Mario Scalzo’s Tech Support Blog
Recent Blog Postings
Next on the soldering to list is soldering to immersion silver (ImAg). ImAg is probably the only solder finish that is sensitive to the profile used for reflow. Usually, the reflow profile can be changed to optimize for voiding, wetting or component limitations. But, ImAg has a track record of causing “champagne” voids.
Champagne voiding is a term used to describe small voids (bubbles) that form along either the intermetallic layer or ImAg surface. These voids are known to be caused by the volatization (out gassing) of the organic co-deposit that is put down with the silver during the plating process, or the out gassing of the tarnish layer after the flux has cleaned it from the surface.
On the other hand, being a metal surface, the in-circuit testing (ICT) is facilitated by the lack of a non-conductive layer (like Organic Solderability Protectant (OSP) over copper). So this means that printing of solder paste is not needed to make the contact points able to be probed. Another advantage is that the shelf life of the bare boards is longer than that of OSP copper and immersion tin (ImSn).
But, there are some reflow profile tricks that can be used to get a good, solid solder joint without the champagne voids. Typically, shorter profile work better, as they prevent the tarnishing of the silver as well as artificially increase the slumping of the paste to increase the spread. The solder spread is usually reduced on ImAg, as well, so the slumping of the paste to increase the spread of the solder helps 2-fold.
Take it from experience; soaking the paste to remove the voiding on ImAg boards does not work. Short, fast and cool is the way to go.
Soldering to copper has been around since the beginning of soldering itself. Copper wiring and copper pipes are where soldering began, and copper is still the most common wire conductor, and soldering to it is still easy.
Copper for printed wiring boards (PWB’s) have been around just as long, dating back to the first “computers”, basically room sized calculators. But, since copper oxidizes, one way of protecting it is coating it. The coating is called Organic Solderability Preservative, or OSP. OSP copper is different from the other surface finishes because it is the only surface finish that covers the solderable surface and is eliminated during soldering, rather than consumed. And since copper is usually the base metal that we are soldering to anyway, why pay for the extra metal, such as tin, silver or nickel/gold, if just a “plastic” coating will work. This is especially true since OSP copper does not require any special reflow profiling or needs, such as a high peak temperature, or long time above liquids (TAL).
Like all surface finishes, OSP copper has some issues that we must look out for. First is the fact that since it is a non-metallic coating, any in-circuit testing must be done on a solder joint, as the test probes cannot pierce the coating to get to the copper underneath. One way around this is to apply solder paste to the test probe pads, and allow the solder to wet through the OSP.
Another potential issue is that since OSP is eliminated during the soldering process, multiple reflows, such as for 2nd side soldering or a final step of selective wave soldering, tend to break down the OSP surface and allowing the copper oxidize. Typically, the copper is pretty well oxidized once the board has been sent through the reflow oven twice, and then sent to the wave machine for selective soldering, requiring the use of a strong flux to remove the copper’s oxidation.
Soldering just plain copper has gone away. Now, there are many different surface finishes that we, as an industry, solder to. This includes Organic Solderability Preservative (OSP) copper, immersion silver (ImAg), immersion tin (ImSn) and Electroless Nickel / Immersion Gold (ENIG).
Each one of these surfaces has its own benefits and downfalls, as well as their own set of requirements to solder to them properly, with the best looking, highest tensile strength and lowest voiding solder joint possible. This list does not include the other surfaces, usually on components, that we must solder to. Such as bright tin, matte tin, Hot Air Solder Leveling (HASL), pure nickel, etc…
We will talk about OSP, ImAg, ImSn and ENIG, and discuss their strengths, weaknesses and how to solder to them. This includes discussions on relative shelf life, history of use, what to look for and special reflow needs.
Lets talk about some issues…
The first thing that I am worried about is the use of a small particle size of Indium-containing alloys. Indium is self-passivating, and will clump and cold weld to itself, even when stored as powder. For this reason we look at each individual case separately. Normally, we do not recommend the use of Indium alloys for solder powders that are smaller than Type 4 (20-38μm). For small aperture sizes, you would need a Type 5 (20-25μm). The smaller the powder size, the larger the surface area, so as the indium-containing powders get smaller, the more tendency to cold-weld in the packaging.
