CA1181471A - Tungsten lamp filament welded joint to lead - Google Patents
Tungsten lamp filament welded joint to leadInfo
- Publication number
- CA1181471A CA1181471A CA000390548A CA390548A CA1181471A CA 1181471 A CA1181471 A CA 1181471A CA 000390548 A CA000390548 A CA 000390548A CA 390548 A CA390548 A CA 390548A CA 1181471 A CA1181471 A CA 1181471A
- Authority
- CA
- Canada
- Prior art keywords
- tungsten
- filament
- lead wire
- lead
- joint
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Landscapes
- Vessels And Coating Films For Discharge Lamps (AREA)
Abstract
TUNGSTEN LAMP FILAMENT WELDED JOINT TO LEAD
ABSTRACT OF THE DISCLOSURE
In a joint between a coiled tungsten filament of an electric lamp and molybdenum lead, the lead connects integrally into a weld section which extends longitudinally within the coiled filament. The weld section may be formed by suddenly melting that portion of the lead involved in the joint, by pulse discharge, within an inert or reducing gas. The melted molybdenum wets the tungsten coil and is drawn by capillarity into the turns of the coil where it is rapidly chilled, and solidifies in a time interval too short for appreciable crystal growth and embrittlement to occur in the tungsten, the time interval being two seconds or less.
ABSTRACT OF THE DISCLOSURE
In a joint between a coiled tungsten filament of an electric lamp and molybdenum lead, the lead connects integrally into a weld section which extends longitudinally within the coiled filament. The weld section may be formed by suddenly melting that portion of the lead involved in the joint, by pulse discharge, within an inert or reducing gas. The melted molybdenum wets the tungsten coil and is drawn by capillarity into the turns of the coil where it is rapidly chilled, and solidifies in a time interval too short for appreciable crystal growth and embrittlement to occur in the tungsten, the time interval being two seconds or less.
Description
~B~
- ~ - LD 8~48 TU _STEN LAMP FILAMENT WELDED JOINT TO LEAD
Our invention relates to the attachment o:E the eilaments to the supports or inleads in electric lamps and sirnilar sealed devices, and provides a high quality welded joint between a coiled tungsten filament and a molybdenum lead which avoids embrittlement of the filament.
BACKGROUND OF THE I~VENTION
The most critical par-t of an electric lamp or light bulb is its filament. It is almost always made of tungsten because tungsten has the highest melting point of all known practical materials. Electric lamps are ordinarily gas-filled in order to recluce the rate oE vapor--ization of the tungsten Eilament, and in gas-filled lamps, eEEicac~ in producing light is increased by coiling the filament. In the common sizes of domestic lncandescent lamps, the so-called bread-and butter lamps, a coiled coil desi~n is generally used.
In low cost mass-produced incandescent lamps~
the ilament is attacherl to the lead wires in the simplest manner possible. The ends oE the Eilament are placed in open hoo~s Eormed on the ends of the lead wires, and these hooks are then folded back and closed by automated apparatus. The folded hooks pinch the filament and since the coiled ends are deformed as a result, the process must be restricted to ductile filaments which have not been heat-treated to achieve recrystallization.
In higher wattage or higher qualitY lamps such .~,, -- 2 - LD-8~48 as haloyen cycle and photo-projection lamps, greater filament stability together with longer :Life duration are sou~ht. In such lamps filaments are used which have been recrystalliæed by firing prior to mounting in order to eliminate internal strains and expel impurities. By so doing, the shape and dimensions of the filament are stabilized for subsequent lamp operation. However such filaments are more brittle and the brittleness makes the attachment of the fllament to the inleads or to the lead wires of the lamp frame structure much more difficult.
Since the simple hook-clamp is ruled out~ various alternative attachment techniques have been developed and that most commonly used is known as spudding. It involves forcing a wire or spud into the end turns or coils of the lS filament to achieve a tight friction fit. In a spudded joint the electrical connection is essentially a mechanical contact.
Various improvements to the simple spud have been made over the years. In the United States patent 2,449,679 van Horn (1948~ uses a spud in which screw threads have been formed either by a screw cutting operation or by winding fine wire tightly around it. U.S. patent 2,830,217 -Hodge (1958) uses a spud in which the end of the conductor is flattened to provide a spade-like tip in order to effect a Eorce fi.t with the filament coil. ~t-tempts hav~ also been made to combine spudding with welding, :Eor .instance, in U.S. paten~ 2,403,070 - Fulton (1946). In the Fulton U.S. patent, in order to assure that the portion of the fila-ment that takes the stresses and strains be free of embrittlement due to welding, the coiled filament is welded to the spud at the end remote from the entry point only.
All spudding techniques and improvements thereon, including spudding combined with welding, are essentially labor-intensive and have not been amenable to automation.
SUMMARY OF T~IE INVEN:TI ON
The object of our invention is to provide a welded joint between a lead wire and a heat-treated tungsten filament without causing deleterlous embrittlement in the filament or in the lead, and -to achieve such joint by a simple process amenable to automation.
