Background of the Invention and Prior Art Statement
Laser soldering has been carried out to date in what is essentially an early stage, a stage directed to proving feasibility. It has demonstrated utility over mechanical soldering techniques because it can reach inaccessible soldering assemblies, such as those recessed in a housing. These early attempts have been applied to the assembly of small electronic components mounted on conventional plastic (epoxy fiberglass impregnated) boards, which prevented heat conduction to the rest of the assembly. Very small diameter wires of conventional construction were soldered using relatively small solder pads dabbed onto easily accessible conductors of a circuit. These early feasibility attempts used very small powered laser beams (7-50 watts) and focused the beams on the solder pad to microsized diameters (0.005-0.010 inch) in order to effect melting. These early feasibility attempts are referenced in the following:
(1) F. Burns and C. Zyetz, "Laser Micro Soldering", Apollo Lasers, Inc. Report.
(2) E. R. Goodrich, "Lasers in Electronics", Circuits Manufacturing, Vol. 21, No. 7, July 1981.
(3) R. Saunders et al, "Lasers-Operation, Equipment, Application, and Design", prepared by Engineering Staff of Coherent Inc., McGraw-Hill, 1980.
(4) T. Kujawa, "Laser Soldering Boosts Productivity", Lasers and Applications, September 1982.
Unfortunately, such early feasibility activities used processing techniques which would not enable an assembler to solder assemblies with the new emerging design constraints of automotive applications. For example, assemblies of sensitive electronic components may now employ larger sized electrical lead conductors (greater than 0.030 inch in width and in many cases flat or rectangular in cross section) which must be soldered with a higher strength to withstand varying thermal expansion forces experienced in the engine compartment of a road vehicle. The prior usage of low powered, neatly focused laser beams provided several obstacles to effective soldering of such emerging design assemblies. The laser beam, so controlled to a microsized diameter, would not melt a sufficient amount of the solder pad to effect a reflowed fillet along the entire bottom surface of a large, flat electrical lead wire, resulting in a weak joint. Reflowing solder about a small, round wire is considerably easier due to surface tension about a round surface. If the interfacing diameter of the laser beam were to be modified or expanded, the beam energy/unit surface area would not be sufficient to melt the solder and effect the reflowed joint.
In some of the emerging automotive electronic units to be soldered (such as an electronic ignition module mounted in an engine compartment) the solderable joints are supported on ceramic substrates used to conduct heat away during normal vehicle operation of the module, thereby preventing destruction of the printed circuit and soldered joints due to momentary high currents. However, such ceramic substrates act as a heat sink and facilitate rapid run-off of heat during soldering which inhibits proper thermal control of the laser diameter, and certainly makes it impossible to use low powered laser beams if proper soldering is to be effected. Increasing the laser beam power by itself is not an answer to this dilemma because a high powered beam, focused at a desired diameter, will heat the solder excessively, leaving little room for processing error and resulting in damage to adjacent sensitive electronic parts.
What is needed is a method of laser soldering which permits recessed soldering assemblies, supported on ceramic substrates, and employing larger sized conductors, to be effectively soldered with increasingly stronger joints and soldered within a wide range of processing parameters making the process easily usable by manufacturing personnel without tight limits.
Summary of the Invention
The invention is a method of soldering one or more electrical lead strands to a printed electrical path. The method comprises the steps of: (a) planting the path on a ceramic substrate and attaching at least one solder pad to a portion of the printed electrical path; (b) forcing a surface portion of the electrical lead strand into full interengagement with the pad with an engaging force sufficient to permit heat conduction therebetween for soldering, the pad having a width greater than the width of the surface portion of the electrical lead strand, and the interengaging surface portion of the electrical lead strand having a width greater than 0.03 inch, the portion of the electrical lead strand and pad constituting a soldering assembly; and (c) directing a defocused laser beam onto the soldering assembly, the beam having a beam power of at least 100 watts, a beam spot diameter no less than the width of the electrical lead strand and no greater than the width of the pad, and a beam on-time effective to exert a controlled thermal radius on the soldering assembly to reflow only a preselected portion of the pad and effect a solder joint between the pad and the strand portion, the joint having a strength of at least 400 grams.
It is preferable that the electrical lead strand have a flat interengaging surface and the pad have a mating, generally flat interengaging surface (allowing for a slight radius due to surface tension). It is advantageous if the laser beam is defocused by an amount of 0.4 inch (with a five inch focal length lens) and have a focal point located above the soldering assembly to thereby protect against reflection of the laser beam to other sensitive electronic components. It is also desirable that the center of mass of the interengaging surface of the electrical lead strand be aligned with the center of the pad against which it is forced.
