I bought some of these 35watt "xenon" (really metal halide) lamps for experimenting. Does anyone know the running parameters (arc volrage/current, ignition voltage) and do they run on DC/AC/HF-AC? I had the impression that they run on DC, can I then just use a series choke from 230VAC + bridge rectifier + conventional HID ignitor?

Thanks for help

This page should be helpful. They use AC. They are apparently very difficult to run.

Thanks! So the only difficulty seems the starting pulse, but one of my other hobbies is high voltage and means of generating it, so this shouldn't be a big problem for me.

From the description with the arc voltage of around 90 volts and low frequency AC waveform, it seems they should be able to run with just a simple series choke ballast from 230V AC. I'll try with a 40watt fluorescent ballast which I think should be OK for testing.

It is not only the starting pulse. Real constant power ballast is a must - on constant current (= series choke ballast) the lamp might become thermally unstable (automotive HID ballasts use constant power low frequency square-wave AC).
What make this lamp not as practical is the xenon fill - the ignition voltage is really high even at cold and the overall efficacy quite poor. The only reason for it's use is automotive requirement of instant full output - only xenon is capable of this at low temperatures and eve this with the expense of huge required power input (up to 90W) upon ignition.

Why the xenon in the automotive MH lamps cause them to become less efficient then regualr MH lamps with argon if the same gas (Xenon) are used in the Philips CPO-TW Cerium/Sodium CMH and does the opposite.

I think I read somewhere the 35watt lamp produces around 3200 lumens, which is around 90lm/W - IMO not bad for such a small lamp.

Anyway I have no other use for these lamps and I bought them cheaply so I'll experiment... maybe they'll run just fine on a fluorescent ballast...

@bluelights:
90lm/W have CMH for general lighting service, with very low buffer gas pressure. The high Xe fill pressure of automotive MH and the limitation to use quartz reduce their efficacy about to one half (below).

Why the xenon in the automotive MH lamps cause them to become less efficient then regualr MH lamps with argon if the same gas (Xenon) are used in the Philips CPO-TW Cerium/Sodium CMH and does the opposite.

The difference lies mainly in far different requirements and their priorities, yielding mainly different xenon fill pressure:
General lighting CMH has the top priority to deliver the highest efficacy and good color rendering, while longer warmup delay and larger luminous surface area are of no issue (in fact the diffusion is a plus, as it even out color gradients), so the ceramic arctube (allowing higher operating temperatures) and very low buffer gas pressure might be used, so it does not interfere with the arc energy budget. The xenon is used, as it allow very low concentration, lower then Ar and still is efficient in providing medium for initial heat-up and protecting electrodes. In fact it focus all power dissipation in the initial state (before mercury evaporate) to electrodes, where the mercury usually reside when cold.

Automotive lamps have in contrast requirement for operation in highly focused beams (what allow to form the required distribution pattern), what require very small "point of light source", what ask for the quartz as the only clear material available for the arctube. It's low (compare to ceramic) allowed operating temperature reduce the range of useful luminous fill materials.
On top of this is the requirement for instant full luminous flux, what is achievable only with xenon at rather high pressure. But such high pressure causes it's density in the arc being quite high, so it absorb significant amount of energy from accelerated electrons (what normally excite luminous atoms like Na, Hg, Sc,...) and emit the energy in full xenon spectrum, so significant radiation of the lamp is then in xenon-IR.
At the same time, as Xenon emit at lot of wavelengths across the visible spectrum, it is capable to absorb the light from the same spectrum, affecting even the radiation of other luminous elements, reducing the useful light output even more.
All that is "paid" (compare to general lighting MH) only in order to be street legal...

I CAN tell you it doesn't light on a stock 50W ballast. I tested the operating voltage at like 500V if my memory serves me.

I, too, have seen these rated for (in some cases)3500lm. The HID flashlights for example. Are they all liars?

I CAN tell you it doesn't light on a stock 50W ballast. I tested the operating voltage at like 500V if my memory serves me.

The problem is in the too low ignition voltage of "regular" ballast

I, too, have seen these rated for (in some cases)3500lm. The HID flashlights for example. Are they all liars?

Yes and no.
Question is, if it was really the same bulb as the automotive one. Welch-Allyn (i hope, i spell it correctly) manufacture special bulbs for these torches, but these are not the same, e.g. do not offer instant full output (as per automotive definition).
Then even when automotive lamp used, it might be overpowered, so it then deliver 3500lm, on flashlights the life in 100's of hours is frequently not as big issue.

What is definitely a lie, if somebody state something like "35W power input, 3500lm in the beam". For that he would need >150lm/W lamp for 10deg beam, for narrower beam even way more.

And including spill light into the torch lumen specification (very frequent) is practically a lie as well (or really at least it is not truth)

My car has Xenons fitted as standard and i can assuse you they are not instant full brightness they take about 6 seconds to reach 100% however they do have hot re-strike