In joining to tungsten, one cannot afford to overheat the tungsten to an extent that allows deleterious crystal growth and the resulting emhr.ittlement to take place. The attachment points of the :Eilament in particular are high stress points and embrittlement at those points must at all costs be avoided. Our success is predicated on utilizing the knowledge that embrittlement of tungsten is a Eu.nction of both temperature and time.
In a welded joint embodying our invention, a lead wire of a metal suitable for welding to tungsten connects integrally into a weld section extending longitudinally within a coiled tungsten filament portion~ The tungsten is not embrittled at the joint and this is achieved by utilizing a burst of energy to heat suddenly above its melting temperature in an inert or reducing enviromnent substantially only the small quantity of the lead that will be used in the joint. Because the quantity of melted metal is small, loss of heat to the surround including the maill part o:E the lead and the tungsten filament which is corltactec~ by the weld metal, causes a drop in temperature so rapid that tungsten embrittlement is substanti.ally avoided.
To avoid losin~ the cooling effect of the lead, the heating time or duration o:E the energy burst should not exceed one second, and we prefer to have it shorter. By virtue oE
the rapid cooling which follows the pulse, there is insufficient time for appreciable grain growth to occur in the tungsten contacted by the weld metal. In addition the molten metal is already cooling when it contacts the tungsten and this also serves to reduce the extent and time duration of heating the tungsten and further limits , . ~
~ .
~ 8~9~7~
- 4 - LD 84~8 crystal growth. The energy pulse may be provided in various ways: a pulse electric discha.rye is preferred;
other energy-emitting sources may also be used, such as an electron beam or a laser. The inert or reducing en-vironment may be an inert or reducing atmosphere or avacuum.
In a preferred technique embodying our invention, we juxtapose the end of a molybdenum lead to the colled end or leg oE a tungsten filament, preferably in such manner that the lead and the coiled leg intersect near their ends. Then, by means of a discharge or arc pulse drawn Erom a tungsten electrode in inert gas (pulsed TIG
welding) to the part of the molybdenum lead projecting beyond the coiled filament leg, the projecting lead part is caused to melt rapidly~ preferably in less than 500 milliseconds. The inert gas sustains the purity of the heated molybdenum and tungsten in the region of the arc and prevents any formation of oxide. A~s a result, the melted molybdenum wets the tungsten and is dra~n by capillarity into the turns oE the coil where it is rapidly chilled and solidifies. The time interval involved from the initiation of the pulse to the moment when the tungsten has cooled to its recrystallization temperature of about 2200C is two seconds or less and is too short to allow objectionable embrittlement of the tungsten. The time :Lnterval is pre:Eerably 750 milliseconds or less to allow no sicJnificant embrittlement to occur. Furthermore, the melted molybdenum progressively cools as it spreads and advances through the turns of the tungsten filament and is coolest at its point of furthest penetration. This ultimate point of advance is also the first point of attachment of the tungsten filament where stress tends to maximize, and our invention thus effectively achieves least embrittlement at that critical point.
DESCRIPTION OF DRAWINGS
FIG. 1 is a perspective view of a single-ended - 5 - LD 8~48 tungsten halogen incandescent lamp having welded joints embodying the invention.
FIG. 2 is a plan view looking down on the frame formed by the leads to support the filament.
FIG. 3 is an enlarged end view of the coiled filament leg resting on the lead with a welding electrode in place over the projecting portion of the lead.
FIG. 4 is a plan section through FIG. 3 along line 4-4.
FIG. 5 is an enlarged plan view of the joint in formation showing the projection portion of the lead formed into a molten ball.
FIG. 6 is a yet more enlarged view showing a completed symmetrical filament joint embodying the invention.
FIG. 7 shows an asymmetrical filament joint embodying the invention drawn to the same scale as FIG. 6.
DETAILED DESCRIPTION
. . .
Referring particularly to FIG. 1 of the drawings, the illustrated single-ended lamp is of the high intensity compact halogen-cycle type comprising a tuhular bulb or envelope 1 containing a longitudinally extending filament
- ~ - LD 8~48 TU _STEN LAMP FILAMENT WELDED JOINT TO LEAD
Our invention relates to the attachment o:E the eilaments to the supports or inleads in electric lamps and sirnilar sealed devices, and provides a high quality welded joint between a coiled tungsten filament and a molybdenum lead which avoids embrittlement of the filament.
BACKGROUND OF THE I~VENTION
The most critical par-t of an electric lamp or light bulb is its filament. It is almost always made of tungsten because tungsten has the highest melting point of all known practical materials. Electric lamps are ordinarily gas-filled in order to recluce the rate oE vapor--ization of the tungsten Eilament, and in gas-filled lamps, eEEicac~ in producing light is increased by coiling the filament. In the common sizes of domestic lncandescent lamps, the so-called bread-and butter lamps, a coiled coil desi~n is generally used.
