Joule Ringer!

[Lynxsteam]

Here is a chart for the four Joule Ringers Circuits I have made. 

SJR2.0 9:90 is an E core type using telephone wire.  Very close to what LaserSaber designed
SJR 2.0 16:200 is an E core type I made with bell wire and 30awg magnet wire
LJL 36:368 is an air core type using 12 awg primary and 20 awg magnet wire
LJL 68:680 is an aircore type using bell wire primary and 30 awg magnet wire
Bulbs were Utiltech 7.5 watt warm LED bulbs
Amps were measured into the DC leads to the primary
Input Voltage was 12 Volts
Brightness varied. Best for the Modded ECore.  Aircore types and SJR 2.0 about the same, at @60% brightness.
My conclusion?  Insulation on the wire is important.  Magnet wire is not the way to go even though it works.  This thicker insulation provides capacitance between turns which helps with the oscillation.  E-Core entraps almost all of the input energy and the single transistor efficiently converts to AC.  Very little wasted energy in this circuit.

[lasersaber]

My conclusion?  Insulation on the wire is important.  Magnet wire is not the way to go even though it works.

You got it.  I did lots of trial and error coming up with this design and one thing I learned was that the insulation on the wire is very important.  I want to explore this effect more.  In the past I have seen joule ringers running lights with the base on the transistor completely disconnected.    If you then connect the base to a large antenna and ground one of the free primary wires you can achieve some pretty strange stuff!  I plan on posting a video showing this kind of stuff soon.

[Lynxsteam]

I have two more things to try.  This may help confirm the importance of a heavier insulated secondary. 

I am going to make another air-core LJL with 20 awg bell wire on the secondary and 14 awg heavy insulated primary wire.  If there is no difference in power consumption we can attribute it to the low inductance of the air core, if there is a lower power consumption we can attribute it to the insulated secondary.

The other thing I will try is an aircore coil wound on a 1" ID x 2.5" OD, (Brooks coil).  I will use insulated bell wire, and wind the secondary with a Tesla Parallel winding on the outside of the primary and one without the parallel winding.

I may quietly try some of your magic tricks but wont still your thunder.

[conradelektro]

I still have not found the right core. Today I tried a toroid which I had from earlier Joule Thief experiments. I could not properly light big LED bulbs (5 Watt or 9 Watt), they blinked. Also unmodified CFLs only lit up a little. I could see a little light with incandescent bulbs, but the circuit behaved strangely.

The frequency was about 28 KHz. The circuit worked very well with a 2.5 Watt LED lamp (see the attached photo and circuit). I tried 3 turns, 6 turns and finally the 9 turns primary. For 10 to 14 Volt operation (12 Volt accumulator), the 9 turn primary worked best.

The little modifications I made to the circuit (copied from Lidmotor and Peanutbutter29) kept the transistor very cool and ensured start of oscillation at power on (down to 3 Volt).

I also tried a transformer core, but it must be a HF transformer core, because the frequency was about 110 KHz. Also, only the 2.5 Watt LED lamp worked nicely. I had the core in my collection since many years and do not know its specification any more.

It seems I need a core with a higher inductance to match the performance I see from Lasersaber's or Lynsteam's experiments. I guess the frequency has to go down to a few KHz and power transfer (through the transformer) up to at least 4 Watt in order to make stronger LED lamps and unmodified CFLs work.

I did not succeed to achieve a low frequency and high power transfer with air cores, therefore I am now exploring ferrite cores and toroids. It is difficult to get the US-cores and US-transformers in Europe, so I look for good replacements available over here.

Greetings, Conrad

[conradelektro]

I think I found a core comparable to Lasersaber's core:

Lasersaber recommends: http://www.surplussales.com/Inductors/FerPotC/FerPotC-2.html

I found: http://at.farnell.com/ferroxcube/etd54-28-19-3c90/core-half-etd54-3c90/dp/3056429 (two required)
             http://at.farnell.com/ferroxcube/cph-etd54-1s-22p/bobbin-etd54-1-section-22pin/dp/137029 (bobbin)
             http://at.farnell.com/epcos/b66396a2000x/clip-etd54/dp/1422754 (clip)

May be some knowledgeable people could comment?

Greetings, Conrad