.

Tyson, you took your glass off right?  If I remember one of your photos correctly, I think there are like 20 LED chips as you described, maybe more in the 60 watt version but I have not taken the glass off of any.

That is all I can add at this time.  Very good discussion here...I am learning a lot.

Bill