US2650322A - High-pressure metal vapor electric discharge lamp - Google Patents

Description

Aug- 5,1953 v. J. FRANCIS ETAL 2,650,322

HIGH-PRESSURE METAL VAPOR ELECTRIC DISCHARGE LAMP Filed July 21, 1947 2 Sheets-Sheet l Waii'gge in KiLowdi'f' a Q 2 a Q Q EFFicienc Inventors in Lumens e'r WdiT G' VlcTor' \J. Fr'dncls,

Evan H. NeLson,

CIS ETAL 2,650,322 HIGHGPRESSURE METAL VAPOR ELECTRIC DISCHARGE LAMP v. J. FRAN Aug. 25, 1953 2 sheet-sneez 2 Filed July 21, 1947 Inventor's Victor J. Francis; 1 EvanHNeLson,

Their A'bcofneg.

Patented Aug. 25, 1953 HIGH-PRESSURE METAL VAPOR ELECTRIC DISCHARGE LAMP Victor James Francis, North Wembley, and Evan Herbert Nelson,

Harrow Weald, England, as-

signors to General Electric Company, a corporation of New York Application July 21, 1947, Serial No. 7 62,420 In Great Britain September 24, 1945 Section 1, Public Law 690, August 8, 1946 Patent expires September 24, 1965 4 Claims. (Cl. 313214) This invention relates to high pressure metal vapour electric discharge lamps, that is to say electric discharge lamps in which, during normal operation of the lamp, the pressure of the metal vapour is about one atmosphere or more. It is particularly, but not exclusively, concerned with lamps for general illumination purposes, for example street lighting.

This application is a continuation-in-part o co-pending application Serial No. 690,632, filed August 15, 1946, and which is assigned to the assignee of this application.

It has long been known that the colour of a mercury vapour discharge operating at about one atmosphere or more is deficient in the red region and that this deficiency can be reduced by the addition of cadmium or Zinc vapour, or a mixture of both of these vapours.

Attempts have therefore been made in the past to produce a high pressure mercury vapour electric discharge lamp giving light containing a useful percentage of red by adding cadmium and/or zinc to the mercury filling. But these attempts were not entirely successful in that either the red content of the light was still not sufficiently great or, if an adequate percentage of red was obtained, the luminous efficiency of the lamp was undesirably low. Accordingly the impression was gained that even if a high pressure mercury vapour lamp giving light containing an adequate percentage of red could be obtained in this way, the sacrifice of efliciency would make the lamp commercially useless.

We have now found that this is notthe case, and the main object of this invention is to pro: vide a design of a high pressure metal vapour electric discharge lamp containing the vapours of mercury and cadmium or Zinc or both with which light including a useful percentage of red is emitted and, nevertheless, capable of giving a satisfactory luminous efiiciency.

For the purposes of this specification a useful percentage of red light may be regarded as a red ratio of 8% or more; here and hereinafter the expression red ratio means the percentage of the incident light which is transmitted through a Wratten 25 filter. The definition of a satisfactory luminous efficiency will be considered in detail later.

In order that the invention may be fully un-- derstood it will be as well to state first the reasons, which our investigations have led us to appreciate, why the earlier efforts to produce satisfactory red-corrected lamps failed. These 3 earlier efforts were made with lamps of two widedischarge column ly different types and the reasons for failure were different for each type.

Thus one type of lamp was the kind having a hard glass envelope, a relatively long arc length, usually several centimetres, and operating at about one atmosphere pressure; the lamps generally used for street-lighting, having an arc length of about ten to twenty centimetres, are examples of this type, and the main reason for failure here was that the glass envelope could not be operated at a sufiiciently high temperature to vapourise enough cadmium or zinc; we do not know whether the red ratio would have been satisfactory if the envelope had in fact been run hotter, but we do know that the efiiciency would have been much less than with the mercury filling only. In particular this type of lamp had too low a value of the ratio W/L, where W was the wattage dissipated by the lamp in normal operation and L the length of the discharge column i. e. the distance between the running electrodes; we have now found it to be essential for satisfactory red-corrected lamps that W/L should not be less than 250 watts per centimetre whereas in these known lamps W/L was only about 20 watts per centimetre.

Here it may be noted that the terms are and are used synonymously throughout this specification.

A second type of lamp, in which we have made improvements, and which is commonly referred to as a high pressure, short-gap lamp is one in which the spacing between the electrodes or the arc length is substantially less than the maximum transverse dimension or diameter of the enclosing envelope or bulb.

The second type of lamp Was of the compact source type used for optical projection, having a quartz envelope, a short arc length, usually less than one centimetre, and operating at about 10 atmospheres pressure; we now know thatthis type of lamp had usually a more than sufiiciently high value of the ratio W/L, theratio usually being as high as 1000 or more, but a sufficiently high red ratio was not obtained since the mercury vapour pressure was always too high compared with the cadmium or zinc vapour pressure, the latter pressure being necessarily limited by the limitations on wall temperature i. e. the temperature of the wall of the envelope. It appears from our investigations that for a satisfactory red ratio to be obtained, the cadmium or zinc vapour pressure must be comparable with the mercury vapour pressure; this does not mean that the cadmiumor zinc vapour pressuremust ciency as the ratio'of from the lamp to the electrical energy dissipated in the lamp, which been required for most practical purposes. The

The vafective efliciency measured as aforesaid is always less than, and may be considerably less than, the true efficiency of the lamp; such'obscuration of the light mainly arises from the seals whereby the electrodes and leads thereto are sealed into the envelope and from the electrodes themselves;

accordingly the obscuration increases as the size ditions which our investigations haveshown to be essential to success. It will-be assumed that the lamp is to be designed to dissipate a given wattage W when run in an evacuated enclosure. These conditions are: a

l. The envelope must be so small that the ratio W/A, where A is the area of the internal surface of the envelope, is greater than (a) 10 watts/square centimetre if the metal filling comprises cadmium or (b) greater than watts/square centimetre if the metal filling comprises zinc without cadmium. Here it is to be no'ted that the internal area A means the internal area with all irregularities, such a those provided by the seals, smoothed out, the general dimensions of the envelope remaining unchanged,

"as explained in British Patent No. 485,489.

'On the other hand the ratio W/A must not be so'great that the lamp has no useful life owing to rapid deterioration of the envelope. This may be expressed by saying that the lamp must be capable'of being operated continuously at the wattage W; here and hereinafter by continuously is meant that the average useful life for the lamps must not be less than 100 hours. For

7 a quartz'envelope in vacuo W/A should not be greater than 40 Watts per square centimetre.

2. The distance L between the running electrodes of the lamp must be so small that the ratio W/L is not less than 250 watts per centimetre.

e 3. The amount and constitution of metal filling must be so great and such that in normal operationa sufliciently highvapour pressure to H achieve a satisfactory efficiency'is obtained.

' 4. The quantity of mercury in themetal filling must be so small that the cadmium or zinc vapour pressure is comparable with the mercury vapour pressure as evidenced by the red ratio of the light from the discharge column being not less i It will be understood that the definitions of A 7 and L hereinbefore given apply throughout the specification.

The question as to what constitutes a satisfactory-efficiency must now be considered and with i this isin volved the question as to how the chiv ciencyis to be measured.

Thus usual method of measurin the efficiency of ala mp is toestimate the light output by'placing the lamp in a large spherical photometer and measuring the mean intensity of illumina'tion produced on the inner surface of the This method then gives the effithe light actually issuing photometer.

is What has hitherto true-efliciency' of the lamp is, however, the ratio of the light issuing from the discharge column to the-electrical energy dissipated in the lamp and where there is any appreciable obscuration of the light issuing from the-discharge column, the ef- (A) of the envelope and the distance (L) between the electrodes decreases.

The efliciency with which we are concerned in this invention is defined below and approximates more to the true efficiency than to the effective efficiency; for where as a matter of fact the eilective eificiency is the important one,

'as' in lamps for general illumination purposes such as street lighting, the siZe of the envelope is usually so large and the distance between the electrodes so great that there is little obscuration and the true and effective efficiencies are substantially the same; but where the obscuration is great, as in small lamps for optical projection in which the electrodes are close together, the

, effective eihciency has little meaning since light is not in general required in all directions from the lamp but usually only over a zone which may be relatively small, and we must therefore consider the true efficiency. Hereinafter in this specification and in the appended claims the term fefiiciency will signify, unless otherwise stated, the true efficiency as measured by the following method:

v The lamp is to be inserted in a spherical photometer with all the envelope obscured by substantially non-reflecting material except for a zonev of known solid angle 0 through which light can emerge without obstruction, where 6 is the solid angle subtended at the centre of the discharge column (i. e. a point mid-way between the running electrodes) by the unob'scured part of the envelope and this angle should be as great as possible when marking this measurement; the

' light output E from the lamp so obscured is then to bemeasured in the usual way and the total light issuing from the discharge column to be taken as 4:7rE/0, the ratio of this total to the electrical energy dissipated in the lamp being taken as the true efficiency.

It may be noted in this connection that for lamps of wattage W not greater than 2000 watts in which the distance between the running electrodes is such that W/L is not greater than 500 watts per centimetre and in which the obscuration provided by the electrodes and seals is small,

0 is so nearly equal to 41 that the effective efficiency. may be taken as the true efiiciency and no special obscuration of the envelope need be made in measuring the efiiciency.

Figure 1 of the accompanying drawing then gives the minimum true efficiencies corresponding to wattages W which are to be regarded as satisfactory for the purposes of this invention.

,In this figure the ordinates give the efiiciencies in lumens per watt whilst the abscissae give wattages in kilowatts and the efliciency corresponding to any wattage W must not be less than that corresponding to the curve shown; thus for example for a lamp operating at 500 watts the minimum efficiency which can be regarded as satisfactory is 20 lumensper watt, at 1000 watts the minimum satisfactory efficiency is 30 lumens per watt; at 4000 watts the minimum satisfactory eficiency is 50 lumens per watt and for lamp-s operating at more than 6000 watts, the minimum .satisfactory efliciency is 55 lumens per watt, it

being understood that the curve flattens out at this ordinate for abscissae beyond 6 kilowatts.

At the other end of the scale the curve drops towards zero as shown, though it will obviously not be practically possible to design lamps in accordance with the invention operating at very low wattages.

metal filling comprising cadmium and 15 wattsper square centimetre for a filling comprising zinc without cadmium must be increased corresponding to the nature of the cooling to attain the same minimum internal wall temperatures as attained with the aforesaid minimum wall loadings in an evacuated enclosure. In practice this consideration will usually involve estimating the size of envelope required and adjusting the cooling, if necessary, to give a sufficiently high wall temperature to vapourise enough cadmium or zinc, as evidenced by the attainment of a red ratio of not less than 8%.

The invention will now be stated generally.

According to the invention a high pressure metal vapour electric discharge lamp comprises a highly refractory envelope, forexample of vitreous quartz, comprises as the metal filling, i. e. the source of metal vapour, mercury together with cadmium and/or zinc, and is so constructed and the internal area A of the envelope, the distance L between the running electrodes and the amount and constitution of the metal filling are all so related that the lamp is capable of operating continuously, with forced cooling if necessary, dissipating a wattage W such that W/A is not less than 10 watts per square centimetre if the metal filling comprises cadmium or not less than watts per square centimetre if the metal filling comprises zinc without cadmium, W/L is not less than 250 watts per centimetre, the true efiiciency of the lamp is not less than the minimum for wattage was hereinbefore defined with reference to Figure 1 of the accompanying drawings, and the content of mercury in the metal filling is so small that the light emitted from the discharge column when i the lamp is operating so as to dissipate wattage W as aforesaid has a red ratio as hereinbefore defined of not less than 8%.

For guidance in designing lamps in accordance with the invention the following general siderations are given.

Firstly the wattage W at which the lamp is to be operated must be fixed and the running conditions of the lamp must be settled, namely the orientation of the discharge column (1. e horizontal or vertical) and the cooling conditions (i. e. forced air cooled, convection cooled or running in an evacuated enclosurel. These considerations will affect the size and shape of envelope relative to the wall loading (W/A) to be used and the distance between the electrodes i. e. the arc length.

Dealing first with the question of th wall I loading, the optimum wall loading tobe ,used

con

depends on the nature of the admixture with the mercury; thus if cadmium alone is to be used a wall loading of W/A=20 watts per square centimetre is preferable; if zinc alone is to be used a wall loading of W/A= 0 watts per square centimetre is preferable; if both cadmium and zinc are to be used a wall loading of W/A=25 Watts per square centimetre is preferable. These optimum wall loadings are for operation of the lamp in an evacuated enclosure and if the lamp is to b cooled in operation the wall loadings should be correspondingly increased so as to attain approximately the same envelope inner wall temperatures; the minimum temperature of the inside of the wall of the envelope at at W/A=2ll watts per square centimetre in an evacuated nclosure is probably more than 1000 C. This shows that the envelope must be highly refractory and should accordingly be of vitreous quartz which is at present the most refractory material known that is otherwise suitable; but if a substantially equally, Or more, refractory substance is found which is otherwise suitable, such substance may be used.

The choice between cadmium and zinc or both for inclusion with the mercury as the metal filling depends mainly on the colour required in the light from the lamp. If a high red ratio is the only consideration in this respect cadmium alone is preferable, as this gives better efiiciency and can be operated with a lower wall loading, which means a longer life for the envelope; the light has however a definitegreenish tinge and if good general colour rendering properties are required a mixture of cadmium and zinc is preferable; zinc by itself gives a good red ratio but the light is deficient in green.

Next We must consider the fixing of the distance L between the running electrodes of the lamp. Here it may be stated that our experiments indicat that the best results are achieved when the voltage drop V between the electrodes is such that V/L lies between about 30-60 volts per centimetre and a value of about 50 volts per centimetre should be aimed at in designing lamps in accordance with the invention. A limitation is already set on L by the fact that W/L must be greater than 250 watts per centimetre and a further factor affecting the choice of L lies in the supply from which the lamp is to be adapted to be run. Thus if the lamp is to be run from ordinary 230 volt A. C. mains it will,as usual with high pressure mercury vapour discharge lamps, in general be desiri in order to minimise light loss by obscuration from seals and electrodes, (2) that L may have to be limited to obtain a lamp of convenient size (3) L should be small so that W/L may be ashigh as possible to get good efficiency, (4) electrode losses may increase as L decreases and (5) that it may in some cases be desirable for L to be small in order that the lamp may provide a compact source for use in optical projection apparatus.

The best compromise between all these factors must be made to suit the requirements of each particular case. It should however be remarked that our experimentshave indicated that the value of V/L should not be less than 30 volts per centimetre; also if L is to be small the possibility of voltageidropin th electrodes'of the lamp may have to be taken into account in considering the voltage to be applied across It remains then to determine the amount and constitution of the metal filling to be introduced into the envelope; here it is convenient to have excess of the zinc and/or cadmium admixture but the mercury should be all evaporated in full operation, except possibly for a small quantity which remains amalgamated with the zinc or cadmium owing to the peculiar properties of' such amalgams.

therefore be sufiiciently small to be all evaporated The quantity of mercury must yet sufficiently great to develop a vapour pressure which is not much less than, and will usually be greater than, the cadmium and/or:

zinc vapour pressure. A working rule which we have found satisfactory in many cases for obtaining the correct metal filling is as follows:

The weight of mercury which would be required to produce by itself in full operation a voltage drop between the electrodes of about 50 volts per centimetre of arc length is first ascertained, either by preliminary trial'or by calculation based on previou experience. To this weight of mercury is added about 65% by weight of cadmium or of zinc or of cadmium and zinc mixed in equal quantities so as to give a metal filling containing substantially 60% 'mercury and 40% cadmium or.

40% zinc or cadmium and 20% zine. It

will however be appreciated that these propor-' tions may be varied somewhat in either direction in order to suit different lamps or to vary the colour of the emitted light or possibly to reduce the deposition of excess cadmium or zinc on the envelope wall.

It will also be appreciated that in designing a particular lamp in accordance with the general considerations set out above, some experiments may initially have to be made in order to ascertain the optimum dimensions for the lamp and the optimum amount and constitution of metal filling to be used; such experiments are of course required in designing any high pressure mercury vapour lamp and once the optimum design has been. fixed, production to that design may proceed without further experiment.

Figure 2 of the accompanying drawing shows, approximately to scale and by way of example, the dimensions which we have found suitable for a lamp adapted to dissipate 1 kw. and to be operated from 230 volt A. C. mains, the running position of the lamp being vertical, as shown in the drawing, and the lamp bein contained within an evacuated outer jacket.

The envelope l of the discharge lamp is of vitreous quartz 2 mms. thick; the greatest width a of the interior of the envelope is 40 mms. and the length b of the interior along the axis of the envelope is 50 mms. The distance 0 between the electrodes 2 and 2 is 25 mms.

The electrodes 2 and 2" themselves are of standard" construction consisting of tungsten rods 3 mms. in diameter carrying starting electrodes in the form of coils of tungsten wire 3', 3' containing thermionieally emissive material and held in grooves in'the sides of the rods 2, 2' by the bigger coils 4, 4 of tungsten wire which fit closely over the rods. The tungsten rods 2, 2 are reduced in thickness to 1.5 mms. where they enter the quartz strip seals 5, '5, which provide the leads to the electrodes.

Th lamp is supported from a metal frame 6 which is itself supported from a hard glass outer jacket I which is evacuated and provided with an Edison screw cap 8 to which the electrodes of the lamp are effectively connected, one via the metal frame 6 and the pinch 9 of the outer 'jacket and the other directly via the pinch 9.

The filling of the lamp consists of 20 milligrams of mercury and 140 milligrams of cadmium, together with the usual small amount of argon, to about 20 mms. mercury pressure, for starting.

With this lamp and this filling we have in operation at 1 kw. obtained an efliciency of 40 lumens per watt and light having a red ratio of 9.5.

It may be noted that the internal area A for this lamp is approximately square centimetres, giving a value of W/A=20 watts p'ersquare centimetre approximately and W/L=l000/25=l00 wattsper'centimetre.

' One practical point which might at this stage be mentioned is that by suitably shaping and 'placin'g'the seal-off pip ID of the lamp the excess cadmium may be caused to'collect in the pip and thereby be prevented from forming a light obscuring film elsewhere on the envelope; care must however be taken to ensure that the pip does not form a cold spot which undesirably limits the mercury vapour pressure.

There remains now to be considered'the possibilitt of further improvement of the colour of the light obtained with lamps in accordance with the invention, which whilst being satisfactory from the point of view of redness is still somewhat cold in appearance for some purposes.

We have appreciated that this possible defect may be overcome by associating a lamp in accordance with the invention with a sodium lamp so that the light from the two lamps mix and the yellow of the sodium lamp supplies the necessary warm appearance to the resultant light. The

' arrangement has the further advantage that the overall 'efiiciency can in this way be increased, since the efiiciency of a sodium lamp is usually greater than that of the mercury'cadmium and/or zinc lamp.

' According therefore to a subsidiary feature of the invention a lighting fitting comprises in combination a high pressure metal vapour electric discharge lamp in accordance with the invention as aforesaid and a discharge lamp having a filling consisting wholly or mainly of sodium.

The intensity of the sodium light should preferably be between l0%"and 20%, measured in lumens, of the whole light emitted from the fitting.

The'fitting may comprise or be associated with a diffusing optical system for'mixing the light from the two lamps, but in some cases, for example for producing decorative 'efiects in the fitting, little or no difiusion may be provided so that the golden colour of the sodium lamp is made apparent in the fitting.

outer envelope, which may be evacuated, so that the heat generated by the mercury lamp may advantageously be used to heat the sodium'lamp.

We claim:

7 1. A high pressure electric discharge device of The two lamps may be mounted in a common the short-gap type comprising a light-transmissive envelope having therein a pair of electrodes spaced at a distance less than the maximum transverse dimension of said envelope, an ionizable medium within said envelope and capable of supporting an arc discharge and comprising mercury and a metal from the group consisting of cadmium and zinc and mixtures thereof, the spacing of said electrodes and the area of the envelope being such that the number of watts per centimeter of arc length is not less than 250 and the wall loading in watts per square centimeter is not less than and not greater than 40 to produce sufficiently high wall temperature to cause vaporization of said metal as evidenced by the attainment of a red ratio measured through a Wratten filter of not less than 8 per cent at an operating efiiciency of not less than 40 lumens per watt.

2. A high pressure electric discharge lamp comprising a light-transmissive envelope having therein a pair of spaced electrodes, the spacing between said electrodes being less than the maximum transverse dimension of said envelope, an ionizable medium within said envelope and capable of supporting an arc discharge and comprising about 60 per cent mercury and 40 per cent zinc, the spacing of said electrodes and the area of the envelope being such that the number of Watts per centimeter of arc length is not less than 250 and the wall loading expressed in watts per centimeter is not less than 10 to produce a sufficiently high wall temperature causing vaporization of said Zinc as evidenced by the attainment of a red ratio measured through a Wratten 25 filter of not less than 8 per cent at an efliciency of not less than 40 lumens per watt.

3. A high pressure electric discharge lamp comprising a light-transmissive envelope having therein a pair of spaced electrodes, the spacing between said electrodes being less than the maximum transverse dimension of said envelope, an ionizable medium within said envelope and capable of supporting an arc discharge and comprising about 60 per cent mercury and 40 per cent cadmium, the spacing of said electrodes and the area of the envelope being such that the number of watts per centimeter of arc length is not less than 250 and the wall loading expressed in watts per centimeter is not less than 10 to produce a sufiiciently high wall temperature causing vaporization of said cadmium as evidenced by the attainment of a red ratio measured through a Wratten 25 filter of not less than 8 per cent at an efficiency of not less than 40 lumens per watt.

4. A high pressure electric discharge device of the short-gap type comprising a light-transmissive envelope having a pair of electrodes spaced at a distance less than the maximum transverse dimension of said envelope, an ionizable medium within said envelope and capable of supporting an arc discharge and comprising mercury and a metal from the group consisting of cadmium and zinc and mixtures thereof, the spacing of the electrodes and the area of the envelope being such that the number of watts per centimeter of arc length lie within the range from 250 to 500, inclusive, the voltage per centimeter of arc length is about 50, and the wall loading in watts per square centimeter lies within the range from 10 to 30, inclusive, to give a sufficiently high wall temperature causing vaporization of enough of said metal as evidenced by the attainment of a red ratio of not less than 8 per cent at an efficiency of not less than 40 lumens per watt.

VICTOR JAMES FRANCIS. EVAN HERBERT NELSON.

References Cited in the file of this patent UNITED STATES PATENTS Number Name Date 2,278,844 Francis Apr. '7, 1942 2,298,965 Pellman Oct. 13, 1942 FOREIGN PATENTS Number Country Date 428,084 Great Britain May 7, 1935 431,382 Great Britain July 8, 1935 431,409 Great Britain July 8, 1935

Images