The invention relates to electric lamps and particularly to electric discharge lamps. More particularly the invention is concerned with electric discharge lamps with restrike capability.
When the arc in an operating high intensity discharge (HID) lamp is momentarily interrupted so as to be extinguished, the high electrical resistance of the hot lamp fill then makes the immediate striking of a new arc difficult. One then has to wait for the lamp to cool, or use a sufficiently high voltage to overcome the temporary high resistance. This is the hot restrike problem. Hot restrike may be overcome in several ways, but usually by the application of a high electric field between the main electrodes, for example, by using 26 kilovolts or more across the 4 millimeter arc gap in an automotive HID lamp. This brute force approach is used successfully in automotive HID lamps, where the need to relight the HID headlamps quickly is critical to safe nighttime driving. However, the high voltage source is expensive, and the high voltage, if not safely contained, may be dangerous. For HID lamps to be competitive in consumer markets, it is essential that the ignition voltage and especially the re-ignition voltage of an HID lamp be low. Consumer safety is of paramount importance, and, from a practical standpoint, many existing lamp fixtures are not safety rated for operation above 5 kilovolts. There is then a need to provide an HID lamp with rapid restrike ability without the use of extra high voltage.
A high intensity discharge lamp is constructed with an envelope having a wall defining an enclosed volume, a fill chemistry and a fill gas positioned in the enclosed volume. A first main electrode having an exterior end and an interior end is extended through the wall in a sealed fashion with the interior end positioned in the enclosed volume. A second main electrode having an exterior end and an interior end, the second main electrode is also extended through the wall in a sealed fashion with the interior end positioned in the enclosed volume. The interior end of the first main electrode is offset from the interior end of the second main electrode. The first main electrode and the second main electrode define between them in normal lamp operation a region of plasma discharge. At least a first starting electrode having an exterior end and an interior end is extended through the wall in a sealed fashion with the interior end positioned in the enclosed volume. The first main electrode may be electrically coupled to the first starting electrode by an impedance element. A second starting electrode having an exterior end and an interior end is also extended through the wall in a sealed fashion with the interior end positioned in the enclosed volume. The interior end of the first starting electrode is offset from the interior end of a second starting electrode; and aligned such that a line from the interior end of the first starting electrode to the interior end of the second starting electrode crosses through or adjacent to the region of plasma discharge formed between the interior end of the first main electrode and the interior end of the second main electrode during lamp operation. The first starting electrode includes a thermo-mechanical element intermediate the wall and the interior end of the first starting electrode such that when the thermo-mechanical element is in a cool state the interior end of the first starting electrode is in a first position, and when the thermo-mechanical element is in a heated state the interior end of the first starting electrode is in a second position. The starting electrode positioning is also such that in a cold state the impedance from the exterior end of the first main electrode through the impedance device, if any, and through the first starting electrode in the first position, to the second starting electrode is less than the impedance from the exterior end of the first main electrode through the interior end of the first main electrode to the second main electrode. The starting electrode positioning is also such that in a hot steady, operating state the impedance from the exterior end of the first main electrode through the impedance device, if any, and through the first starting electrode in the second position, to the second starting electrode is greater than the impedance from the exterior end of the first main electrode through the interior end of first main electrode to the second main electrode.
The envelope 12 may be made from any of numerous light transmissive ceramics as known in the art, including such materials as vitreous silica, also known as quartz, polycrystalline alumina (PCA) and others. The envelope 12 includes a wall 24 defining an enclosed volume 26, and one or more seal regions 28, 29.
Contained in the enclosed volume 26 is an appropriate chemical fill 14 that is non-reactive with respect to the two (or more) enclosed elements comprising the bimetallic starting electrode element assembly. Such fill 14 chemistries may include the metal halide fills (salts) well known in the art of lamp making. A preferred chemistry is a mixture of iodide salts of such metals as Thallium, Dysprosium, Holmium, Sodium, and Thulium along with metallic mercury. A neutral fill gas is included in the fill 14, such as a noble gas. The preferred fill gas is argon with a cold pressure of 20,000 Pascals (about 150 torr). Other fill 14 chemistries and gases are known and may be used, provided the components are not reactive with the bimetallic starting electrode assembly elements.