In low cost mass-produced incandescent lamps~
the ilament is attacherl to the lead wires in the simplest manner possible. The ends oE the Eilament are placed in open hoo~s Eormed on the ends of the lead wires, and these hooks are then folded back and closed by automated apparatus. The folded hooks pinch the filament and since the coiled ends are deformed as a result, the process must be restricted to ductile filaments which have not been heat-treated to achieve recrystallization.
In higher wattage or higher qualitY lamps such .~,, -- 2 - LD-8~48 as haloyen cycle and photo-projection lamps, greater filament stability together with longer :Life duration are sou~ht. In such lamps filaments are used which have been recrystalliæed by firing prior to mounting in order to eliminate internal strains and expel impurities. By so doing, the shape and dimensions of the filament are stabilized for subsequent lamp operation. However such filaments are more brittle and the brittleness makes the attachment of the fllament to the inleads or to the lead wires of the lamp frame structure much more difficult.
Since the simple hook-clamp is ruled out~ various alternative attachment techniques have been developed and that most commonly used is known as spudding. It involves forcing a wire or spud into the end turns or coils of the lS filament to achieve a tight friction fit. In a spudded joint the electrical connection is essentially a mechanical contact.
Various improvements to the simple spud have been made over the years. In the United States patent 2,449,679 van Horn (1948~ uses a spud in which screw threads have been formed either by a screw cutting operation or by winding fine wire tightly around it. U.S. patent 2,830,217 -Hodge (1958) uses a spud in which the end of the conductor is flattened to provide a spade-like tip in order to effect a Eorce fi.t with the filament coil. ~t-tempts hav~ also been made to combine spudding with welding, :Eor .instance, in U.S. paten~ 2,403,070 - Fulton (1946). In the Fulton U.S. patent, in order to assure that the portion of the fila-ment that takes the stresses and strains be free of embrittlement due to welding, the coiled filament is welded to the spud at the end remote from the entry point only.
All spudding techniques and improvements thereon, including spudding combined with welding, are essentially labor-intensive and have not been amenable to automation.
SUMMARY OF T~IE INVEN:TI ON
The object of our invention is to provide a welded joint between a lead wire and a heat-treated tungsten filament without causing deleterlous embrittlement in the filament or in the lead, and -to achieve such joint by a simple process amenable to automation.
In joining to tungsten, one cannot afford to overheat the tungsten to an extent that allows deleterious crystal growth and the resulting emhr.ittlement to take place. The attachment points of the :Eilament in particular are high stress points and embrittlement at those points must at all costs be avoided. Our success is predicated on utilizing the knowledge that embrittlement of tungsten is a Eu.nction of both temperature and time.
In a welded joint embodying our invention, a lead wire of a metal suitable for welding to tungsten connects integrally into a weld section extending longitudinally within a coiled tungsten filament portion~ The tungsten is not embrittled at the joint and this is achieved by utilizing a burst of energy to heat suddenly above its melting temperature in an inert or reducing enviromnent substantially only the small quantity of the lead that will be used in the joint. Because the quantity of melted metal is small, loss of heat to the surround including the maill part o:E the lead and the tungsten filament which is corltactec~ by the weld metal, causes a drop in temperature so rapid that tungsten embrittlement is substanti.ally avoided.
To avoid losin~ the cooling effect of the lead, the heating time or duration o:E the energy burst should not exceed one second, and we prefer to have it shorter. By virtue oE
the rapid cooling which follows the pulse, there is insufficient time for appreciable grain growth to occur in the tungsten contacted by the weld metal. In addition the molten metal is already cooling when it contacts the tungsten and this also serves to reduce the extent and time duration of heating the tungsten and further limits , . ~
~ .
~ 8~9~7~
- 4 - LD 84~8 crystal growth. The energy pulse may be provided in various ways: a pulse electric discha.rye is preferred;
other energy-emitting sources may also be used, such as an electron beam or a laser. The inert or reducing en-vironment may be an inert or reducing atmosphere or avacuum.
In a preferred technique embodying our invention, we juxtapose the end of a molybdenum lead to the colled end or leg oE a tungsten filament, preferably in such manner that the lead and the coiled leg intersect near their ends. Then, by means of a discharge or arc pulse drawn Erom a tungsten electrode in inert gas (pulsed TIG
welding) to the part of the molybdenum lead projecting beyond the coiled filament leg, the projecting lead part is caused to melt rapidly~ preferably in less than 500 milliseconds. The inert gas sustains the purity of the heated molybdenum and tungsten in the region of the arc and prevents any formation of oxide. A~s a result, the melted molybdenum wets the tungsten and is dra~n by capillarity into the turns oE the coil where it is rapidly chilled and solidifies. The time interval involved from the initiation of the pulse to the moment when the tungsten has cooled to its recrystallization temperature of about 2200C is two seconds or less and is too short to allow objectionable embrittlement of the tungsten. The time :Lnterval is pre:Eerably 750 milliseconds or less to allow no sicJnificant embrittlement to occur. Furthermore, the melted molybdenum progressively cools as it spreads and advances through the turns of the tungsten filament and is coolest at its point of furthest penetration. This ultimate point of advance is also the first point of attachment of the tungsten filament where stress tends to maximize, and our invention thus effectively achieves least embrittlement at that critical point.