Which leads us to my second concern, which is the higher metal percentage in wafer pastes. Usually, in order to print through the smaller apertures (and lower area ratios) for wafer bumping, the solder paste has a higher metal percentage. For these wafer pastes, the metal percents are usually >92%. Which makes them very prone to cold-welding.
For example, the area ratio for an aperture opening of 140μm with a 90μm thick stencil is 0.39. Area ratios that are below 0.50 are not recommended.
We can physically manufacture the paste, but whether it will be useable when you get it is the problems.
Solder paste for printing follows the same guidelines as solder paste for dispensing. The good news about solder paste for printing is the apertures that are printed through are usually significantly larger than the needles used for dispensing. The BIG difference is that the paste is not as susceptible to air bubbles that would cause skips or clumping that would cause clogging.
Although stencils make a difference in the amount of paste applied, it is the paste itself that makes all the difference. Stencil release, often-called transfer efficiency or TE, can be tracked through a paste measurement system. By feeding the stencil details into the paste measurement system at onset, the system can calculate the theoretical amount of paste that should be deposited, and can create a percentage (efficiency) from measuring the amount of paste that was actually deposited.
Transfer efficiency is just now becoming something that we are tracking scientifically (read statistically). Some variables that can affect transfer efficiency are stencil type, atmospheric conditions and the paste itself.
For stencils, material makes the most difference. There are 3 types of stencils that we normally come across when visiting customers, they are Laser cut, laser cut with electro-polish, and electro-deposited (or e-fab). Also, the transfer efficiency commonly increases from laser cut, laser cut with polish and e-fab. The manufacturing cost usually increases across the three types, respectively.
Room temperature, and sometimes humidity, also affects transfer efficiency as the viscosity usually drops when solder paste is warmer, as well as the paste also becomes less tacky at warmer temperatures. Humidity affect water washable paste in the same ways, so much so that cold slump may be induced.
Most of the time, it is the paste itself, and the rosin, thickener or solvent constituents that affect the stencil release of the paste. As mentioned before, it is only in recent years that transfer efficiency is being statistically tracked to the point that the formulation may be tweaked to attain higher numbers.
More information may be found at IKB: Indium Knowledge Base.
Again, there has been a trend in the past few days that shows me that there is something happening. People are looking towards a new application or project, and realizing that they need solder paste and don't know what size powder they need. As an engineer, we have graphs and posters of data on the walls at our desks for easy reference, but i think that there should be more reasoning behind what we recommend other than just cross-referencing.
Dispensing is a good place to start, because solder paste for dispensing is more subject to the process constraints than solder paste for printing. Making a recommendation for solder paste dispensing is fairly straight forward, and mostly (yes, mostly) depends on the size needle that is to be used. Case in point, the Solar Materials Manager came to my desk and explained that a customer was having serious issues dispensing. After some discussion, we found out that the customer was using a 20-gauge (0.023" ID) needle. this itself is nothing out of the ordinary. But, after digging, we found out that they were using a Type 2 (-200+325 mesh or 45-75um) powder and a metal percent of 87%. This would explain why they were seeing clogging and "skips".
For future reference, I have attached a great picture of what we use to determine which powder size. In a nutshell, I would try to fit at least 7 spheres of powder across the ID of the needle, if the powder was all on the large side of the specification. For example, for a Type 3 mesh (24-45um), the smallest size diameter needle I would recommend is a 23-gauge (330um). This is because ~7 45um powder spheres would fit in the 330um diameter needle.
Some people would ask why not just go to the smallest size powder, then you would not have any issue dispensing through any size needle greater than a 30-gauge. This is a no-no, as there are too many drawbacks to this approach. These include the high cost of the smaller powder sizes and the higher oxide content of the smaller particles, which may cause drying of the paste in the tubes.
PART 4- Cool Down
The final element of maximizing tensile strength through a proper reflow is the cool down. Cool down is last line of defense against a poor solder joint. This is because the cool down ramp, and it alone, controls the formation of the crystalline structure of the metal lattice. The smaller, tighter and denser we can make the crystal lattice is, the higher the joint strength. Because, it is along these facets of the crystal that the joint breaks, and the longer, larger and sparser the crystal facets are, the easier they are to cleave.
