How To Achieve Perfect Soldering
Perfect soldering is easy! It’s also very rare. What can explain this contradiction? If perfect soldering is easy, why is there so much touchup and rework?
The answer, of course, is that it’s easy if you know how. And most of what the soldering “experts” have told us is wrong. The “industry standard” soldering rulebook guarantees failures and high costs.
Surprisingly, although soldering is the core process of electronics assembly, few people do know how to solder reliably. They often do become proficient at hiding defects, but that’s a completely different and unacceptable ability. Visually acceptable connections are not necessarily reliable connections.
The very large amounts of time and money the electronics assembly industry devotes to training and certification have largely been wastes of resources. No one ever learned perfect soldering by attending “industry standard” training.
This and upcoming articles will explain why soldering reliability is so dismal, how we got here, and the necessary corrective actions.
Why We Don’t (Typically) Have Perfect Soldering
Here’s the problem: Training focuses on the desired appearance of solder rather than how the connection is achieved. And “acceptable” appearance may hide failures in waiting. How the connection was made determines not only whether the solder connection itself is reliable but whether catastrophic damage was inflicted on the component being soldered.
At soldering iron temperature, solder will stick to oxides and contaminants to produce a visually acceptable connection. However, the connection lacks an intermetallic bond and the high temperature degrades bonds inside components. The altered bonds change electrical values and shorten component life. In just a few seconds of improper application, a soldering iron can shorten the life expectancy of components by decades.
But because the connection looks acceptable and the component damage can’t be seen, the truly dismal state of modern soldering is largely unrecognized.
A Brief History of Soldering Procedures
Electronics has not always consisted of solid-state components. For decades before the arrival of devices like transistors and microprocessors, vacuum tubes represented state of the art. Electrical connections were made by soldering wires to lugs of sockets into which the tubes were inserted. Some of the wires and lugs were quite large and absorbed considerable heat. Meanwhile, soldering irons were not very efficient at turning electricity into heat. The thermal challenge in soldering, therefore, was how to prevent the solder from freezing before it completed its flow. So techniques were developed to maximize the amount of heat applied. (Protecting the tubes from heat was not necessary. The tubes were not inserted into the sockets until after soldering. They were never exposed to soldering heat.)
The arrival of solid-state components meant that, for the first time, solder was applied directly to the component rather than wires and sockets. In other words, the components were subjected to the soldering heat. And this had profound reliability consequences because heat degraded the component electrical properties.
To prevent heat damage during soldering, metal clamps were attached to leads next to the component body. Heat flowed from the soldering iron towards the component but was absorbed by the clamps before it could reach the component body. The clamps were called “heat sinks” and they provided absolute protection against heat damage.
Every work instruction since the dawn of solid-state electronics has called for the use of heat sinks. (See J-STD-001G, Sec. 4.6, for example.) But no one uses heat sinks! How can they? The leads (if there even are leads) are too small. There’s no room to attach a heat sink. But all training programs still tell students to apply heat like in 1960!
Reflow is Not Soldering
It gets worse. During the years when soldering procedures were being written, almost all component leads had tin or tin/lead plating. These surfaces melted during “soldering” and the melted solder simply flowed together with the melted surface metal. Oxides, being lighter than pure metal, floated on top of the liquid metals where they contacted flux (also lighter than metal) and were eliminated. Making connections by mixing melted metals is quite easy but it is not soldering. (The term “reflowing” was often, and properly, employed.) Soldering is the process of making the intermetallic bond with metal surfaces that do not melt. (They do not “reflow.”) And this requires additional processing steps not needed to mix molten metals. (Sadly, “reflow” continues in widespread use even though it is no longer accurate.)
The difference between soldering and reflow (simply mixing melted metals) acquired great relevance when Europe banned lead in electronics. The transition to a lead-free world focused on the new alloys. Aside from a few quirks, however, the new solders do not present tremendous problems. Lead-free solder is less forgiving of a defective process than the traditional tin/lead alloy but performs reasonably well with a properly controlled process. (Since the soldering processes in most companies have been defective, the switch to new alloys was accompanied by difficulties that were wrongly attributed to the solder rather than the process.)
The greater challenge concerns the new lead platings. Tin/lead plating disappeared, of course. But, because of tin whiskers, leads of fewer and fewer components (especially multi-leaded surface mount parts) come with tin surfaces. These new surfaces do not melt at soldering temperatures. In other words, they must be soldered. But our industry too often sticks to the limited steps that only work for reflow. And the most common training and certification simply guarantee defects and failures.
Soldering is a Simple Science – If We Let It
The reality is that soldering is science – mostly chemistry but considerable amounts of metallurgy and physics as well. The people who wrote the rule book, though, didn’t approach it that way. They operated on the basis of observation, not realizing that the critical foundations of the science are not obvious. If they got results that seemed to look right, that’s what they institutionalized. If we want a product that works and efficiencies that make profitability possible, things need to change.
Interestingly, reliability varies inversely with the amount of handling. The most reliable products are produced most efficiently. Our industry has the worst of both worlds – excessive cost and too many failures. In the next few articles, I will explain how to get things right.