<mrdata>
nmz787, a 40 watt bulb at ... 10 kilovolts and 4 milliamps, say?
<mrdata>
you will want to control the current
<mrdata>
is it AC or DC?
<mrdata>
power is V^2 / R so R = V^2 / P .. so that bulb has a resistance of 360 ohm; but at 10 kV that will allow 27.8 amps to flow
<mrdata>
so you definitely need to control current
<mrdata>
maybe by use of a capacitor which can only deliver a certain charge
<mrdata>
then you end up with an RC circuit; but the time constant is going to be very short
<mrdata>
to the bulb's thermal resistance matters
<mrdata>
*so the
<mrdata>
you can blow it up if it heats too quickly
<mrdata>
otherwise go ahead an operate it intermittently
<mrdata>
so a 20 nF cap charged to 10 kV then discharged through that bulb will supply 1 joule per pulse with tau = 7.2 µs; pulse it 40 times a second for 40 watts
<mrdata>
what could go wrong? (a) risk of internal arcing (b) heating too fast
mofh has quit [Remote host closed the connection]
_whitelogger has joined #homecmos
mofh has joined #homecmos
X-Scale has quit [Ping timeout: 246 seconds]
X-Scale has joined #homecmos
jimn has quit [Ping timeout: 268 seconds]
jimn has joined #homecmos
BitEvil has joined #homecmos
SpeedEvil is now known as Guest54818
Guest54818 has quit [Ping timeout: 255 seconds]
<nmz787>
mrdata: thanks, this is an arc lamp power supply so the kilovolts are just for striking a plasma before the voltage drops to 39 VAC
<nmz787>
apparently some power supply topologies base things on impedance, since the resistance of no-plasma is much much higher, it allows the supply to charge up to kV, but once the plasma is active the impedance gets way way lower and the supply cannot sustain the kV anymore
<mrdata>
nmz787, fire it up with a cheap bulb and let me know the dmage
<nmz787>
but that was just one type of supply that I read a journal article about
<nmz787>
others use a parallel topology, where the 39VAC is spliced with the kV supply via a big diode
<nmz787>
but it does seem like either way the kV shouldn't last *too* long
<mrdata>
more than a few microseconds could be too long
<nmz787>
mmm, yeah I was also doing some power calcs yesterday based off assumed resistance
<nmz787>
the amps def get crazy at 30kV
<mrdata>
gtg; tty later...
BitEvil is now known as SpeedEvil
john_cephalopoda has quit [Ping timeout: 258 seconds]