One way of visually investigating whether the solder joint is tight is to look at the post-reflow surface finish of the solder joint. A joint that seems to have good wetting and good flow yet is grainy and gray may have been exposed to a slow cool down. One way to test this is to heat it up with a soldering iron. After it goes molten, remove the heat. If it becomes brighter and shinier, it probably needs a faster cool down. This may also happen if the joints around the perimeter of the board, or where the components are lightly populated, are bright and shiny and the densely populated areas have solders joints that are dull and grainy. This is because the more densely populated areas take longer to cool off, and affect the cool down rate of the board. I would reposition the thermocouples used in profiling to the denser area, and re-map the profile to meet their cooling needs.
More information may be found at Online Help: Indium Knowledge Base (IKB).
PART 3-TAL & Peak Temperature
For our purpose here, Time Above Liquidus (TAL) and Peak Temperatures both have the same affect on the solder joint. Look at it as “total heat input”, as you can have a longer TAL and lower peak, or a higher peak, and shorter TAL. As it is, together they play arguably, the most vital role of the reflow process. The name of the game is heat. Heat is responsible for solid intermetallic formation and a homogeneous solder joint, as well as proper flux deactivation.
A short TAL or low peak may result in insufficient intermetallic formation, which results in low tensile strength. It is the intermetallic that gives the joint its strength, as you always want the joint to fail during testing at either the board-side of the pad, or in the middle of the solder joint, not along the intermetallic. This is the same for the homogeneity of the joint, which is a metal solution. If the joint is not thoroughly mixed, then it is where the edges of the metal layer is where it fails, which is poor intermetallic formation. Another issue with a short TAL or low peak is not deactivating the flux. Improper flux deactivation causes a multitude of sins, including poor Surface Insulation Resistance (SIR) and continued etching of the metals.
On the flip side, a long TAL or high peak temperature may increase the dissolution of the base metallizations, and possibly increase the MP of the final joint. Too much dissolution of the base metals also forms a higher number and larger of intermetallics. Eventually, this may lead to the complete dissolving of the pad or component lead. Any time you increase the size of the intermetallic crystals, it is easier for them to fracture along said layer. A long TAL or high peak also increases joint stress, again giving another avenue for fracturing.
Ramp Rate
Ramp rate is literally the first step in the four-part reflow process and plays an important role in the formation of the intermetallics. Ramp rate, from room temperature to peak, needs to be watched for a few reasons. The ramp rate determines both the spread and volatization of the flux, and has a hand in voiding and oxidation build up.
A slow ramp tends to allow more solvent volatization, or “out gassing”. Slow ramps for solder pastes are usually 0.75-1°C per second. (For reference, a “typical” reflow profile has a ramp rate of 1-2°C per second, which generally poses a balance between spread and out gassing.) This slow ramp keeps the flux close to where it’s been applied, reducing spread and slump. This also gives enough time for the full volatization of the solvents in the flux, usually reducing voiding, as well as keeping the ΔT of the board well under 10°C. All this extra time may have a detrimental effect on some other points of interest, though, especially oxide build up of both the component and substrate metallizations, as well as the solder alloy itself.
On the flip side, a faster ramp reaches the softening temperature of the flux quicker, and therefore the flux (and paste) spread to cover a greater area, which increases the area of the joint. It may also allow for some of the activators to be saved for the actual liquidus of the alloy. Of course, there are downsides to this approach, which are the possibility of voiding (sometimes severe) and a high ΔT across the board.
At a recent customer visit, I had the opportunity to discuss “the process”. What we typically call “the process”, is that magic that happens from when the separate parts go in at the start of the line, and the finished product comes out of the reflow oven. This discussion was focused on reflow, and why it is important. Reflow is the high-wire balancing act of the SMT circus. Reflow is a balancing act because a good profile is a split between too little and too much.
Typically, we configure the reflow profile to work with the available solder and components, to give the highest tensile strength possible. So, we know what the end goal is, and we adjust what we have to achieve that goal. Besides tensile strength, some secondary goals are good wetting, solid intermetallic formation, homogeneous solder joint and a small, tight crystal structure. All of these are achieved through process management of the reflow process.
There are four parts of the reflow process that are adjusted to achieve the goals we have in mind, namely highest possible tensile strength. They are ramp rate, time above liquidus (TAL) peak temperature and cool down rate. Each one of these has its own effect on the final solder joint, and each one is important.
Want to read more? Browse the archive of past entries.
