MB Lamp Development

Once the first Philora HP300 lamp was beginning to sell in quantity, two principal features of its design were found to be unfavourable and required modifying. First, the high arc voltage necessitated a special transformer ballast which was heavy, expensive and inefficient, and it was a severe obstacle to the marketing of the lamp. Secondly, the arc tube was so small that it blackened rather quickly and the useful lamp life was short. The pressure was on to create a new lamp which solved these two issues.

In the same way that all lampmakers of any significance had started making MA lamps within a few years of its invention by GEC, the same applied after Philips created the quartz MB lamp. Philips, British Siemens and British Thomson-Houston in particular were three companies that devoted particular attention to solving the problems.

To get around the blackening issue it was decided to increase the arc tube diameter and make it longer, thus increasing its internal surface area. To meet the requirements of a lamp which would work on an open circuit voltage of 220-240V, the arc voltage necessarily had to drop to around 120V. This happened naturally as a result of the increased arc tube dimensions.

There then arose the problem of how to strike the discharge on such a low open circuit voltage. The practical solution was to return to a third auxiliary electrode located within the arc tube. With skill it was found that another graded seal to a third tungsten wire could be added to the side of the arc tube, rather than running all of the sealing wires through the ends of the arc tube as had been standard practice until this time. Such a seal is displayed in Figure 39 to the right, and an example of a lamp made with this kind of seal can be seen by clicking here.

A tremendous contribution to the future of the lamp industry was made in the early 1930's by Dr. Dennis Gabor OBE, the famed Hungarian inventor who is perhaps most famous as the father of holography, for which he won the Nobel Prize in 1971.

In 1931 he took up employment in the Berlin labs of Siemens & Halske (the predecessor of the present-day German Osram company, no connection with Osram-GEC of England). While there he was charged with developing a high pressure cadmium lamp, which it was felt could be a viable alternative to mercury. He realised that quartz had to be used as the arc tube material because cadmium tends to be quite corrosive towards other glasses, and devoted much of his attention to sealing metallic conductors through quartz. In 1931 he was awarded a German patent for a method based on very thin, flat conductors of molybdenum. These molybdenum strips were so thin that despite having a coefficient of expansion some 8 times greater than quartz, the foil was thin enough that the small amount it expanded by was not sufficient to rupture the seal. Osram employed these seals in the manufacture of UV lamps for laboratory and medical applications and for many years they were unknown to other lighting firms who were struggling on a similar topic. German Osram had little activity in mercury discharges for lighting applications at that time and surprisingly failed to realise the significance of Gabor's seal.

With the rise of Hitler in 1933, Gabor left Germany and obtained employment as an inventor in the research laboratories of the British Thomson-Houston company in Rugby, England. He was a great asset to the discharge lamp development team, where he introduced the company to his molybdenum foil seals and went on to develop many other ideas for BTH - one of the most famous being a ballastless discharge lamp. Shortly after Philips put its first HP300 lamp on the market, BTH showed a very similar product but which was made with Gabor's molybdenum foil seals. The reliability of the seal was improved by sharpening the sides of the foil to knife-edges by an electropolishing method, giving it an almost ragged appearance along the edges. This eliminated the seal leaks that otherwise occur if the edge of the foil is perfectly straight or not feathered, as a tiny airline would often be formed along these edges. With the moly foil seal it became easy to accommodate a third wire for the auxiliary striking electrode, and Gabor's arc tube, first marketed through BTH Mazda, set the standard for all quartz arc tubes right up to the early 1950's. An arc tube made with molybdenum foil seals is shown below in Figure 41.

A slight improvement on this seal was perfected by British Siemens, who recognised that a serious bottleneck in the lampmaking process was the handling of these thin foils and the welding of an electrode to one side, and a molybdenum lead wire to the other side. The welding process may justifiably be described as completely out of control in the early days; getting tungsten to stick to molybdenum when spot-welded is quite an art! The technique employed by Siemens was to take a molybdenum wire and press it flat in the centre. That flattened region was then rolled to a very fine strip, and its edges sharpened as usual. The component thus formed was a single piece consisting of the foil with a wire on either side, to which an electrode could be attached and external connections made. The Siemens seal is shown in Figure 42. It was only short-lived though, because before long the welding problems were overcome by using intermediate so-called 'referee' metals which helped the parts stick together more strongly, as well as through improved control of the welding current. Various methods were used and indeed still are used today. The foil is often platinised on the side to be welded, or a small amount of platinum enamel may be applied before welding, or as an alternative a thin wire or strip of tantalum may be sandwiched between the parts to be welded. The waveform of the current supplied to the welding electrodes is also of great importance in making a good quality joint.

The most recent update to the method of forming a mercury arc tube was to mechanise the process of sealing the electrodes into the arc tube through a remarkably simple process innovation. Instead of using atmospheric pressure to collapse the quartz down over the foil, a technique was developed which allowed the lampmaker to pinch the quartz flat around the foils on a fully automatic machine. Oxidation of the molybdenum and electrodes was prevented by flushing argon through the tube to be sealed continuously during the heating process.

To ensure that the correct shape was formed around the electrodes at the ends of the arc tube, the tube diameter was first reduced at each end. This was usually carried out by one of three methods. The first was to manually stretch the end regions down to a smaller diameter while heating the tube, this method requiring skilled glassblowers and being rather time-consuming (Figure 43a). Another method was to attach short lengths of narrow bore tubing to each end of the arc tube (Figure 43b). Alternatively the tube was simply rilled-in at the region around the electrodes to form the correct shape here, and below this point the tube remained at its full diameter (Figure 43c).

The electrode mount assemblies were then located inside the vertically mounted arc tube on a special machine, a flow of argon was flushed into the assembly through the exhaust tube, and oxy-hydrogen fires rapidly brought the end of the tube up to its softening temperature over a period of about ten seconds. Oscillating fires were normally employed to spread the heat out well over the entire surface of the area to be pinched. Then in a split second heating would cease and a pair of brass hammers would fly in from either side and pinch the quartz flat around the foils. The process was then repeated at the opposite end of the arc tube.

The process of pre-shaping the arc tube ends prior to pinch sealing was rather tedious and something that lampmakers were keen to abandon in efforts to further speed the production. In 1960 it was successfully abolished by incorporating a V-shape region into the top of the hammers used to make the quartz pinch-seal. By heating a greater length of quartz at the end of the tube it could be partially pushed inwards as part of the pinching action, thereby forming a very neat conical end chamber around the electrodes. This was the last process improvement made in the fabrication of MB style arc tubes, and since then the production method has remained largely unchanged. To speed up the process alternative methods of heating have been introduced though. The oxy-hydrogen fires are quite slow to heat up the quartz, and recent methods involve ionising the argon gas which is flushed over the electrodes to form a plasma torch, rather similar to the flame found in an atomic absorption spectrometer. This method is much faster and cheaper to operate than traditional fires, but a large capital expenditure is required to construct the machinery.