On the early satellite earth stations, the first part of the receiver system was nearly always a cryogenically cooled parametric amplifier. In order to keep the noise temperature down to around 15 K to 18 K, the actual amplifier modules were housed inside a stainless-steel Dewar (vacuum flask) and attached to the cold end of a two-stage displacer/regenerator. The displacer piston was operated by a Scotch yoke crank from a motor turning at about 50 rpm.
Every 2,000 hours the unit had to be warmed up to ambient and then dismantled to replace various bearings and seals. Then it had to be restarted under a vacuum pump until it was cold enough to condense nitrogen, after which the external vacuum pump was disconnected and the vacuum maintained cryogenically. Any subsequent tiny leaks of air immediately condensed as frost on the cold part, thus preserving the vacuum.
Most times the elaborate servicing procedure went to plan. The paramp would pass its post-service performance test and be declared fit for use, after which we would switch paramps and service the other unit. It was always a nervous time during servicing, with worry that the in-circuit paramp might fail, thus interrupting traffic. Despite the cryogenic pumping action, eventually the frost could build up so much that a tiny flake would drop off, land on a warm part, and evaporate, thus breaking the vacuum and causing an unexpected warm-up of the whole thing.
On one occasion the planned servicing session went from bad to worse. When cooled down, the 500-MHz-wide pass-band was all over the shop and not fit for traffic. After several cooling and warming cycles the problem hadn't been solved. As each procedure used up another set of seals and bearings, it was getting out of hand. Eventually the news reached the deputy station manager and he decided to roll up his sleeves and get involved with the paramp crew. That was Brad, and let's just say that sometimes his enthusiasm outpaced his judgment.
For the fourth time, we opened up the business end of the paramp and inspected individual modules carefully. We found nothing obvious. Brad reasoned aloud that the uneven pass-band could be caused by a problem in the connections between the three paramp stages and the one final tunnel diode stage. They were connected by a 3.5-mm-diameter semi-rigid coaxial cable and tiny gold-plated screw-on connectors. Maybe there was a mismatch in one of the connectors and that was causing standing waves?
Without consulting the manual, Brad removed each short section of cable in turn and inspected each one minutely. Finding nothing obvious, he reached for a multimeter, set it to ohms range, and checked the links for continuity. With a yelp of glee he said, "That's it -- this one is open-circuit!" Sure enough, one of the identical-looking links was indeed open-circuit. Milking the situation for all it was worth, Brad gently chided his engineering staff for not checking for something as obvious as an open circuit. Most of the paramp team looked duly impressed. "Good work, Brad."