small medium large
extra large extra
extra-large hey everyone I started to
get into xenon strobe lights and so I
wanted a way to measure the voltage
current and light output of a strobe as
it flashes I took apart a speed of Tron
4803 head unit and discovered that
there's a pulse transformer inside the
head so there's two large wires which
supply the capacitor voltage the whole
bank voltage out to the strobe light and
then when the unit wants to cause a
flash to happen it sends a pulse out to
the transformer which kicks up probably
a few thousand volts to initiate the
strobe the easiest measurement is just
to measure the voltage across the strobe
light as it fires but I already ran into
a problem the speeder Tron is wired so
that the neutral of the mains supply is
actually connected to the middle of the
capacitor bank and so there's plus 450
and - 450 around the neutral and the
case is also grounded I tried lifting
the ground just by disconnecting it and
still had a fault when I connected the
negative side of the capacitor bank the
negative 450 to my Scopes ground I
proceeded to find a woefully inadequate
isolation transformer on my shelf and
connected it up between the speedo Tron
and the line with just some alligator
clips I knew that this was not going to
be sufficient to run the unit for any
length of time but the way the speedo
Tron works it charges up and draws a lot
of power for a relatively short amount
of time and then draws very low power
while it's idling so I thought if I was
careful enough and didn't fire it too
often I'd be okay later on the
transformer actually started to smoke a
little bit so it was definitely getting
hot but did the job with the isolation
transformer in place I was able to
connect my scopes ground to the negative
side of the capacitor bank so
essentially the entire speedo Tron unit
is now pushed about
ground so it would be a very bad idea to
touch the case well it's in this
condition I have a wooden work bench and
was felt well aware not to touch the
metal case while this was all going on
I used a high voltage probe to detect
the voltage across the strobe light and
fired it with it just a simple trigger
on the oscilloscope and got this chart
here I was very surprised at how slow
the voltage changes so you know you can
see the charge cycles starting kind of
near the ends there and it actually
takes over 50 milliseconds to get down
to a steady state voltage the 75
millivolts on the Left column is
actually the steady state voltage where
that small cursor is so it levels off at
about 75 volts and the capacitor bank
aim is charged up to about a thousand
volts to start before the flash is
initiated next I wanted to measure the
current in addition to voltage so one
way to do this is with a shunt so I have
a Milian meter and some 14 gauge solid
wire and I slid the alligator clips
around on the wire until I had a section
that was about 1 million resistance the
idea being that for every amp that flows
through this section of wire there will
be a 1 millivolt drop so I soldered the
wire in between the negative of the
flash lamp and a negative supply and the
speed of Tron and then slid the
alligator clips around again and marked
it with a sharpie so that I knew where
the the section the appropriate section
of wire it was as you can see the
voltage and current are fairly linearly
related during the flash
I also produce another graph showing the
power tracy is the is one times two
which is voltage times current and D is
the integral of C so D is actually the
total power that the unit has put out
over the course of the flash and as you
can see that levels off and the cursor
shows that there's about five point one
seven units there and the units are in
fact kilojoules so this this agrees with
reality pretty well it's a forty eight
hundred watt second unit which is a
Joule and the measurement is coming in
pretty close five point one seven
kilojoules here's a close up of the
current drawn by the strobe during a
flash as you can see there's a peak and
the peak is around three thousand amps
and then the graph levels off right
around a thousand amps and continues on
sloping very slowly downward for for
quite a long time the x-axis in this
case is 50 microseconds per division
finally I wanted to see how much light
was actually coming out of the strobe so
I took an LED and put a 100 ohm resistor
across the leads and then connected that
directly to the scope the idea is that
when light hits the PN Junction in there
it functions kind of like a tiny solar
cell and the 100 ohm resistor keeps the
junction from saturating so initially I
tried this without any resistor on there
and the amount of light going into that
LED saturates it's so fully that the
voltage goes up to you know one point
six or seven volts or whatever the the
junction is set up for and it basically
just hangs there for a long time because
the junction is so incredibly saturated
so with a stiff 100 ohm resistor across
it the junction won't saturate and we'll
get a very very accurate quick quick
time measurement of the light going into
the junction so as you can see though
the light is not quite proportional to
the current in this graph B is the
current and a is the light output from
the LED and you can see that the current
is decaying you know kind of linearly
over this scale but the light is
actually decaying much more quickly the
time here is
milliseconds per division so as you can
see there's about five divisions are
about ten milliseconds for the you know
the approximate flash duration I'm
actually surprised at how long of the
tail is on this thing I was expecting
the flash to be very very quick quick
down to zero and then a lot of you know
zero but this actually does taper down
quite a bit so it's hard to say how long
the flash duration actually is because
there's so much tapering okay well I
plan to continue these experiments with
xenon flash tubes hopefully we'll get to
do some Ruby laser stuff and high-power
aerial photography at night perhaps
something setting things on fire
you never know okay see you next time
bye
s
Posts
No comments:
Post a Comment