|
The solder I used was the
63./37 tin-lead eutectic. Experimental seals were made using solder pot
dipping, radiant heating, and RF heating. They all required flux which
could contaminate the device. Ultra-sonic dipping was tried but required
specially optimized equipment. About that time, Joel Ollendorf came up
with the idea of sealing with an interference fit between the solder coated
cases and bases (aka stems): no heat and no flux. I came up with ways to
properly coat the mating parts and modified a pair of pliers to jam them
together. The seals passed leak tests and devices performed well on life
test. With some evolutionary improvements, the process became standard for
production devices. Joel and I were awarded U.S.Patent #2,965,962, which we
all referred to as the cold solder seal. Fred Hunter and other engineers,
after exploring many possible coatings, came up with a successful junction
passivation coating using a Dow Corning silicone resin, SR98. To increase
power dissipation capability to meet customer needs, other engineers came
up with a successful potting compound. Aluminum oxide powder was mixed into
silicone fluid (1:1) and injected into the inverted case before sealing.
Dr. Li of the Tube Division statistically designed experiments that
minimized sample sizes for these etch, rinse, bake, coat, and potting
compound optimization tests. Next, a short project with Aaron Frank and
Gerry Sullivan, resistance weld sealing transistors for a Signal Corps
Preparedness contract. The Tube division was in high volume manufacture of
resistance welded metal can tubes which were about 1" in diameter. It
took some modifications and help from our machine shop, from Jim Bibby and
Paul Del Priore, but we finally got decent seals with our small relatively
delicate packages.
|