Hi all,
I bought this xenon flash tube off eBay. Here is the Russian datasheet: http://rumata.net/books/31.pdf
I just can't believe something that small can take such extreme amount of energy. 2kJ can literally vaporise a piece of wire in a cloud of plasma and a bright flash, and the loud bang leaves your ears ringing (Done that earlier with a bank of capacitors).

I don't want to damage this lamp as it wasn't exactly cheap. Are there any precautions I should take before firing it up at full energy? Any specific operating conditions, mounting position etc?
Please see the attached schematic, do you think it is OK to use with this lamp?

PS. I cleaned the lamp with alcohol, I hope this didn't harm it in any way...

Thanks for your replies.
-Jan

Operating position is not restricted (end of chapter II).

Pulse transformer: Primary = 25turns, Secondary 2500turns, wound on the 21mm diameter spool, no magnetic core.
Ignition capacitor 1uF/1000V

The lamp is apparently rated for three different regimes (table 2):
Regime A:
Nominal energy: 2kJ
Flash capacitor: 16mF
Charging voltage of the flash capacitor: 500V
Maximum ignition supply (not sure, what that mean, probably the max voltage on the ignition pulse transformer primary): 250V
Light output: 80klm.s
Pulse duration: 4ms
Minimum repetition time: 1.5s
Lifetime: 5000 pulses
Minimum voltage for spontaneous ignition (without the pulse on the auxiliary electrode): 2200V

Regime B:
Nominal energy: 400J
Flash capacitor: 8mF
Charging voltage of the flash capacitor: 320V
Maximum ignition supply (not sure, what that mean, probably the max voltage on the ignition pulse transformer primary): 250V
Light output: 14klm.s
Pulse duration: 2ms
Minimum repetition time: 1.33s
Lifetime: 100000 pulses
Minimum voltage for spontaneous ignition (without the pulse on the auxiliary electrode): 2200V

Regime V (V is the 3'rd letter in the Azbuka):
Nominal energy: 400J
Flash capacitor: 800uF
Charging voltage of the flash capacitor: 1000V
Maximum ignition supply (not sure, what that mean, probably the max voltage on the ignition pulse transformer primary): 250V
Light output: 18klm.s
Pulse duration: 0.25ms
Minimum repetition time: 1s
Lifetime: 30000 pulses
Minimum voltage for spontaneous ignition (without the pulse on the auxiliary electrode): 2200V


So the schematic should be OK. I would maybe adopt the ignition circuit, as charging the ignition capacitor to 250V would be rather inconvenient:
Ignition capacitor 470nF
Ignition coil primary: 33turns (instead of the 25 turns)
Charging voltage for the ignition capacitor: 320V (= mains peak voltage ;-) )

And for experiments I would not use the full energy (regime A), as the lifetime is quite limited there, unless the high flash energy is really needed...

Medved,
do you think the pulse transformer must output a specific pulse rise time/voltage, or can be arbitrary as long as it's sufficient to ignite the lamp? My original idea was to use an automotive ignition coil, or I could use a TV flyback transformer for smaller size/weight.

I'd guess arbitrary. Perhaps you can use  fluorescent ballast (of any power) and superimposed HPS ignitor for the starting wire - just power it up when you want the flash

I don't know, what the "maximum ignition supply" mean.
There may be upper limit to not break down the glass between the auxiliary electrode and the arc.
And the lower limit is the successful ignition of the flash.

I would start with flash tube ignition transformer from another tube and then with the TV transformer.

The HPS ignitor would be insufficient in the peak voltage for the external electrode.
And ignite the lamp via the internal electrodes may work (connect the ignitor between the flash capacitor and the lamp, the "Neutral" from the ignitor is connected via the trigger switch), but there is very high pulse current during the flash, what may mechanically (by the large electromagnetic forces in the pulse transformer) damage the ignitor.

I have quite high expectation from the HPS ignitor. It strikes a HPS lamp which i hold in my hand by the glass, touching the tip of the base to the isolation of the ignitor's wire

At the very last sentence it says that fingerprints should be wiped off with alcohol. So you did right.

I was just testing the lamp. I connected a TV flyback transformer to the starting wire, which alone makes some 1.5cm sparks. I put a 5 uF film cap across the main terminals, charged it to 300 Vdc, fired the TV transformer and nothing happened, just very faint purple sparks inside the tube. I tried swapping the polarity of the flyback primary but this didn't help, I also tried a different transformer - nothing... The 5 uF cap remained fully charged.

Is there a simple test I can do to make sure the lamp is OK?

Edit: I tried running it on a NST, it made an arc so I think it is ok... The lamp fired at 3800 volts peak from the transformer.

In any case observe the polarity.
For he ignition: It may be, the external electrode need high dV/dt the TV flyback is not able to support.

The pulse transformer rise time is important, as it has to couple through the capacitance formed by the glass with enough voltage to ionise the gas fill in the lamp, which then initiates the main discharge. You need a very fast rise pulse to do so, and the max voltage of the pulse is recommended to not exceed the breakdown voltage of the glass.

The storage capacitor used with the tube is very important, as it must be specified for the high current discharge the tube requires. A standard electrolytic or film capacitor will not last long, some will explode within 10 shots, or will go open circuit. The 800uF photoflash capacitor is built with thick foils, and is formed to a high voltage ( much higher than the rating says) with multiple leadout tabs to the terminals, along with a very conductive electrolyte composition. They still only have a limited life. There are extended foil capacitors for photoflash applications with high values, but they are very big units. Remember the 2kJ flash requires a capacitor of 16 000 uF, rated for 500V minimum, discharged in 4ms into the lamp, at an average current of 2000A. Thick wire is needed in the current paths to the tube, and short wires as well.

Best source of a prewound transformer is a disposable camera, which gives both the pulse transformer, a charger that operates off a single AA cell and the capacitor. You strip the unit, and wire the tube in place of the original, after discharging the capacitor with a 1k 5W resistor across the terminals ( safer than a screwdriver, and less likely to shock you, provided you insulate the resistor body with some PVC sleeve where you are holding it, and less noisy). Just note which end of the tube is nearest the discharge ring, and wire the big tube the same. Will not give the 2kJ discharge but will be bright enough.

Ok I will try to construct another transformer, maybe a small Tesla coil will work. As for the caps, I will be using 10 of these, 2 in series and 5 strings in parallel.

You may reduce the peak current by using a series coil with the flash tube. That will lessen mainly the mechanical stresses (coming from electromagnetic forces) on the internal assembly of used capacitors.
The coil should be "air core" and well mechanically build - it would create strong field with high forces mainly to itself.

Thanks guys, the lamp just made a first flash  I had to charge the cap to 330 volts (though the minimum required ignition voltage is 250 volts) and I pulsed the cap to the transformer's primary with a SCR. However then I found out it also works without the SCR, just with the mechanical switch. So at least I know the lamp is working 

In camera flash it is common to charge a second small capacitor parallel to the 1st one through large resistor, and pulse the transfromer from the 2nd one. This limits the energy that the transformer gets

Sorry I haven't put it too clear, of course I use a different cap for the ignitor (1.5 uF) which is charged separately from the main one.