Joule Thief

[NickZ]

 Sky:
  Although I wind both halves of my yoke cores with a secondary coil to use on gutted Cfls, I find that it not really needed for leds bulbs, as too high a voltage can damage the internal circuits, especially the small green Ac caps, or the led bulbs themselves.
  The Lyxnsteam's version of the Lasersaber's 3.0 circuit is probably the most useful circuit out, for practical lighting of ungutted Ac bulbs, like the Utilitec 7.5 watt, or the Cree bulbs, also.
  I use the ferrite beads to make my Ringer core, as they are free from old monitors and Tvs, and are available to me. Pictured below.
  Both pictures are of ferrite circuits cores used on my Ringer circuit. I normally use the TIP 3055 or similar transistor, as I've burnt out several of my 2n3055 very quickly for some reason.
  The Ringer circuit is a modified JT circuit, using only a single transistor, and no other components. Hard to beat, for simplicity, and efficiency, as well.  About 10- 12 volts works best for me, without any over heating of the transistor.
But, for gutted Ac Led bulbs 4 volts lead acid batteries, work fine, even without any JT circuit.

  I'm surprised that you are able to obtain higher light output (higher lumins) from the Cree bulb, but remember that it can also burn it out in time, if using too high a voltage.

[SkyWatcher123]

Hi crow, thanks for kind words and tips, though based on voltage drop of battery compared to straight dc loads, the digital meter is probably close to accurate.
Hi nick, thanks for sharing your thoughts.
The cree leds are not outputting higher lumens, the circuit is just using the input watts more efficiently (lumens per watt).
Certainly it seems, the 100 nanofarad capacitor in series with led bulb is helping greatly.
With this setup, i can switch the secondary output leads around and the led bulb is only slightly brighter the opposite way.
Yes, already had in mind next, to wind the ferrite yoke as you show, though am going to replicate the turns ratio of this original setup.
Which is actually a 10 to 1 ratio, 40 primary turns to 400 secondary turns.
So am going to have 200 secondary 30awg. turns per yoke core half and 20 bifilar turns per side.
Goal is to see if having the copper closer to the core and spread out more evenly over the entire core, will yield more efficient lumens per watt at 4 volts input, 12 volt testing may come later on.
I read online, that cree has managed to test a 250 lumen per watt led, so this circuit is approaching that with much less efficient leds.
Will post results when finished rewinding the yoke core.
peace love light
tyson

[Pirate88179]

With this setup, i can switch the secondary output leads around and the led bulb is only slightly brighter the opposite way.

tyson

Interesting.  I have found, using the modified flash circuit boards, that output polarity makes no difference in light output.  (Same on the Jeanna circuit) I mark all of the boards for polarity when modding them but, just for the heck of it, when I was wiring the gutted Cree bulb, i switched the wires and, no difference.

Just for whatever that might be worth.

Nick, Tyson, great work.  I might try adding an output cap as well just to see what it does with the flash circuits.  As I mentioned in an earlier post, Gadget was placing a ceramic cap in series to the transistor base and he said it helped lower the amp draw without hurting the light output.  (If I remember that correctly.)  I will look up what value cap he was using.  So much testing to do and so little time.

Keep up the great work fellows.  I am working on a new project this evening that, if it works, I will post here.  If it does work, I think it will be really cool.

Bill

[NickZ]

  Thanks Bill:
   What I may try next is to see how to incorporate a feed-back path from the secondary coil of the Ringer circuits to a drop down ac to dc adapter, like a wall adapter. Then to  send that back to the JT circuit's input run battery, to see if it can be recharged, while the circuit is providing some light output, as well.
  Why use a wall adapter? Because I've found that since they are made for 110v Ac (in my country), and my Ringer circuits are also made to light 110v ac bulbs, the wall adapters will effectively drop the HV to about the right level of low voltage Dc, (about 7 to 13v), depending on which adapter is used, in order to charge the run battery back.
  I find that I can charge cell phones, AAA to D cell batteries and 4 volt to 12 v lab that way. Why not try to feed back some juice back into the circuit run battery at the same time to see what happens. My Ringer circuit's secondary cores output over 1000v, so, there should be more that enough voltage to do so, although not all that much current.  Anyways.... just a thought, as I'm always looking for the "self runner", and not just a JT that light a couple of Leds bulbs.

[Pirate88179]

Nick:

Funny you should mention feedback.  OK, I tried an idea this weekend supplied by TK in a comment a month or so ago where he suggested using the AA battery as the core for a JT.  I decided to try this using the Jeanna circuit high voltage JT circuit.  I spent many hours this weekend building this and...I can't get the circuit to work at all!  Not even as an aircore!  (I figure if the battery does not work as a core, I can just slide in a high perm ferrite rod)  The transistor (tip3055) just gets hot and I get 0 output no matter what I use for the base resistor.

This is not any sort of proof that a battery can not be used for the core, as I think it can.  It certainly has iron in the case as we can hold them in place using neos so, it should output something even if not very efficient.  I have built many of the Jeanna circuits and this is the first time I have had this problem.  I even swapped in a new tip3055 in case the first one was bad....no luck.

Feedback:  I think that it is very possible that one could use "Inductive Feedback" (My new term) to send some of the energy from the collapsing mag. field to the battery when the circuit switches off.  Now, maybe the battery will not store this energy....maybe it will aid in making the battery last longer...or maybe it will make the battery run down faster....who knows?  I will not know until I can figure out what is wrong with my circuit!

Anyway, I think this was a brilliant idea from TK and it deserves others to attempt this to see what happens.  I will continue to try to fix my circuit.  Here are some photos....and I will post a video later when it gets uploaded.

Bill