This modification takes the Double
Barrel Energizer - Luxeon Star Modification (above) I completed
earlier and adds an LM2621 DC to DC Converter
modified for variable voltage output. LM2621 boards are available
for purchase from National
Semiconductor. Luxeon Stars are available from Arc
Flashlight. The advantage of this arrangement is that with
the LM2621 regulator board in place, the light will not dim
(or will dim a minimal amount) as the batteries are depleted.
The board and the light will simply shut down when the batteries
are completely drained.
Here are a few pictures (above and below) of the assembled
system. The Luxeon Star needed to be trimmed around all of
the edges. This was completed with a Dremel. The board experienced
the same. This time I mounted the 100K potentiometer on the
top of the board and a 47K resistor on the side (in series
with the potentiometer) to reduce the vertical depth of the
board.
Below and to the left you can see
that the wires from the board are plugged directly into the
bulb socket as was done before with the previous DB Mod. The
board rests in the head upside-down with a coil of wire between
the board and the Luxeon Star. It was necessary to cut a groove
in the white plastic switch housing of the light in order to
make room for the potentiometer. Again, a Dremel was used to
cut the groove.
Above and to the right we see a beam
comparison picture of several lights. The light on the left
is a Luxeon Star running on a regulated 3V. On the right is
a standard little 2-AA faceted reflector incandescent with a
Krypton PR base bulb. The center is the DB mod. The DB mod was
originally running at 3V and had identical light otuput to the
beam picture on the left. However, after 17 minutes running
continuously the light began flashing - this is a sign that
the LM2621 board was overheating. The head of the light was
noticeably warm and the Luxeon Star burned the experimenter's
fingers when grasped. When the board in the DB mod had its
output reduced to 2.9V it easily survived a 20 minute runtime
test with no problems and the head of the light was only slightly
warm. Light output was noticeably reduced, but is still sufficient
for any task which needs to be completed in a low-light situation.
