Joule Ringer!

[TinselKoala]

  Tinselkoala:
   You might want to try to change the polarity of both the primary as well as the secondary wire connections, as it can make all the difference. Mine would not even start until I changed the polarity.
I also have the same problem of not being able to connect the bulb back to the transistor, without it ringing, or not even starting.  But the way I have it now it does not ring at all, but I still need to put my hand on the bulb to get it to light more brightly. So, I'm still working on that.  Any ideas are welcome.

   NickZ

I've tried all four possible combinations; two work and two don't. The two that do, I can't tell any difference.

But... I was able to get my NErattler that powers the Ring Oscillator to start oscillating right away when power is applied, by adding a 10K resistor between the base and the positive rail. It also makes it run down to lower voltages before shutting down. It starts reliably now, even with the lowest applied voltage.

[NickZ]

  Tinselkoala:
  Well, that's good news.
 
   I also have some good news today. I finally got my Joule Ringer to work like it should, and produce the brightness that I imagined was possible. I redid all my connections, and added an earth ground to the positive rail. Now I can connect the Cfl bulb's negative back to the transistor collector, as Lasersaber mentioned. and don't need to hold or touch the bulb to have it fully lit.  The output is so bright it's practically blinding.  I'm lighting my whole house, from a single 65 watt Cfl bulb now, and it's light even goes out across the back yard.. Unbelievable!
The pictures don't do it justice...

[Groundloop]

I've tried all four possible combinations; two work and two don't. The two that do, I can't tell any difference.

But... I was able to get my NErattler that powers the Ring Oscillator to start oscillating right away when power is applied, by adding a 10K resistor between the base and the positive rail. It also makes it run down to lower voltages before shutting down. It starts reliably now, even with the lowest applied voltage.

TK,

I also found that the circuit did need a low resistance load to start oscillating. But I wanted the circuit to start anyway
because some of the loads I needed to test on was a high resistance load. Some of the Osram LED bulbs sold
here in Norway does not have any DC resistance at all, and probably have a capacitive input. The circuit will not
start at those bulbs. So I used a 3K3 Ohm resistor as bias and later found out that I could put a LED in series with
the resistor to get a nice ON indicator also. Works great, my circuit start up every time if the input voltage is above
1 Volt. Once started, I can run the circuit at input voltages as low as 0,32 Volt. And I'm currently testing how low I can
go before the circuits stops. At 0,32 Volt circuit input, I still have a good green light in my power on indicator LED
and my output LED lamp still has two of the LEDs lighted, but very dim. I'm working on a new circuit and coil where
I will try to make a night light. I will be using one LED only and run the circuit from a 50 Farad 2,5 Volt capacitor.
By using the known "capacitor self recharge effect" I will see if the circuit can light a ultra bright LED (dim) for a
very long time since the circuit only need a very low voltage to run and a very low current to keep the LED light.

ADDED: Could not get this circuit to perform well. Seems to me that a smaller diameter coil with less copper wire
is NOT the way to go in this circuit. So I just connected my 50 Farad 2,5 Volt capacitor to my first version circuit and
will see how low voltage the circuit can run on.

ADDED: My first version circuit did run until the voltage over the capacitor went down to 0,188 Volt.
I have attached the circuit drawing I'm using in this test.

GL.

[Groundloop]

  Tinselkoala:
  Well, that's good news.
 
   I also have some good news today. I finally got my Joule Ringer to work like it should, and produce the brightness that I imagined was possible. I redid all my connections, and added an earth ground to the positive rail. Now I can connect the Cfl bulb's negative back to the transistor collector, as Lasersaber mentioned. and don't need to hold or touch the bulb to have it fully lit.  The output is so bright it's practically blinding.  I'm lighting my whole house, from a single 65 watt Cfl bulb now, and it's light even goes out across the back yard.. Unbelievable!
The pictures don't do it justice...

NickZ,

Very nice done! :-)

How much input voltage and current do you need to keep that CFL fully bright?

GL.

[NickZ]

   Right now I'm using a wall adapter with 14volts, and 1 amp input to the circuit.
  Since my analog meter is not working on the higher mA setting I'm not sure what the circuit is drawing. Tomorrow, actually today as it's 1am, I plan on connecting this set up to 4 small motion light solar panels, that are all connected in parallel. They probably output about 300mA, or so, at 11.5 volts, in full sun.
The main thing for me is that the MJE 3055 is stone cold, as if no current is going through it. So, I would think that it could take even higher voltage/current levels, than I'm giving it now.
I don't know if the bulb is fully 100% lit, but it is certainly bright enough for me. Since I have several 65watt Cfls, I'll see just how many I can light with this system as is. There is slight ringing now, even though I put masking tape between the two yoke halves. But, other than that,  I'm a happy camper now.