You know you’re in trouble when you start thinking that shooting the designer of a troublesome system is a serious solution
By Ken Herrick, Contributing Writer
I dredge up this investigative instance from 50 and more years ago. But even in today’s solid-state age I think it pertinent still. I was a field engineer assigned to the U. S. Navy overseas and there had been a continuing problem in an airborne radar. The problem was that the radar would shut down in flight, and not recover, and the mission would have to be aborted. It would then be found by maintenance personnel, back on the ground and in the shop, that a particular component–a relay in the radar’s power supply–appeared always to be the cause of the problem.
The small relay was in the radar’s magnetron power supply resonant-charging network. Its job it was to interrupt the driving-thyratron’s circuit if the thyratron did not turn off when it should have. It was a truly cockamamie design in that the relay’s normally-closed contact was wired in series with its own coil, the whole connected into the thyratron circuit so that when excess current flowed, the relay was to “buzz,” just like a doorbell. The buzzing was intended to repeatedly interrupt the current flowing through the thyratron, very briefly but enough to allowing it to, in its own sweet time, turn off and then resume the resonant-charging cycles.
I found, first, that the problem did seem to lie within the hermetically-sealed, overload-relay itself–but only in relays of a particular brand. Those used in the radars were from several manufacturers. In the failed ones, the normally-open contacts measured always-closed on the bench, and no sound emanated when I tried to electrically actuate them. That this situation was always the case, and only appeared in the one manufacturer’s relay, was what I found significant. And when I took one of the relays and cut off its hermetic can with a hacksaw, the cause of the frozen condition became clear: The little insulating “button” on the armature, which pressed against the moving contact, was made of a thermoplastic material and it had partially melted, freezing the contact closed and the armature depressed.
I concluded that the failure sequence in those relays was as follows:
1. The circuit-current was maintained, and not interrupted as was supposed to happen, by an arc established between the opened relay contacts-maintained because the relay’s own coil was in the circuit, holding them open
2. The contact-blades heated up from the arc
3. The thermoplastic material of the actuator softened from the heat, allowing the moving contact to descend toward the closed armature, eventually re-connecting to its opposing contact
4. The contact blades cooled because the arc thus became extinguished, and the button material re-hardened; d) the relay then became frozen, with its normally-closed contact permanently closed and its armature permanently actuated; and then
5. The thyratron stayed on because its current had not been sufficiently interrupted, incapacitating the radar because the resonant-charging no longer worked.
The long-term fixes for the arcing were clear:
1. Shoot the designer–and his supervisor
2. Either wholly change the method of resetting the thyratron or at the least, add arc-suppression circuitry across the relay contacts and/or its coil-while, of course, specifying only relays with a thermo-set type of button material.
But my mandate was for the short term (and I didn’t have a gun), I filed my report to Washington, and heard nothing more about it. I only stayed in the job six months or so after that, but surely soon thereafter had to have come a change-order, directing prompt replacement of all relays of the offending manufacturer. Surely…, surely…
Contributing Writer Ken Herrick has always been in his words a techno-nerd. “I spent 18 months in the Navy, then a BSEE from U.C. Berkeley. Worked for “The Boeings” in Seattle but didn’t fancy the drudge-job–or the rain. Then as an airline station-agent in Alaska (not much rain but too cold). Then it was the Rad Lab in the Bay Area, designing stuff for the 1st A-bomb test at Eniwetok. Then I got married and we took off for our field-engineering sojourn in Japan. Then back to the U.S., but this time the long way, docking at NYC flat broke and with first-born bringing chicken pox. Took a 4 year job in RI as a project- and system-engineer, then back to CA for various engineering employments over the years. Along the way, my mother’s art-gene kicked in. Did kinetic and neon art for 25 or so years, then spent some years (retired) Tesla-coiling, and…here we presently are.”