Joule Lamp

[NickZ]

  That's towers, like in mini towers,  not towels.  I saw the error but couldn't edit it.
  So, towels it is... 
                           
   

[JouleSeeker]

Today I added a 6th LED bulb (of the corn-cob type I've described earlier) and the Lumens/Watt went UP. 

Vid posted here:  http://www.youtube.com/watch?v=b7lUFPh4ttM&feature=youtu.be

Text:

  Lynxsteam has previously noted in his air-core experiments (with the Lasersaber SJR 2.0 circuit) that when one adds an additional LED bulb (after the 2nd one), the DC input current drawn does not go up by much.   Here are some data taken at 11.0 V DC input (and I have added three ferrite rods near the central axis of the "cranberry" coil as previously described):

4 bulbs draws 342 mA (so 3.76 W) and produces approx 374 Lm, so 100 Lm/W

5 bulbs draws 344 mA (so 3.78 W) and produces approx 420 Lm, so 111 Lm/W

6 bulbs draws 353 mA (so 3.88 W) and produces approx 443 Lm, so 114 Lm/W,
this is the case shown in this vid.
114 Lm/W is well above the Lm/W I expect from these LED bulbs; so this result adds to my interest in this ongoing study.

When I hook up my DSO, I see a frequency (output) of approx 21 KHz, and the voltage across the bulbs (i.e., across the power-strip plug feeding the LED bulbs) is 460 Vrms -- which is pushing my  DSO I think.  This is tapping into the 75th winding of the primary; so I would expect around 130 V AC output just based on the ratio of secondary to primary windings... so I'm trying to understand the output voltage.  My little DMM set to 750 VAC registers "overload" when I try to read the output AC voltage with it.

Again, I thank Lynxsteam for excellent observations and ideas and Lasersaber for the basic circuit being scrutinized!
______________

If anyone can explain the observation of 460 Vrms on the output, and the overload on the little Cen-Tec multimeter, I would appreciate it!  An accurate measurement of the output voltage (and current) is one way to measure the output power, so I'd like to understand the output voltage!  has anyone else measured it?

[Lynxsteam]

Joule Seeker,

Congratulations on 114 L/W!

If this were a standard transformer then you would expect AC voltage input to be transformed by the turns ratio.  12 volts ac in 120 volts ac out.  I do use this for calculating turns ratio, however since this is a blocking oscillator you will get high voltage spikes.  These spikes can be further accentuated by voltage amplification in the secondary.  In your "cranberry" model LJL 4.0 the voltage amplification is double.  That voltage gets converted to amperage in the load. 

And it just so happens the voltage drop across diodes (LEDs) is between 3-5 volts, perfect for reverse biasing a transistor which is basically the breakdown voltage for the base to collector.  The brighter we light the LEDs the more we reverse bias the transistor and the lower the frequency and higher the output (self regulating load).  So by increasing voltage at input you are increasing the bias voltage closer and closer to 5 volts.  Push the DC bias voltage beyond 5 volts and you will probably fry the transistor.

On my latest LJL 5.0 the voltage amplification in the secondary is even higher.  I have designed in self capacitance "ala Tesla" and now I am pushing the transistor and LEDs to their limit.  The transistor gets hot and so do the LED bases.

When you run on 11 volts you are probably avoiding this wasted heat energy.  We can de-tune the coil arrangement to run on 12 volts in that sweet spot you are finding.

I am just about done detailing the design, schematic and instructions.  I could wait for more testing or turn everyone loose replicating what I have so far.

[JouleSeeker]

Thanks for the input, Lynxsteam -- makes sense.

Also, Slider writes by email, it may partly be a "Tesla tower phenomenon."  I would like to learn more about that!

[Lynxsteam]

It is a Tesla tower we are talking about here in this thread.  It is a little different because we are confining the field instead of allowing the field to radiate as RF.  A typical Tesla Coil for making plasma and wireless lighting has a very small primary at one end and the voltage amplifies through self inductance up the coil to very high voltage and miniscule amperage.  What we want is about 120 volts and 60 ma per bulb for a 7.5 watt LED.  Your 353 ma is on target for 6 bulbs.  The secondary spikes may actually be 480 vac and the amps are only 9 ma.

Its similar to the way a Bedini SSG can charge a 12 volt battery with back emf spikes of 200 volts.  The battery clamps the voltage to 12 and the spikes get converted to amps at 12 volts across the plates.

This is part of the reason mixing LED bulbs doesn't work so well.  It is like trying to charge a 6 volt and 12 volt battery in parallel.  One is going to consume more power than the other.