It is advantageous if the thermal radius is increased by at least one of the following: (a) a coating of energy absorbing paint, or (b) by increasing the temperature of the soldering assembly by preheating to a temperature of at least 100.degree. C. It is further advantageous if the laser beam spot diameter is controlled within the dimensional limits of 0.03-0.09 inch and that the beam on-time be limited to 50-2000 milliseconds (0.05-2.0 seconds). The beam parameters can be optimized by use of the following mathematical relationship: ##EQU2## where: c is the critical thermal radius (melt radius)
a is the Gaussian radius at 1/e.sup.2 point
ln is logarithm
T.sub.m is the melting temperature of the solder minus the specimen preheat temperature
P is the laser beam power in watts
A is the surface absorptivity of the solder at 10.6 microns
R is the thermal resistance per unit area of the system
t.sub.c is the critical time to bring the solder to the T.sub.m temperature (beam on-time is approximately t.sub.c /0.975)
C is the heat capacity of the system.
It is advantageous if a shielding gas is used during soldering to prevent the flux from burning for easy removal of the flux after soldering.
It is preferable that the ceramic, supporting the electrical path, be selected as one or both from the group consisting of alumina and beryllia, and advantageously have a film thickness of at least 0.02 inch.
It is advantageous if the electrical lead strands extend from an annular integral lead frame which is encapsulated within an annular nonconductive housing. The leads may be preferably comprised of a tin coated brass material, the pad comprised of a tin/lead/silver or tin/lead alloy, and the printed circuit be comprised of palladium/silver material. A flux is typically employed between the pad and strand and is preferably comprised of nonactivated rosin. It is also preferable that the laser beam be a CO.sub.2 generated laser.
Summary of the Drawings
FIG. 1 is a photograph illustrating a microelectronic device having several joints soldered by the method of this invention.
FIG. 2 is an enlarged photograph illustrating some of the electrical leads of FIG. 1 successfully soldered in accordance with the present invention.
FIG. 3 is a schematic sectional elevational view of a portion of the device of FIG. 1 showing one soldered joint and an insert (FIG. 3a) illustrating a plan view of the soldering joint.
FIG. 4 is a graphical illustration depicting the variation of beam on-time with beam power at various combinations (the cross hatched portion showing the acceptable manufacturing combinations that are useful).
FIG. 5 is a heat flow diagram between the various materials forming the soldering assembly of FIG. 2.
FIG. 6 is a schematic diagram illustrating a Gaussian energy beam distribution across the solder pad.
FIG. 7 is a graphical illustration of critical thermal radius varying as a function of beam on-time at various power levels.
FIG. 8 is a graphical illustration of beam power as a function of soldering assembly temperature at various preheat temperatures T.sub.o.
Detailed Description
As a preferred mode, the invention is applied to the soldering of integral electrical leads to a thick film ignition module useful in controlling the starting of an internal combustion engine. The device to be soldered is more particularly shown in FIGS. 1 and 3, and more completely disclosed in copending U.S. Pat. No. 4,470,648, "Interconnection Construction to Thick Flow Substrate", invented by D. Davis and C. Henritzy, assigned to the assignee of this invention.
Essentially, the microelectronic device 10 comprises an annular lead frame 11 which is encapsulated or buried within a plastic annular housing 12 with integral lead strands 13 projecting inwardly through the sides of the housing and depending downwardly therefrom. This device here has nine electrical lead strands projecting inwardly that need to be soldered to a printed electrical circuit 14 supported on a ceramic substrate 15 and recessed below the top opening of the housing. One of the substrate materials is alumina and occupies 75% of the total substrate; the other is beryllia. The ceramics are glued to a supporting aluminum plate 16 which acts as a heat sink.
Each of the electrical lead strands is comprised of a brass conductive strip nominally comprised of 70% copper and 30% zinc. The projecting portion of the lead strand is given an inverted U-shape 17 with the terminal portion defining a foot 18 to be soldered to a portion 14a of the printed circuit 14. The foot of the electrical strand has a flat cross sectional configuration here sized to be 0.04 inch wide and 0.009 inch thick. Numerous adjacent sensitive electronic components are present such as resistors 19, capacitors 20, and electronic chip 21.
A ceramic substrate is required in this device because of the necessity for maintaining a heat sink during operation of the ignition module for engine starting. Ceramic has a higher conductivity than a plastic (epoxy fiberglass-impregnated) substrate, typically used by the prior art. The alumina is nominally comprised of 94% Al.sub.2 O.sub.3 and used in a thickness about 0.02 inch (0.56 millimeters); the beryllia substrate is nominally comprised of 99.5% minimum BeO and in a thickness of about 0.02 inch. The aluminum supporting plate 16 is comprised of AA3003H14 aluminum alloy and used in a thickness of about 0.06 inch.
