EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: BUAROD on February 06, 2022, 09:52:46 am
-
A decommissioned xenon lamp has been lying in my basement for a number of years, waiting to be put into operation again.
It is a 15kW Ushio UXW-15KD short arc lamp (picture 1) of the type formerly used in IMAX cinemas. With a burning voltage
of 37.5 volts, a nominal current of 400 amps and approx. 45 kV ignition voltage. In continuous operation it requires water
cooling!
In order to generate the ignition voltage, a special ignition device is required that is connected in series between the power
supply and the lamp (picture 2). It generates the high voltage before the lamp is ignited and must be designed for an operating
current of 400 amps (!) if operation at max. power is intended.
For example, the AS-40045 ignition device from IREM (Italy) is suitable. Despite an intensive search in recent months, I have
not found such a device in used condition on the known platforms. In contrast, the AS16040 ( (picture 3,4) is offered quite
frequently.
But this unit can only carry 160A continuous current and the ignition voltage is 40kV (picture 5). This voltage will not be
sufficient for this type of xenon lamp, specially at the end of its service life. So I was not very surprised that he unit was
not able to ignite my old lamp.
After analyzing the circuit and searching the network, the functional principle became clear to me (picture 1): a mains
transformer TA converts the 230 volts into approx. 15kV. This voltage reaches the primary side of the high-voltage transformer
via a 4-stage spark gap. When the half-wave of the 15kV reaches the breakdown voltage, an oscillating circuit is created
(3xC + HT) with extremely high current peaks, which generates high voltage pulses on the secondary side of transformer HT.
Despite special materials on the spark gap surfaces, burn-off and contamination had occurred over the (apparently) long
service life. After dismantling and cleaning the spark gaps (picture 6,7), the idea was to consciously set the gaps a little
larger than found. The intention was to increase the breakdown voltage and thus also to increase the voltage and energy
in the resonant circuit.
This not only successfully ignited the lamp (still without connection of a DC power supply), but also destroyed several
LED indication lamps in my control panel! I had connected the lamp with approx. 1m long cables and they obviously worked
very well as antennas! The LED-lamps without any EMC capacitors in the small E10-socket could not withstand the
radiated energy. I then realized that I must have continued Marconi's experiments...