DESCRIPTION OF DRAWINGS
FIG. 1 is a perspective view of a single-ended - 5 - LD 8~48 tungsten halogen incandescent lamp having welded joints embodying the invention.
FIG. 2 is a plan view looking down on the frame formed by the leads to support the filament.
FIG. 3 is an enlarged end view of the coiled filament leg resting on the lead with a welding electrode in place over the projecting portion of the lead.
FIG. 4 is a plan section through FIG. 3 along line 4-4.
FIG. 5 is an enlarged plan view of the joint in formation showing the projection portion of the lead formed into a molten ball.
FIG. 6 is a yet more enlarged view showing a completed symmetrical filament joint embodying the invention.
FIG. 7 shows an asymmetrical filament joint embodying the invention drawn to the same scale as FIG. 6.
DETAILED DESCRIPTION
. . .
Referring particularly to FIG. 1 of the drawings, the illustrated single-ended lamp is of the high intensity compact halogen-cycle type comprising a tuhular bulb or envelope 1 containing a longitudinally extending filament
2 of coiled tungsten wire, here shown as a coiled coil.
The bulb contains a gas filling consisting primarily of an inactive gas such as nitrogen, or an inert gas such as argon, or mixtures of such gases at a pressure of several hundred torr and even substantially exceeding atmospheric.
Incluc~ed in the gas filling is a smal]. quantity of iodine or bromine vapor which serves as a reyenerative getter to maintain the walls of the bulb free from deposits of tungsten vapori~ed from the filament. In order to maintain the tungsten halogen regenerative cycle in operation, the bulb walls must be maintained at a relatively high temperature in excess of 250C, for example about 500C.
Accordingly, bulb 1 is made of glass having a relatively high softening point, such as one of the well-known ;;~
so-called hard glasses like borosilicate or aluminosilicate glasses, or quartz glass. The filament 2 is of any desired capacity sufficient to main-tain the required bulb wall temperature, for example about 150 watts or more for operation from a conventional power supply source. The illustrated filament is actually for operation at 150 watts from a 115 volt supply.
The filament 2 is supported along the longitudinal axis oE the bulb by short and long inner lead wires 3 and 4, respectively, terminating in transverse arms 3a and 4a extending towards the axis of the bulb. The lead wires are ~uitably made of ductile molybdenum sufficiently rigid to support the filament, and they are embedded in and extend through a pinch seal 5 in the lower end of the lamp. If desired, the portions of the lead wires which are embedded in the seal 5 may be precoated or beaded with glass in known manner to facilitate sealing. Although the lead wires 3, 4 are here shown as made of a single continuous length of molybdenum wire, composite wires are frequently used in which the externally projecting portions of the leads consist of wire of different composi-tion such as a nicle-iron alloy. Also when the envelope 1 consists of quartz or fused silica, the leads 3, 4 will include thin foil portions to assure a hermetic seal in the region of the compressed pinch seal portion 5. After the envelope has been pinch-sealed, the lamp is flushed and filled with the gas mixture through an exhaust tube attached to the top of the bulb and which leaves the vestige 6 when tipped off.
The present l.amp is intended as a high performance lamp and, accordingly, the filament ~ has been fired in order to stabilize the fibrous crystal structure of the tungsten and in order to drive out any contaminants. However, fired tungsten filaments are brittle and in the past this has made the attachment of the filament to the lamp frame difficult. The technique 7~
Eormerly used with the present lamp was to insert spuds into the filament legs, secure them with resistance welds and then weld the spuds to the lead wires. The process was expensive and relatively unreliable because it required precision handling of many parts.
Our invention provides a welded jolnt which is more reliable than a spudded joint and cheaper to make.
Molybdenum whose melting point is about 2620C is a good metal for welding to tungsten which has a melting point oE
about 3370C. In welding the tungsten filament to the mol~bdenum lead wires 3, 4 by the process of our invention, it is convenient to utilize a so-called hairpin wire frame as shown in FIG. 2. The Erame includes the leads 3 and ~ joined by a bridging portion 7 which maintains the parts in the desired spacial relationship during assembly and welding. By way of example the leads 3 and 4 are approximately 0.02a" wire and the filament consists of 0.002" tungsten wire. The coiled coil filament 2 is laid to bridge the gap between the transversely turned arms 3a and 4a of the frarne with the linear coiled leg por-tions crossing and extending beyond the arms. A suitable ~ixture, not shown in the drawing, holds the filament and frame in -the relative positions desired for them subsequent to welding. Referring to FIGS. 3 and 4, the length L of the underhang or projection of end portion 3b (or 4b) oE
the arm beyond the filament is determined on the basis o~
the volume o~ molybdenum required for the weld. E`or a connectlon that will extend in both directions from the cross-over point X, there must be a volume of molybdenum suEficient to fill several turns of the filament leg ex--tending in both directions from the intersection. The weld section extends within the filament a distance at least equal to and preferably greater than the diameter of the lead. The length of coiled filament that will be filled by molybdenum is indicated by M in FIG. 4.
The hairpin frame is grounded, preferably at a ., ,~
point of arm 3a close to the intersection with the filament leg, as schematically indicated in FIGS. 3 and 4, and in similar fashion (not shown) at a point of arm 4a. The tungsten electrode 8 of the tungsten insert gas (q`IG) welding equipment is placed verticall~ over the projecting portion 3b of the molybdenum lead. The electrode tip should be close enough to the filament 2 to allow some heat splash or spillover to the filament but the distance from tip to lead should be less than that from tip to filament to assure that the arc discharge occurs to the lead and not to the Eilament. The electrode is sur.rounded by a ceramic tube or sleeve 9 through which an inert cover gas is discharged to envelop the weld region and shield it from air. A suitable gas is argon, preferably with a small percentage (5%) of hydrogen that produces a hotter arc and provides a reducing atmosphere. The reducing atmosphere removes oxides in the braze or weld area and promotes metal flow.
To make the joint, the inert cover gas is turned on and a pulse of electric current, suitably about 100 milliseconds in time duration, is discharged between the electrode and the molybdenum leadO The molybdenum lead end 3b melts and balls up into a sphere 3c over which the Eilament leg 2 extends as shown in FIG. 5. Within a very short time interval the molten molybdenum wets the tungsten filament and is drawn by capillary action into its turn~s approximately equal distances 3d and 3e on each s:ide of the intersection with the lead 3a as shown in FIG. 6~ Prior to welding, the :Eilament leg 2 passed above the lead 3a as shown in FIG. 3 but the fixture (not shown) which holds the :Eilament in place tended to force the leg into the same plane as the lead. When the projecting portion 3b becomes molten, the pressure exerted by the fixture and also surface tension cause the filament leg to center itself with respect to the lead leg 3a so that their centerlines now share the common point X. Loss of heat by radiation, to the filament and to the frame by conduction, and to the cover gas by convection, rapidly cools and soli.difies the weld section. The weld from the long lead 4 may be made simultaneously in the same way.
The cover gas is now stopped and the cycle is complete.
After the hairpin frame has been pinch-sealed in-to a lamp envelope as shown in FIG. 1, the outer bridging portion 7 is cut off to leave leads of appropriate length emerging from the lamp envelope.
In FIG. 6 a symmetrical joint is illustrated in which the weld section, that is the extent of molybdenum penetration into the filament leg, on each side oE the lead is equal. FIG. 7 i.llustrates an asymmetrical joint which is achieved by shortening the extent of projection of the filament leg beyond the lead prior to ma~ing the weld. ~he molybdenum then readily fills the short distance 3d' to the cut end on the left because there is less heat loss from it and it is hotter~ The molybdenum also penetrates the distance 3e to the right which may be longer or shorter depending upon the volume of melted metal~ An asymmetrical weld may be used to reduce waste oE filament since it is only the portion to reduce waste o:E filament since it is only the portion of the Eilament leg continuing into the coil that is useful. In fact the lead may be weldecl into the very end of the filament leg i:E desired to reduce waste to a minimum. On the other hand, a symme-trical joint as illustrated in FIG. 6 ha~ the adva.ntage that the variability in effective filament length and the resultant volume oE weld metal due to tolerance or error in the length of the underhang is reduced in half.
The more important process parameters which must be controlled to make good joints between lead and coiled tungsten filament are the following:
l. Length of underhang.
The length of the underhang controls directly "^ !
7~L
the volume of melted metal, and the dimension Mo It must also be sufficient to permit the electrical discharge to take place to the lead portion 3b and not to the tungsten filament.
2. Heating A quantity of electrical energy is supplied sufficient to melt the volume of molybdenum chosen for the weld. The arc power, that is the rate at which heat is supplied, is preferably optimized to melt the molybdenum as rapidly as possible without causing uncontrolled spattering or excessive vaporiza-tion of material.
The bulb contains a gas filling consisting primarily of an inactive gas such as nitrogen, or an inert gas such as argon, or mixtures of such gases at a pressure of several hundred torr and even substantially exceeding atmospheric.
Incluc~ed in the gas filling is a smal]. quantity of iodine or bromine vapor which serves as a reyenerative getter to maintain the walls of the bulb free from deposits of tungsten vapori~ed from the filament. In order to maintain the tungsten halogen regenerative cycle in operation, the bulb walls must be maintained at a relatively high temperature in excess of 250C, for example about 500C.
Accordingly, bulb 1 is made of glass having a relatively high softening point, such as one of the well-known ;;~
so-called hard glasses like borosilicate or aluminosilicate glasses, or quartz glass. The filament 2 is of any desired capacity sufficient to main-tain the required bulb wall temperature, for example about 150 watts or more for operation from a conventional power supply source. The illustrated filament is actually for operation at 150 watts from a 115 volt supply.
The filament 2 is supported along the longitudinal axis oE the bulb by short and long inner lead wires 3 and 4, respectively, terminating in transverse arms 3a and 4a extending towards the axis of the bulb. The lead wires are ~uitably made of ductile molybdenum sufficiently rigid to support the filament, and they are embedded in and extend through a pinch seal 5 in the lower end of the lamp. If desired, the portions of the lead wires which are embedded in the seal 5 may be precoated or beaded with glass in known manner to facilitate sealing. Although the lead wires 3, 4 are here shown as made of a single continuous length of molybdenum wire, composite wires are frequently used in which the externally projecting portions of the leads consist of wire of different composi-tion such as a nicle-iron alloy. Also when the envelope 1 consists of quartz or fused silica, the leads 3, 4 will include thin foil portions to assure a hermetic seal in the region of the compressed pinch seal portion 5. After the envelope has been pinch-sealed, the lamp is flushed and filled with the gas mixture through an exhaust tube attached to the top of the bulb and which leaves the vestige 6 when tipped off.
The present l.amp is intended as a high performance lamp and, accordingly, the filament ~ has been fired in order to stabilize the fibrous crystal structure of the tungsten and in order to drive out any contaminants. However, fired tungsten filaments are brittle and in the past this has made the attachment of the filament to the lamp frame difficult. The technique 7~
Eormerly used with the present lamp was to insert spuds into the filament legs, secure them with resistance welds and then weld the spuds to the lead wires. The process was expensive and relatively unreliable because it required precision handling of many parts.
Our invention provides a welded jolnt which is more reliable than a spudded joint and cheaper to make.
Molybdenum whose melting point is about 2620C is a good metal for welding to tungsten which has a melting point oE
about 3370C. In welding the tungsten filament to the mol~bdenum lead wires 3, 4 by the process of our invention, it is convenient to utilize a so-called hairpin wire frame as shown in FIG. 2. The Erame includes the leads 3 and ~ joined by a bridging portion 7 which maintains the parts in the desired spacial relationship during assembly and welding. By way of example the leads 3 and 4 are approximately 0.02a" wire and the filament consists of 0.002" tungsten wire. The coiled coil filament 2 is laid to bridge the gap between the transversely turned arms 3a and 4a of the frarne with the linear coiled leg por-tions crossing and extending beyond the arms. A suitable ~ixture, not shown in the drawing, holds the filament and frame in -the relative positions desired for them subsequent to welding. Referring to FIGS. 3 and 4, the length L of the underhang or projection of end portion 3b (or 4b) oE
the arm beyond the filament is determined on the basis o~
the volume o~ molybdenum required for the weld. E`or a connectlon that will extend in both directions from the cross-over point X, there must be a volume of molybdenum suEficient to fill several turns of the filament leg ex--tending in both directions from the intersection. The weld section extends within the filament a distance at least equal to and preferably greater than the diameter of the lead. The length of coiled filament that will be filled by molybdenum is indicated by M in FIG. 4.
The hairpin frame is grounded, preferably at a ., ,~
point of arm 3a close to the intersection with the filament leg, as schematically indicated in FIGS. 3 and 4, and in similar fashion (not shown) at a point of arm 4a. The tungsten electrode 8 of the tungsten insert gas (q`IG) welding equipment is placed verticall~ over the projecting portion 3b of the molybdenum lead. The electrode tip should be close enough to the filament 2 to allow some heat splash or spillover to the filament but the distance from tip to lead should be less than that from tip to filament to assure that the arc discharge occurs to the lead and not to the Eilament. The electrode is sur.rounded by a ceramic tube or sleeve 9 through which an inert cover gas is discharged to envelop the weld region and shield it from air. A suitable gas is argon, preferably with a small percentage (5%) of hydrogen that produces a hotter arc and provides a reducing atmosphere. The reducing atmosphere removes oxides in the braze or weld area and promotes metal flow.
To make the joint, the inert cover gas is turned on and a pulse of electric current, suitably about 100 milliseconds in time duration, is discharged between the electrode and the molybdenum leadO The molybdenum lead end 3b melts and balls up into a sphere 3c over which the Eilament leg 2 extends as shown in FIG. 5. Within a very short time interval the molten molybdenum wets the tungsten filament and is drawn by capillary action into its turn~s approximately equal distances 3d and 3e on each s:ide of the intersection with the lead 3a as shown in FIG. 6~ Prior to welding, the :Eilament leg 2 passed above the lead 3a as shown in FIG. 3 but the fixture (not shown) which holds the :Eilament in place tended to force the leg into the same plane as the lead. When the projecting portion 3b becomes molten, the pressure exerted by the fixture and also surface tension cause the filament leg to center itself with respect to the lead leg 3a so that their centerlines now share the common point X. Loss of heat by radiation, to the filament and to the frame by conduction, and to the cover gas by convection, rapidly cools and soli.difies the weld section. The weld from the long lead 4 may be made simultaneously in the same way.
