[NewCandle] Long and winding LED road

[Nick Reiter]

Hello all,

Sorry this took me a bit longer than I had originally
thought...

I had to stockpile LEDs week before last, finding I
didn't really have enough left of my old ones.  So I
bought 20 GaN superbright greens, 20 superbright
whites, and I found an exotic orange style that called
to me, so I got 20 of those.

I spent part of this past weekend trying to get an
oscillator circuit running ala the schematics shown
here and there on Hartmann's site.  No luck - I know
they work, but it may be I don't have the right turns
ratio on my loopstick or something.

At any rate, I went to a stopgap plan B.  I pulled out
my old two transistor flyback oscillator and clipped
it up to a 9V battery.  (Uses about 2.6W...Ah, now we
are making Tesla wireless power:)  I can use a simple
foil ball capacitance to couple from a foot or two
away, and light at least the string of 20 greens. 
Yeah, when I touch the pickup, they get bright, and
current through the LEDs goes up.  I have a simple
d'arsonval milliammeter in series with the LED string.
 I can say that I see NO problem or weirdness
measuring this in such a way!  When I switch off my
"Tesla station", the LEDs stay lit for a few seconds
due to the large capacitor across the AV plug.  The RF
from the oscillator is not bothering the meter, so I
believe what it tells me.

At the same time, I've watched the evolution of the
discussion threads on Hartmann's site, and they seem
to be de-cohering (which oddly is not necessarily a
bad thing - its how serendipity happens)

At this point its hard to say just how one can
replicate anything, as the drivers and configurations
are constantly morphing.  Single transistor step up
oscillators don't seem to have any novelty, nor does
the Avramenko plug part.  So I guess for now, I ought
to focus on the properties of LEDs in long series
strings.

I've been making some initial measurements using a
calibrated Si photocell, and have been so far just
re-acquainting myself and sanity checking...

1.  Luminous output from an LED does appear to be
lawful and linear to current, whether I have a short
string of LEDs (3) or a string of 20.  So thus up to
20 LEDs do not seem to be violating any rules.

2.  What is linear with current is NOT linear to the
eye!

3.  The voltage drops across the LEDs running off of
the AV plug / pickup have been lawful, and for
comparison, I took measurements whilst powering the
LEDs from a regulated 0 to 500V DC supply.

4.  Now here may be a clue.  The voltage drop across
the LEDs does vary slightly with current!  At 1mA,
average V across the green GaN LEDs was 2.83V.  At
.1mA it was 2.60V.

5.  With these LEDs, (the greens) even 500 micro-amps
/ .5mA is resulting in brightness probably equal to
what the fellows on Hartmann's list are seeing and
photographing.

6.  The LED conversion efficiency - roughly calculated
from taking LED V x current versus cell power with the
Si cell, given a rough classical conversion 
efficiency of 20% for the cell, seems about 65%.  So
unfortunately, calling OU using photocells as power
sources for the oscillator is going to take some
BOOKOO >100% anomaly region.

I will add the whites next, and then oranges, to see
if anything goes "weird" when you get to higher series
counts.

If nothing else, the lights are real purty.

N



The Holy Grail 'neath ancient Roslin waits.
The blade and chalice guarding o'er Her gates. 
Adorned in the masters' loving art, She lies;
She rests at last beneath the starry skies.


      ____________________________________________________________________________________
Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping