Issue link: https://iconnect007.uberflip.com/i/1453746
100 PCB007 MAGAZINE I FEBRUARY 2022 from post soldering residual material. It was feared that these conductive paths could eas- ily develop, especially in humid environments, from acid flux residues. However, we could not see rosin fluxes used in our kind of mass pro- duction. HP had actually started using acid fluxes with the introduction of eyelets, but that general process had also introduced reliability ques- tions that had not been fully answered. e big- gest issue with the use of non-rosin fluxes (usu- ally acid based) was that they required a water wash. is brought up more concerns about moisture and acid flux residue being trapped under components and amplified the issues concerning potential immediate failure due to the conductive paths, and long-term reliabil- ity problems. In addition, there was a concern about the component reliability issues due to corrosion. In HP's case, all this was further complicated by Dave Packard's decision to use gold as the etch resist. is happened when he inspected the first double-sided plated-through board and realized that the gold-plated circuitry on the top side would be exposed. He turned it over in his hands, looked at the glint of the light in the holes, and said, "It looks quality, and when you buy HP you should not only get quality but it should look quality." e only reason he was looking at a gold- plated board in the first place was because I had not yet developed a solder plating process, and just used gold as a stop gap in order to dem- onstrate the value of plated-through-hole tech- nology as compared to eyelets. In addition, I had not yet developed a nickel-plating pro- cess. As a result, the gold was plated directly over the copper. One of the unintended conse- quences of this decision was the basically hid- den total cost of the gold used. A small number multiplied by a big number can still be a very big number, and, at one time, HP bought one million dollars of gold on the Canadian spot market. In addition, due to the solubility of gold into solder, the solder pots became literal gold mines, and we had "prospectors," before we figured out how to secure them and have the solder sent to a gold reclamation company. We also had to deal with another more basic issue due to gold on plated boards. e prob- lem was that a relatively small amount of gold in the solder can raise the melting temperature significantly. In addition, gold in the solder can lead to creating more than one intermetal- lic and contribute to brittleness in the joint. In general, this is not a big issue in a large solder pot, but the solder in a small hole with heavy gold plating can have a much higher percent- age of gold. No other company took on this combination of problems in soldering. To me this was the equivalent of a war on two fronts: 1. Developing a plated-through-hole fabrication process. 2. Developing a process to solder to gold plated boards. I was doing it alone. For the soldering issue, the first thing was to find out whether anybody knew how to sol- der to gold-plated circuit boards. I sent the same letter to three solder companies asking for information on the issue. e first response was just a copy of their catalog; so much for that. e second response was a letter from the chief technologist of the company who wrote, "If anybody is stupid [he really used the word] enough to want to solder to gold they should get out of the business." He then added, "In The biggest issue with the use of non-rosin fluxes (usually acid based) was that they required a water wash.