The cover gas is now stopped and the cycle is complete.
After the hairpin frame has been pinch-sealed in-to a lamp envelope as shown in FIG. 1, the outer bridging portion 7 is cut off to leave leads of appropriate length emerging from the lamp envelope.
In FIG. 6 a symmetrical joint is illustrated in which the weld section, that is the extent of molybdenum penetration into the filament leg, on each side oE the lead is equal. FIG. 7 i.llustrates an asymmetrical joint which is achieved by shortening the extent of projection of the filament leg beyond the lead prior to ma~ing the weld. ~he molybdenum then readily fills the short distance 3d' to the cut end on the left because there is less heat loss from it and it is hotter~ The molybdenum also penetrates the distance 3e to the right which may be longer or shorter depending upon the volume of melted metal~ An asymmetrical weld may be used to reduce waste oE filament since it is only the portion to reduce waste o:E filament since it is only the portion of the Eilament leg continuing into the coil that is useful. In fact the lead may be weldecl into the very end of the filament leg i:E desired to reduce waste to a minimum. On the other hand, a symme-trical joint as illustrated in FIG. 6 ha~ the adva.ntage that the variability in effective filament length and the resultant volume oE weld metal due to tolerance or error in the length of the underhang is reduced in half.
The more important process parameters which must be controlled to make good joints between lead and coiled tungsten filament are the following:
l. Length of underhang.
The length of the underhang controls directly "^ !
7~L
the volume of melted metal, and the dimension Mo It must also be sufficient to permit the electrical discharge to take place to the lead portion 3b and not to the tungsten filament.
2. Heating A quantity of electrical energy is supplied sufficient to melt the volume of molybdenum chosen for the weld. The arc power, that is the rate at which heat is supplied, is preferably optimized to melt the molybdenum as rapidly as possible without causing uncontrolled spattering or excessive vaporiza-tion of material.
3. Cooling The heat input into the joint is cut off entirely when the pulse is ended and the rate of cooling is controlled primarily by the conductivity of the frame, by the heat sink provided by the components, and by the rate of cGver gas flow. The rate of cooling is maximized in order to protect the tungsten filament grain structure from undesirable crystallization and embrittlement.
4. Oxidation Oxidation of the tungsten filament or of the molybdenum is prevented by sufficient flow of inert cover gas. Addition of a few percent of hydrogen to the argon cover gas is desirable to promote quick movement of the molten molybdenum into the coiled filament.
5. Material Selection The material selected for the frame or lead wi~re which forms the weld should meet the requirements for a lead imposed by the lamp in addition to having a melting temperature appreciably below that of the coiled tungsten filament. Molybdenum is a good choice because it can be sealed directly to hard glass or quartz and is refractory enough for the lamp application. Nickel and iron may also be used.
We have utilized the foregoing criteria in tests of our invention and have achieved good process 1.4L73 ~ LD ~448 consistency. Metallographic photographs indicate no significant changes in the crystallographic struc-ture of the filament. This is confirmed by mechanical tests which have shown no deterioration in the strength of the filament as a result of the weld. ~t will be apparent that the welding process which we have described does not include any operation requiring a high level of manual skill or dexterity on the part of an operator, and our process is readi.ly adaptable to automation for lamp manufacture.
We have utilized the foregoing criteria in tests of our invention and have achieved good process 1.4L73 ~ LD ~448 consistency. Metallographic photographs indicate no significant changes in the crystallographic struc-ture of the filament. This is confirmed by mechanical tests which have shown no deterioration in the strength of the filament as a result of the weld. ~t will be apparent that the welding process which we have described does not include any operation requiring a high level of manual skill or dexterity on the part of an operator, and our process is readi.ly adaptable to automation for lamp manufacture.
Claims (9)
1. A tungsten filament to lead wire joint comprising:
a tungsten filament of recrystallized wire having a coiled portion, a lead wire of metal other than tungsten entering said coiled portion and continuing integrally into a weld section extending longitudinally within said coiled portion, said weld section comprising metal from the lead wire coalesced with the substantially filling the turns of the filament within said section, said tungsten filament being substantially undeformed and exhibiting no significant change in its crystallographic structure where it enters the weld section.
a tungsten filament of recrystallized wire having a coiled portion, a lead wire of metal other than tungsten entering said coiled portion and continuing integrally into a weld section extending longitudinally within said coiled portion, said weld section comprising metal from the lead wire coalesced with the substantially filling the turns of the filament within said section, said tungsten filament being substantially undeformed and exhibiting no significant change in its crystallographic structure where it enters the weld section.
