Joule Lamp

[JouleSeeker]

  I'm still gearing up, and have been searching for high lumens-per-watt LED bulbs today.  I figure that it is a good goal to seek for high Lm/W output (with minimal input), and that one might start with a high Lm/W bulb to begin with...

"This Utiltech pro LED warm 450 lumens, is the LED I bought at Lowes for $9 each.  Its not wasted money because they will last a long time.  This is a nice bulb and it looks very much like the one Laser Saber uses in his video.  Lumens is stated as 450 at 7.5 watts."

This works out to 450 Lm/ 7.5W = 60 Lm/W, which is fairly low for LED bulbs actually as I look around on line. 
Tmart.com sells a variety of LED lamps, often with ratings around 100Lm/W  (they give you Lumens and Watts typically; just divide).  Below is a lamp that is quite good -- if the ratings hold true:  400 Lm/3W = 133Lm/W, about twice the Lm/W for that lamp from Lowes.  133Lm/W is the highest for any LED bulb I found... And inexpensive at $7.  I ordered a couple; evidently the demand is high and these are back-ordered.

[JouleSeeker]

Over at EF.com, my friend Slider goes the other way -- makes the coil SMALLER, and has some success. Note his compliments to Lynx also:

Moving on to building. I was intrigued by an email link to the OU forum, of the Joule Lamp by Lynxsteam and had an idea to decrease the physical sizing of the excellent work seen so far. I was besotted with b_rads work, it looked like a crystal radio was running a CFL !!! I mean, how cool is that ! (http://www.energeticforum.com/images/smilies/biggrin.gif)
Just a small step of an idea, but definitely wishing to move toward what is possible without including ferrite. Being as these are a lot like Slayer exciters, it intrigues me to know how small these can go, yet still produce useable results.
So, a previously wound secondary was selected - 4.5" length of 1/2" PVC pipe, 360 winds of 30 gauge. A toilet roll was then cut down its length and adjusted til it fit over the secondary coil. The excess was trimmed off and then it was taped to hold the shape. Onto which went 32 turns of 26 gauge.
Transistor is a D2641 Darlington power transistor, which came from a bag of trannies and still has a 1N4007 across Base and Emitter, so I just left it there.
Power is 12V, from a converted old PC power supply.

Here are two pics.
The first is an LED nightlight, which although the draw was a monstrous 500mA, lights the light brighter than when plugged directly into the wall. Sorry for the lack of light box type exactness there, which is a fine idea for comparisons testing.

http://www.energeticforum.com/renewable-energy/7051-joule-ringer-40.html

[b_rads]

This is a short video showing the current draw using different number of turns on the primary.  Can someone explain why, when I put LED's on the circuit, the current goes way up compared to using CFL's.  The current using LED's measured with a digital multimeter was 1.36 Amps.  The camera ran out of disk space while filming, sorry for the sudden ending.  Hope you enjoy.

Current Draw on LJL

Brad S   

[Lynxsteam]

B-rads

I watched your video.  I have noted the same performance.  The circuit has to be setup differently for LEDs as opposed to setting up for florescent tubes.  Florescents take 190 volts or so to start.  LED's need about 90-130 volts. 

To get the circuit started the DC voltage has to energize the primary with enough field to induce a voltage high enough in the secondary to reverse bias the transistor.  The primary has to act as a capacitor to allow the current somewhere to go.  I think this is why magnet wire doesn't work as a primary in this circuit.  You need a good thick insulation on the primary and some spacing between turns.

When the transistor turns off due to the magnetic field blocking the DC, the field collapses and you get a high voltage spike with no where to go except across the bulb. 

As frequency slows the transistor is fully on longer and the amp draw goes up.  LaserSaber shows this in his video where you can hear the frequency dropping as he screws in more bulbs.  This automatically adjusts the power to the load.  In your video you are forcing frequency lower with fewer turns on the primary thereby causing a higher amp draw until finally the capacitance of the primary isn't enough to keep the oscillation going.

To prove/disprove this idea, try keeping turns constant and add a second bulb and note amp draw.  Move down the taps and see what happens.  Try a heavier gauge stranded wire with thicker insulation.  I have had good results with double insulated stranded electrical wire used for home wiring.

Note:  I barely know what I'm talking about, hopefully an EE will straighten me out.

[JouleSeeker]

 Very interesting vid, Brad, and suggestions, Lynx.

Over here, I've plotted the data points for the calibration of my light-box-2, described earlier:
http://www.overunity.com/12340/joule-lamp/30/

The plot shows the linearity of the response in Lux for various Lumens output, with Lumens given on the packaging for each bulb.

The slope of the graph is the conversion factor FOR THIS LIGHT BOX.  You should do such a quick plot for each light box you build.  For this one, the conversion factor is 0.08(0) Lumens/Lux. 
Now I can put an unknown light source in there, such as a bulb during Lynx-Lamp testing, and actually measure the Lumens!  I will be able to tell HOW MUCH the light is actually getting brighter or dimmer as I change things like the tap on the primary.
(PS == Not hard to build a light-box! and easy to calibrate.  I think Nerzh said he is building one.)