Joule Lamp

[NickZ]

  I've also been getting my Joule Lamp to light with no connection the the pnp3906 transistor base. The transistor still gets hot at 12 volts though, and is what I'm going to try different transistors to see if I can tame the heat issue. RS has no 2n3055 in stock here, or the TIP31 so I'm going to go through my junk parts to see what I've got.

  My last video:
  http://www.youtube.com/watch?v=ZnySZEno9W8&feature=youtu.be

[Magluvin]

I found something interesting.  Playing with the larger aircore LJL I hooked up a 120 v LED bulb with no transistor.  No collector at , no base connection.  The bulb lights at .3 watt or .7 watt depending on the dim switch setting.  I will post a video later.  Off to go water skiing.  Its hot!

HOW CAN YOU GO WATER SKIING NOW!!!   LOL  just kidding  ;]

That is interesting.

Has anyone opened one of these led bulbs yet to see what the driver consists of?  I havnt gotten down to getting one yet.

Does the bulb have a dimming switch?  I just dont see one in the circuit you show.


And is that correct? A 12v batt as input? Or 120v ac?   I could see that if it were ac input that the open windings would act like a capacitor and pass current to light the bulb.

But if it is dc 12v and the coils are truly open ended and not conductive contact, then I would want to see the circuit in the led bulb for answers. ;]

Mags

[Lynxsteam]

I found a slight short between the secondary and the primary.  It is  a 9 Ohm short.  When the circuit is truly open as drawn it doesn't work.  Its starting to look like a tank circuit, with LCR.  When I put a resistor between the normal base tie and collector tie the circuit oscillates.  Or if I just alligator clip across it works.  This 3 watt bulb was running nicely on .7 watts!  This same bulb draws 1.2 amps on the mains.
I didn't draw in the dimmer switch for simplicity.  I simply toggle between 2 or 3 secondary coils (3 for high).

Here's the video

http://www.youtube.com/watch?v=RSICg7KHmBI

[Magluvin]

Hmm, have you tried the battery directly to the bulb?

Yeah sorry bout the short. It just had to be.

Mags

[Lynxsteam]

Yes, directly connected to 12 volts the bulb blinks dimly.  This is probably what starts the oscillation in the aircore transformer.  The pulsing gets magnified in the transformer to higher voltage and lower amps.  This is just one of those things you stumble onto.  May not be very useful considering a 2N3055 is about $2. 
These things spark ideas in other people who then go and build on an idea.
I am still impressed that the amp draw is so slight compared to this same bulb drawing high amps on the normal Joule Ringer type circuit.  I may get a couple more of these bulbs and see what kind of output vs input is possible on a simple 10:1 aircore winding.  Hey, at least it doesn't ring and buzz!

Nick,

If you are driving CFL tubes with the reverse bias circuit (ala super joule ringer) there is no protection for over driving the transistor.  You may be better off using a more standard approach with a resistor capacitor from the positive rail to the base, and a protection diode from negative rail to base (base emitter).  This will be more of a standard Joule Thief with the flyback coming off the collector.  If you reverse bias a npn transistor much over 4 volts it will over heat and probably burn up.  When driving LED bulbs with this circuit the voltage drop is right at 3.5-3.9 volts.  When driivng CFL tubes the voltage drop varies with voltage input. 
I think you would be very happy with a circuit suitable for CFL tubes.
Or you could modify your design to prevent the really high voltage spikes.  I was able to use the reverse bias circuit because I covered up more of the secondary with a 20:1 ratio in the original LJL posted in this thread.  OUBrad replicated this very well and showed high brightness and how many turns to use.  But to use a Tesla type exciter with its very high voltage spikes is asking for trouble with the reverse bias circuit.