2. A tungsten filament to lead wire joint as in claim 1 wherein said lead wire is of molybdenum.
3. A tungsten filament to lead wire joint as in claim 1 wherein said weld section extends through the coiled portion a distance no less than the diameter of the lead wire.
4. A tungsten filament to lead wire joint as in claim 1 wherein said lead wire enters the coiled portion from the side at an intermediate point.
5. The method of making a joint between a tungsten filament having a coiled portion and a metal lead wire which comprises:
juxtaposing said coiled portion to an end portion of the lead wire having a volume adequate to fill several turns of coiling of the coiled portion, suddenly melting said end portion under inert or reducing conditions whereby the molten metal wets the tungsten and is drawn by capillarity into the coiling and solidifies therein, and controlling the rate of supplying energy to said end portion and the rate of cooling the molten metal thereafter to achieve first melting and thereafter solidification in a time interval too short to allow significant change in the crystallographic structure of the tungsten where it is contacted by metal from the lead wire.
juxtaposing said coiled portion to an end portion of the lead wire having a volume adequate to fill several turns of coiling of the coiled portion, suddenly melting said end portion under inert or reducing conditions whereby the molten metal wets the tungsten and is drawn by capillarity into the coiling and solidifies therein, and controlling the rate of supplying energy to said end portion and the rate of cooling the molten metal thereafter to achieve first melting and thereafter solidification in a time interval too short to allow significant change in the crystallographic structure of the tungsten where it is contacted by metal from the lead wire.
6. The method defined in claim 5 wherein the coiled portion of the filament and the lead wire are juxtaposed in such fashion that they contact and intersect near their ends and the projecting end portion of the lead wire is melted and drawn by capillarity into said coiled portion.
7. The method defined in claim 5 wherein the rate of supplying energy to said end portion and the rate of cooling the molten metal thereafter results in a time interval of less than 2 seconds from melting to solidifica-tion.
8. The method defined in claim 5 wherein said end portion is melted by an electric discharge pulse thereto in inert or in reducing gas.
9. The method defined in claim 5 wherein said end portion is melted by an electric discharge pulse thereto from a tungsten electrode in inert or in reducing gas, and the weld arc power and time duration and the rate of cooling the molten metal are controlled to achieve a time interval of less than 750 milliseconds from melting to solidification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CA000390548A CA1181471A (en) | 1981-11-20 | 1981-11-20 | Tungsten lamp filament welded joint to lead |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CA000390548A CA1181471A (en) | 1981-11-20 | 1981-11-20 | Tungsten lamp filament welded joint to lead |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1181471A true CA1181471A (en) | 1985-01-22 |
Family
ID=4121459
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA000390548A Expired CA1181471A (en) | 1981-11-20 | 1981-11-20 | Tungsten lamp filament welded joint to lead |
Country Status (1)
| Country | Link |
|---|---|
| CA (1) | CA1181471A (en) |
-
1981
- 1981-11-20 CA CA000390548A patent/CA1181471A/en not_active Expired
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5557169A (en) | Electric lamp with high and low melting point current supply conductor | |
| US4136298A (en) | Electrode-inlead for miniature discharge lamps | |
| US4550269A (en) | Electric discharge lamps | |
| JPH1173921A (en) | Metal halide lamp with ceramic discharge tube | |
| JP2005514741A (en) | High pressure discharge lamp and method for producing electrode feedthrough of such a lamp | |
| US4703221A (en) | Electric lamp and method of making | |
| US4661739A (en) | Welded tungsten filament to lead joint | |
| US3588315A (en) | Quartz-to-metal seal | |
| JP2992765B2 (en) | Light | |
| HU222936B1 (en) | Method for manufacturing an electric light bulb | |
| US5905338A (en) | Electric lamp | |
| CA1181471A (en) | Tungsten lamp filament welded joint to lead | |
| GB2086286A (en) | Brazed tungsten filament to lead joint | |
| US4370589A (en) | Filament connector means for electric incandescent lamp | |
| US5821678A (en) | Electric incandescent lamp having an improved filament support | |
| JPH11502663A (en) | Light bulb and method of manufacturing light bulb | |
| US3530329A (en) | Filament support and heat shield construction for electric lamps | |
| US4066926A (en) | Gas-filled incandescent lamp with integral fuse assembly | |
| KR960004756B1 (en) | Welding auxiliary material | |
| JPH0589839A (en) | Welds for bulb | |
| US903860A (en) | Incandescent lamp. | |
| JP2650190B2 (en) | Manufacturing method of high power incandescent lamp | |
| JPH0711403Y2 (en) | High power incandescent bulb | |
| JP3015457B2 (en) | Cathode structure for magnetron | |
| JP3398299B2 (en) | Incandescent lamp mount and incandescent lamp using the mount |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |