In the fourth part of this series, we look at the development of metal halide sources and their effect on lamphead design...

The Spark.

To those of you who have to use and carry lighting equipment, the advantages of reducing the size of equipment are self-evident. As discussed in the last article, the introduction of the tungsten halogen lamp allowed big reductions in lamphead size - similar benefits were gained for metal halide discharge lighting with the introduction of the single-ended lamp. But an even greater step was the development of the metal halide lamp from the arc lamp.

Arc lamps have existed for nearly two hundred years and were still widely used in the film industry in the 1960's. Although air is normally a good insulator, it was discovered that an arc could be formed if two electrodes, for example two carbon rods connected to a supply of electricity, were touched together momentarily then drawn apart. Adjusting the gap between the electrodes could regulate the size and brightness of the arc, although constant adjustment was necessary to maintain the gap size as heat from the arc burned the carbon away.

The arc lamp produced a very bright light and, because of its small size, was a good substitute for sunlight - particularly able to produce hard shadows. Besides requiring the constant attention of an operator, there were other disadvantages: poisonous fumes; a high UV content and only limited control of the light colour. Sealing the electrodes into a glass envelope made it possible to control the mixture of gasses which minimised burning of the electrodes, eliminated fumes and give greater control of the colour and consistency of the light produced. This method was used to develop Xenon lamps commonly used for film projectors, but as these lamps operated at extremely high gas pressures they were far from ideal for location use. Metal Halide lamps operate at lower pressures than Xenon and are not at high pressure when cold, making them much safer to handle. They were first used with great success at the 1972 Olympics.

Sealing the electrodes into a capsule created two obstacles. Because it was no longer possible to touch the electrodes together, a source of high voltage - the igniter - was used to jump the gap. This high voltage, up to 65,000V, is only required for between a half and two seconds at most. When the lamp is initially 'struck', the high voltage breaks down some of the gas inside the lamp into positively and negatively charged components or 'ions'. These ions effectively make the gas conductive so high voltage is no longer needed to maintain the arc. Because the connections between the igniter and lamp require expensive high voltage cable, the igniter is generally inside the lamphead.

Secondly, the gap between the electrodes could not be adjusted to regulate the current. Once the igniter forms an arc, a virtual short circuit is formed which would simply blow the supply fuse. An additional component - the ballast - is needed. Originally, electrical chokes were used which consist of a coil of wire on a metal former. As the flow of energy to the lamp increases, the coil forms a magnetic field of increasing strength. This field opposes the flow of current. If current increases, the field increases to oppose it, restricting the current to a defined value. This type of ballast is referred to as a 'choke', 'wire-wound' or 'conventional' ballast.

In the 1980's electronic ballasts were introduced. These are far more complex than chokes but have important advantages: they are smaller and lighter; because they have a control system they can compensate for variations in factors such as varying mains voltage or lamp ageing. Most importantly they offer 'flicker-free' operation - particularly important for shooting on film.

Our normal mains supply is AC - alternating current changing from positive to negative fifty times each second in the form of a sine curve. If supplying say, a 1kW filament lamp, there is virtually no change in light output as the mains varies since the filament would take time to cool. With a 25W pygmy bulb there will be a measurable variation in output since it will cool as the mains voltage reduces, getting hotter and brighter as it increases again. Similarly, in a discharge source, the light is coming from a very small mass - a cloud of gas, which cools very quickly. With 50 cycles per second (ie 50Hz) mains, as in the UK there is no problem when filming at 25fps. The light source produces one hundred bright peaks every second - one for each positive and negative maximum in the supply. In 1/25th of a second, that gives four peaks - two complete peaks shine onto the film and two through the viewing system. It does not matter that the light varies during the exposure, because each exposure will be the same. However, if we filmed at a speed that was not related to the variation of the light, say 75fps, some frames would be exposed during peaks and others during troughs causing some bright and some dark frames.

Diagram courtesy of B & S Electronische

One notable disadvantage of electronic ballasts is increased noise from both head and ballast. To reduce this, most ballasts allow the selection of various low-noise modes. It is vital to understand that in these modes, the light is no longer flicker-free.

Original metal halide lamps were linear in form with an electrical contact at each end. The lamps were comparatively long to protect the seals from heat and, because of the high striking voltage, there had to be substantial clearance between the ends of the lamp and the sides of the lamphead. With the single-ended design, for example MSR and HMI-SE lamps, a shortened version of the linear lamp is mounted on a two-pin lamp base similar to those used on many tungsten halogen lamps. The lamp capsule is enclosed in an outer glass envelope, which improves temperature stability.

The technological demands of this design are much higher. As the two pins of the base are close together, a high degree of insulation is needed to prevent arcing between the pins. Similarly within the capsule, arcing may occur between the short and long electrodes, both outside and through the wall of the inner capsule. As the seals where the conductors pass through the glass are closer to the heat source, these too must be more carefully designed and manufactured.

The critical time where the lamp is most stressed is during hot re-striking. As the gasses are heated up and begin to emit light, the pressure inside the lamp increases until it reaches normal working temperature. The problem arises when the lamp is extinguished. The ions recombine quickly which removes the easy path for electricity to flow. However, the gas in the capsule is still very hot and the pressure is still high. During this 'window' it is very difficult to re-strike the lamp and a much higher striking voltage is required. But this higher voltage will try to find the easiest path to complete its circuit. Any weakness in insulation either in the lamphead or the lamp itself may allow electrical energy to leak away. In extreme cases, an arc may even be formed outside the capsule. Some judgement has to be exercised by the lamphead designer to allow a high enough igniter voltage for hot re-striking without causing damage to lamp or head.

Sometimes, there is simply not enough voltage to form an arc and the lamp must be left to cool. Repeated attempts to strike a reluctant lamp can cause the glass envelope to become coated with tungsten from the electrodes. This can make the lamp even harder to strike because the current will track across the metal deposit without forming an arc. The only hope for a lamp in this condition is to let it cool down thoroughly before trying to re-strike it. Once lit it should be left running to allow it to heat up fully so the metal deposit can be vaporised.

As a lamp ages, the arc gap increases and it can become progressively harder to strike. In fact one mode of failure of a discharge lamp is that it becomes so hard to strike that it is unusable. Often, by this time other effects of age such as incorrect light colour will also be evident.

Lamp manufacturers all have their chosen additives to improve hot re-striking characteristics, although there can be variations in performance between, not only different manufacturers but different batches of lamps. It is important to understand there is a big difference between a hot re-strike lamp which is for whatever reason, reluctant, and a non-hot re-strike lamp (eg an HQI lamp) which is not designed to and would be damaged by attempts to strike it before it had cooled down.

To stand the stresses of hot re-striking, the HT cables must be in good condition with no damage to the insulation. Black spots can be an indication that an arc has been formed between cable and housing. Similarly, the ceramic of the lampholder and lamp base must also be clean and in good condition. In some cases, fire clay may be used to make emergency repairs to ceramic parts.

From the lamphead manufacturer's perspective, single ended lamps allow the high voltage to be confined within the lamp base, allowing for a smaller lamphead. The construction of lampheads where the lamp passes through the reflector is simplified - namely, narrow beam units using parabolic reflectors, for example the Arrisuns and Pocket Pars. In the next article, we will look at some of the available types of lamphead and the different jobs they can do.

By Andy Barnett, ARRI GB.

next >>