Xenon flash tubes in series

[Xenom]

Hi.. i have one of my crazy ideas. My idea is to link conventional 13 mm arc cheap xenon tube scraped from disposable cameras in series, to form a big massive flash lamp.
I have 8 flash tubes with relative electronics and capacitors.

It is a good or a bad idea to link these tubes in series and power them with a cockcroft-walton voltage multiplier? i'll trigger the tubes with one HV pulse generated from one HV pulse transformer.

Each tube support 300V, 10 J max i think, so the final configuration i think could be about 2500 V with 9 330 V 10 uF capacitors in series, charged via voltage multiplier.

Any suggestion?

thx

PS: i know that this experiment is very dangerous, these energies and voltages are very dangerous.

 

[Things]

Only issue I can think of, is the outer 2 flash tubes have to be able to carry the current for the middle one as well without exploding or completely destroying their electrodes.

On the other hand, parallel is almost certain to not work, as it'll just arc down one tube and dump all the energy meant for the other 2 into that 1.

 

[BShanahan14rulz]

Only issue I can think of, is the outer 2 flash tubes have to be able to carry the current for the middle one as well

Is this a special property of discharge tubes where they "eat" current?

 

[Things]

They do, they turn it into heat and light

 

[Polonium210]

What is the intended purpose for this? Are you trying to pump a YAG rod?

 

[Cyparagon]

They do, they turn it into heat and light

Current is the flow of electrons - what goes in the lamp must come out. Electrons don't disappear when they've gone through the first lamp. The current across all the lamps will be the same.

 

[Xenom]

What is the intended purpose for this? Are you trying to pump a YAG rod?

No.. (maybe one day XD ). I'm just curious to see the result... It is an experiment.

if i connect the voltage multiplier directly to main capacitors and the tube, when i trigger the tubes the impedence drops, it is near a short circuit; Could the voltage multiplier damage from that?

 

[HIMNL9]

current for tubes in serie is the same, but you need to multiply the voltage for the number of tubes (and finding a 2400V capacitor can be a bit difficult )

For the other side, i occasionally build "emergency flashers" for friends (resin-sealed units, for use for emergency in scuba or nautical conditions, so made with "disposable cameras components" ), and i never had problems using 2 or also 3 lamps in parallel, they flash all together, dividing the current (each lamp, ofcourse, flash less bright, but my purpose, in the emergency units, is that they have to be visible from each directions)

 

[Xenom]

i have problem with parallel configuration: i tried with 8 tubes in parallel, one small capacitor, one hv trigger pulse.

Each time i trig, lamps flash randomly. I was not able to flash all lamps together and simultaneusly...

 

[Xenom]

I have just tried series configuration as i described before, but with less power and energy.
I linked 4 tubes in series and linked them to a voltage multiplier (4 stages), so i have about 1200VDC on the Vout.
I used a pulse capacitor and trigger scraped for a disposable camera for trigger.

It seems to be working, now i'm using a very small capacitor (15 nF) for the main flash capacitor.

I'm goning to increase the capacity

PS: obtain a 2500V capacitor is not difficult, it's sufficient to link in series 8 or 9 330V photoflash capacitors XD the problem is that 120 uF at 2500V is too high energy (above 400J). Extremely dangerous and the tubes will explode. i have to buy or find 1 to 10 uF 400V capacitors

 

[HIMNL9]

8 tubes are too much, for parallel work, probably ..... my units usually uses 2 or 3 tubes in parallel, the max i've tried working, was a big unit with 4 tubes (90% of the times works, randomly one don't flash)

 

[Cyparagon]

Could the voltage multiplier damage from that?

Possibly. I would just shut it off before you fire the lamps.

 

[HIMNL9]

^ or use a resistor fro charge the caps.

BTW, about the 8 or 9 capacitors in serie ..... yes, it increase the total voltage (better use identical ones, each one with a balance resistor), but at the same time, decrease the capacitance (capacitors in serie are the same as resistors in paralel, for the value ..... 10 x 120uF 330V ends in 12uF 3300V ), so the total energy don't change too much

 

[Xenom]

Possibly. I would just shut it off before you fire the lamps.

how? the easiest way?

^ or use a resistor fro charge the caps.

this can be the easiest way i think. right?

BTW, about the 8 or 9 capacitors in serie ..... yes, it increase the total voltage (better use identical ones, each one with a balance resistor), but at the same time, decrease the capacitance (capacitors in serie are the same as resistors in paralel, for the value ..... 10 x 120uF 330V ends in 12uF 3300V ), so the total energy don't change too much

loool i didnt remember that omg... yes it's true... capacitance decrease with the series configuration. lool.
So.. i can use without problem the photoflash capacitors, right?
120 uF 330 each, so 9 of these and i have about 10 uF, very good!

How about the balance resistor? the value? is it necessary?

 

[HIMNL9]

With serial configuration for high energy banks, balance resistors are not indispensable, but always suggested ..... also, they discharge the capacitors all together when not in use, preventing that one remain partially charged and "blow out" for overcharge the next time you use it.

for 330V capacitors, i normally use 150 Kohm resistors (one in parallel for each capacitor), and always try to use identical capacitors (still for avoid "KABOOM" effects ).

Unfortunately, with components from disposable cameras, it's not too easy to be sure about them ..... the better thing is trying to get a lot of them from a photo-lab, and choose the ones from the same brand, model and, possibly, year of production.

Another trick for "safeguard" a bit the elevator, can be using a small coil in serie to the flash lamps bank ..... nothing special, it only need to take care about the "spike" initial, and need to hold the current involved ..... as example, if you get some coil wire (maybe from an old mangled transformer ?), enamel insulated single-core wire of 1mm or more, you can wire 10 or 15 turns of it on a 20mm support, keeping the loops very close, and usually it's enough (don't forgot the diode in antiparallel ..... for 2500V, 3 or better 4 x 1N4148 in serie are enough).

If i get my webcam work again, i may make some pics for show that thing better.

 

[Xenom]

With serial configuration for high energy banks, balance resistors are not indispensable, but always suggested ..... also, they discharge the capacitors all together when not in use, preventing that one remain partially charged and "blow out" for overcharge the next time you use it.

for 330V capacitors, i normally use 150 Kohm resistors (one in parallel for each capacitor), and always try to use identical capacitors (still for avoid "KABOOM" effects ).

Unfortunately, with components from disposable cameras, it's not too easy to be sure about them ..... the better thing is trying to get a lot of them from a photo-lab, and choose the ones from the same brand, model and, possibly, year of production.

Another trick for "safeguard" a bit the elevator, can be using a small coil in serie to the flash lamps bank ..... nothing special, it only need to take care about the "spike" initial, and need to hold the current involved ..... as example, if you get some coil wire (maybe from an old mangled transformer ?), enamel insulated single-core wire of 1mm or more, you can wire 10 or 15 turns of it on a 20mm support, keeping the loops very close, and usually it's enough (don't forgot the diode in antiparallel ..... for 2500V, 3 or better 4 x 1N4148 in serie are enough).

If i get my webcam work again, i may make some pics for show that thing better.

no problem for the capacitors and balancing resistors: all my capacitor are identical (i have about 15 identical disposable cameras). ok for balancing resistor, i have a lot of 150k.

i think i understood for the coil; the core must be ferrite or air works well?
BTW resistor in serie works anyway?

1N4007 diodes can carry the spike current well? i think the surge current can be 20A max at 2500V (omg, i have to watch out from that)