The electrical lead strands are of a size much greater than that normally employed by the prior art, having a width 22 greater than 0.03 inch, and have a flat, rectangular cross section (i.e., 0.04 inch.times.0.009 inch). Such electrical lead strands require different processing technology to ensure that a completely solid solder joint takes place throughout the entire interface (under-surface) of the electrical lead foot as well as around the edges thereof.
The method herein comprises essentially three steps:
1. A printed electrical circuit path 14 is planted on a ceramic substrate 15 and at least one solder pad 24 is attached to a portion 14a of the printed electrical path. The solder pad may be deposited by conventional techniques which normally include silk screening. The pad here is a soldering composition comprised of 10% tin, 88% lead, and 2% silver. The pad is formed to have a generally flat interfacing surface 25 (except for curvature due to surface tension) with a width greater than the width of the electrical lead. Here, each of the pads have a rectangular configuration of 0.08 inch wide, 0.1 inch long, and 0.006 inch high.
2. The surface portion 23 of the electrical lead is forced into full interengagement with the soldering pad (against surface 25) with an engaging force of 50-150 grams. It is important that the pad 24 have a width 26 greater than the width 72 of the electrical strand and the interengaging surface portion of the electrical lead have a width 22 greater than 0.03 inch. To bring about the imposition of force by the electrical lead strand on the pad, the electrical lead strand is bent with a radius or U-shape as shown in FIGS. 2 and 3; this allows the strand to impart a spring bias against the top surface 25 of the pad. The end of the pad has a curved foot 18 portion to permit mating of the interengaging flat surfaces 23-25.
3. A defocused laser beam is directed onto the soldering assembly, the beam having a beam power of at least 100 watts, a beam spot diameter 27 which is no less than the width 22 of the electrical strand and no greater than the pad width 26, as well as a beam on-time effective to exert a controlled thermal radius on the soldering assembly to reflow only a preselected portion of the soldering pad and effect a soldered joint between the pad and the electrical strand portion, the joint having a strength of at least 400 grams.
The laser beam 28 employed here was generated by a 370 watt CO.sub.2 laser apparatus and was focused at a point 29 a distance 30 (0.4 inch) above the soldering assembly. The beam impinged downwardly on the foot 18 and solder pad 24 assembly. The central axis of the laser beam was aligned with the center of mass 31 of the electrical lead at its engaging portion and also aligned with the center 32 of the soldering pad. A small amount of soldering flux was applied to the area between the soldering surfaces prior to soldering, such flux being typically nonactivated rosin microflux.
The soldering assembly was preheated to 150.degree. C. to lower the power requirement and to eliminate the possibility of thermal shock. A coaxial nitrogen gas stream was used to (a) shield the solder area from oxidation and thereby protect the soldering joint from a black coating deposition, and (b) to protect the focusing lens from being fogged by smoke. The beam itself, as it emanated from the generating device, had a diameter of about 0.6 inch, and was defocused slightly above the soldering assembly to have a spot diameter on the soldering assembly of 0.030-0.080 inch. The best results were obtained when the spot diameter was 0.050-0.060 inch.
Various soldering operations were undertaken with a variation of beam power and time to see whether satisfactory soldered joints of sufficient strength could be effected. It was found that soldered joints with a strength of 2000 grams or greater could be effected with a power variation of at least 100 watts and up to 500 watts, a beam on-time of 50-500 milliseconds (0.005-0.25 seconds). These combinations of beam power and beam on-time that worked using large sized electrical leads and soldering pads are plotted in FIG. 4 (spot diameter in this case is 0.060 inch). The shaded area of FIG. 4 represents the manufacturing window wherein tolerable combinations of parameters can be employed to render a satisfactory fillet 33 around the entire joint (see FIG. 2) without permitting the entire solder pad to melt nor the substrate to melt.
Room temperature specimens which were not preheated required a longer beam on-time or a higher power level than preheated specimens. A room temperature specimen required 50-75% more power than a specimen preheated to 150.degree. C. for the same beam on-time. Accordingly, it is preferred that a preheat temperature of the soldering assembly be employed which is at least 100.degree. C. At this temperature, the power may be reduced by 25% compared to the room temperature (20.degree. C.) specimen (see FIG. 8).
The energy requirement or power level of the laser beam can be decreased by increasing the surface absorptivity of the soldering lead. The power level can be reduced, for example, from 400 watts to 200 watts, by applying a black coating on the lead foot or by eliminating the nitrogen gas shielding.
The effect of the beam spot on beam on-time is such that the minimum time for satisfactory fillet formation is progressively decreased as the beam spot diameter is increased. The widest time range for satisfactory joints was obtained when the beam spot diameter was about 0.05 inch (1.27 millimeters). An excessively large spot diameter (such as 0.085 inch) coupled with a long beam on-time (such as 0.3 seconds) sometimes caused a beam reflection problem; such reflection causes heat damage to the adjacent components such as resistors or capacitors. The reflection problem, can be reduced by use of a proper defocused beam (proper spot diameter) with the focal point above the soldering assembly. The shielding gas flow increases the power requirement; but without a shielding gas, a black residue is left by the burned flux. Since the black residue cannot be removed easily, it is recommended that the shielding gas be employed. A gentle blow of the nitrogen gas is adequate to prevent the flux from burning or charring.
Critical Thermal Radius
The beam spot diameter has been used as a common measure in laser thermal processing to define the beam size on the work piece. However, this diameter by itself cannot be used in predicting the amount of thermal work accomplished. For example, in the drilling of holes with laser beams in various materials, such as glass, the drilled hole is consistently different from the beam spot diameter on the glass material.
The critical thermal radius of this invention is the radius of an area on the work piece within which the desired thermal effect is achieved. It represents the radius of the melt area in laser soldering.
To obtain a mathematical relationship that will provide an accurate critical thermal radius, an analysis of input heat distribution is necessary. A simple thermal model of one-dimensional heating and a Gaussian beam profile were assumed in the analysis. When a laser beam impinges on a metallic absorbing surface, the laser light is absorbed by interaction with electrons. A quantum of optical energy is absorbed by an electron which in turn dissipates its energy by colliding with lattice phonons and other electrons. Since the mean free time between the collision of electrons is on the order of 10.sup.-12 to 10.sup.-13 seconds, one may consider that the optical energy is almost instantaneously turned into heat. The penetration depth of the light field in a good conductor, such as used in the preferred mode of this invention, lies in the range of 10.sup.-6 centimeters or 10-100 atomic layers. Thus, laser absorption is considered to occur at the surface in cases of laser soldering.
A few other additional assumptions were adopted to simplify the heat flow analysis of the system: (a) the thermal properties of the materials and the thermal resistance were assumed not to change with temperature, (b) the convection and the radiation losses were considered negligible, (c) the heat flow was considered one-dimensional, (d) the solder pad was at a uniform temperature throughout the thickness, and (e) the aluminum base was considered the heat sink. With these assumptions, and as shown in FIG. 5, the heat flow system can be proportioned with heat flow into two components: accummulated heat and conducted heat. The heat input q is partially accummulated in the solder pad (q.sub.s) and the alumina substrate (q.sub.A), and the rest of the heat is conducted away to the heat sink q.sub.3 (aluminum plate). As shown in FIG. 5, T.sub.2 is the average temperature of the alumina substrate through the thickness and, if one assumes a parabolic temperature profile within the alumina substrate, the average temperature of the substrate T.sub.2 is given by the relationship T.sub.2 is equal to T.sub.1 /3. The energy balance of the heat flow system in FIG. 6 yields a first order differential equation: ##EQU3## where: C.sub.1 =heat capacity of the solder pad
C.sub.2 =heat capacity of the ceramic substrate.
From this we can solve for T.sub.1 and, if we define a further quantity called q.sub.c as the rate of heat flow per unit area to bring the solder temperature from T.sub.1 to T.sub.m in T.sub.c seconds, we have the expression: ##EQU4##
After the solder reaches T.sub.m, additional energy is needed to melt the solder, and this additional energy can be supplied by an extended beam on-time approximately 3% longer than the time t.sub.c shown above.
The radius of a Gaussian beam is typically defined as the radius at which the intensity falls to 1/e.sup.2 of its peaked value (86.5% of the total beam energy is contained within the area of this radius). The intensity profile of a Gaussian beam is illustrated in FIG. 6. When a Gaussian beam melts a spot of radius c on the solder pad, as shown in FIG. 6, the beam intensity at the edge of the melt spot is given by the following equation: ##EQU5##
This radius c is selected as the critical thermal radius because it defines the area where the desired thermal effect is achieved. Employing previous equations, the c radius (critical thermal radius) can now be reduced to the following mathematical relationship: ##EQU6## where: c is the critical thermal radius
a is the Gaussian radius at 1/e.sup.2 point
ln is logarithm
T.sub.m is the melting temperature of the solder minus the specimen temperature
P is the laser beam power
A is the surface absorptivity of the solder at 10.6 microns
R is thermal resistance per unit area of the system
t.sub.c is the critical time to bring solder to T.sub.m
C is the heat capacity of the system.
The minimum power to melt the solder at a given time (T.sub.1) can be obtained by setting c=0 in the previous equation.
FIG. 7 shows the theoretical c radius as a function of beam on-time when the beam spot diameter is 0.060 inch. The plot was generated by substituting proper values of the parameters and material properties into the critical thermal radius equation.