The first main electrode 16 has an exterior end 30 and an interior end 32. The first main electrode 16 is extended through the seal region 28 of the envelope 12 wall 24 in a sealed fashion with the interior end 32 positioned in the enclosed volume 26. The first main electrode 16 may have any of the numerous material and structural configurations of HID lamp electrodes. The preferred electrode has a tungsten inner shaft end.
The second main electrode 18 also has an exterior end 34 and an interior end 36, and is extended through the seal region 29 of the envelope wall 24 in a sealed fashion with the interior end 36 positioned in the enclosed volume 26. The interior end 32 of the first main electrode 16 is axially offset from the interior end 36 of the second main electrode 18. The first main electrode 16 and the second main electrode 18 define between them in normal lamp operation an arc region 38, generally surrounding the least straight line between interior ends 32, 36 of the two main electrodes, between which a plasma discharge forms.
The first starting electrode assembly 20 has an exterior end 40 and an interior end 42, and is extended through the wall 24 in a sealed fashion with the interior end 42 positioned in the enclosed volume 26.
The second starting electrode assembly 22 also has an interior end 44 and an exterior end 48, and is extended through the seal region 29 of envelope wall 24 in a sealed fashion with the interior end 44 positioned in the enclosed volume 26. The interior end 42 of the first starting electrode assembly 20 is laterally offset from the interior end 44 of the second starting electrode assembly 22. The interior end 42 of the first starting electrode assembly 20 and the interior end 44 of the second starting electrode assembly 22 are aligned in a cold state such that the distance 52 from the interior end 42 of the first starting electrode assembly 20 to the interior end 44 of the second starting electrode assembly 22 is less than the distance 50 from the interior end 32 of the first main electrode 16 to the interior end 36 of the second main electrode 18. Further, a least line 52 from the interior end 42 of the first starting electrode assembly 20 to the interior end 44 of the second starting electrode assembly 22 crosses through or near the defined arc region 38 formed between the interior end 32 of the first main electrode 16 and the interior end 36 of the second main electrode 18 during normal lamp operation.
In one embodiment, the thermo-mechanical element 60 was formed from a 0.228 (0.009 inch) thick molybdenum substrate joined to a corresponding 0.033 millimeter (0.0013 inch) thick tungsten ribbon to form a 25 millimeter long by 4 millimeter wide structurally sound tungsten-molybdenum bi-metallic shaft like structure. The bimetallic structure was fabricated by laser welding a continuous seam around the rectangular perimeter of the composite strip. The welding was done while flowing an argon atmosphere around the weld region to preclude oxidation of the molybdenum and tungsten components. The molybdenum and tungsten strips could have been welded together by ultrasonic means or by hot roll bonding.
To facilitate deflection testing of the bimetallic strip by external radiative heating from a torch flame, without the risk of oxidation damage, the bimetallic strip was mounted cantilever style inside an evacuated and sealed cylindrical quartz capsule (15 millimeter outside diameter by 13 millimeter inside diameter by 64 millimeter long). Heating the capsule to about 900 degrees Celsius produced about a 2.5 millimeter lateral deflection of the free end of the tungsten-molybdenum bimetallic strip, bending away from the molybdenum side. The bi-metallic bending demonstrated that significant lateral deflection of the free end of the tungsten-molybdenum bimetallic strip could be implemented as an auxiliary variable length arc gap in an HID arc tube. The difference between the thermal expansion coefficients of molybdenum and tungsten is quite small, on the order of 1.5 ppm/C, over the temperature range of interest (room temperature to about 1000 degrees Celsius).
The thermally induced deflection of the free end of the bimetallic strip, the other end being clamped, is nominally proportional to the product of the difference in thermal expansion coefficients of the two metals and the change in temperature to which the bimetallic strip is exposed. In many other lower temperature applications using other metals, the difference between the thermal expansion coefficients is greater, for example, by about 20 ppm/C, but the operational temperature range is smaller.
Theory predicted that the thermally induced deflection could be substantially enhanced by reducing the overall thickness of the strip or by lengthening the strip. For validation purposes, a second tungsten-molybdenum bi-metallic element, 50.8 millimeter long, 3.2 millimeter wide, and 0.066 millimeters (0.0026 inch) thick, was fabricated using the same laser welding process. The bi-metallic strip was then sealed in a larger cylindrical quartz capsule (50 millimeter outside diameter by 47 millimeter inside diameter by 80 millimeter long). Subjected to heating conditions similar to the first case (900 degrees Celsius in 100 torr of nitrogen), the tip of the single bi-metallic strip deflected laterally about 20 millimeters, more than enough for use in a 400 watt metal halide HID arc tube (about 18 millimeters inside diameter).
External current limiting resistors may be used in series with the auxiliary bimetallic starting electrodes. Because the free ends of the bimetallic strips are closer to each other than are the tips of the corresponding main electrodes, a small startup arc forms between the free ends of the bimetallic strips when an appropriate ignition pulse is applied to the arc tube. An extra tungsten tip is preferably welded to the free end of each bimetallic strip to withstand the elevated arc temperatures when the starting arc is initiated.
The startup arc heats the bimetallic assemblies, causing the respective free ends to deflect away from the mid line of the arc tube and toward the side wall of the cylindrical capsule. At some point as the bimetallic starting electrodes laterally deflect and the tip to tip distance increases, and the startup arc extinguishes. This occurs because the effective resistance of the arc gap increases with increasing arc gap length. As a result, the voltage (or power) across the starting electrode gap increases, until the needed voltage (or power) exceeds the output capability of the ballast power supply, thereby extinguishing the startup arc.
The residual ionization produced from the startup arc of the surrounding gas between the main electrodes 16, 18 reduces the impedance between the main electrodes 16, 18 (maintained at full open circuit potential), allowing the main arc between the main electrodes 16, 18 to ignite. Once ignited the main arc heats the lamp, including the starting electrodes 20, 22. The sustaining heat generated by the main arc causes the bi-metallic elements of the starting electrodes 20, 22 to remain deflected or to splay farther apart, causing the starting electrode tips 42, 44 to reside near or at the side wall of the arc tube during steady state operation. The main arc may ignite before the startup arc extinguishes. If that happens, the voltage across the main arc decreases from open circuit ignition values to much lower steady state running values, effectively shutting off the startup arc. The net result is the same as if the startup arc had extinguished by sufficient lateral deflection of the starting electrodes.
When power to the arc tube is shut off, the free ends of the fully-splayed bimetallic elements 60 cool. The interior ends 42, 44 of the starting electrodes 20, 22 begin to re-approach each other because heat is drawn from the bimetallic elements 60, primarily by thermal conduction through the corresponding starting electrode components to the exterior of the arc tube, and also to a lesser extent by radiation and convection. The conduction heat transfer process is more rapid (estimated to be at least 10 or more times faster) than the heat transfer from the entire arc tube (primarily by radiation) to the external environment. As the effective gap between the bimetallic tungsten tips decreases toward the original room temperature midline condition, re-ignition of the startup arc is possible, repeating the cycle described above.
The cool down of the bimetallic elements 60 (and the consequent reduction in their effective arc gap length) is faster than the overall cool down of the arc tube enclosure and the enclosed gas fill. The ability to relight (re-ignite) the main arc after the main arc is extinguished is then substantially improved. In heating up experiments with the 50.8 millimeter long tungsten molybdenum bi-metallic strip in a quartz test vessel, the free end of the starting electrode returned about 20 millimeters in about 25 seconds. This return deflection time delay is significantly shorter than the typical 10-minute re-ignition time required for a 400-watt HID quartz lamp of conventional construction (without the bimetallic starting electrodes). It is anticipated that the return deflection time of the bi-metallic starting electrodes in an arc tube as described herein will diminish because heat transfer from the extinguished main arc to and through the bi-metallic starting electrode will be dominated by thermal conduction rather than by radiation.
A ceramic arc tube using the same principles can be made using two single-ended arc tubes. Two modest disadvantages are the additional cost of the bimetallic strips, and the slight shading (light obscuration) caused by the bi-metallic strips. It is believed that the slight lumen loss is a small price to pay for rapid restrike ability. It should be understood that While there have been shown and described what are at present considered to be the preferred embodiments of the invention, it will be apparent to those skilled in the art that various changes and modifications can be made herein without departing from the scope of the invention defined by the appended claims.