sdr20

Add your text here

Years had elapsed. ETM had delivered three of Al's CWI test sets to Hughes. The program had been a disaster. Bad luck worsened when an earthquake shook the back door of SN 335 open. This killed a shippable tube. No tube had yet run the full gambit of tests successfully. It had. Coincident to the earthquake and the door opening, the output-window broke: it was dilectric overheating due to harmonic overabundance. Oscilation of that frequency-range was symptomatic of operating with the cathode voltage way too low. A full investigation was launched. ETM sent test-technician, Jim Zeranski, to help Harvey. Jimmy was at Hughes for maybe a year. A big list of retrofits came out of it. Jimmy got promoted upon his return. I was sent to replicate listed changes, in the two remaining units. I was there all summer mopping up.

Alas, Al Johnson got laid off. I learned a lot from him.

Harv' is an electronics-enthusiast extrordinaire. Harv' Paraded me through all the test sets we could get into. Then I got down to work.

From Harvey's Riddle Archive:
Q: A winding has 19.5 turns: how?

I said: by using an E core: you can pass the lead through one window in the core an odd-number of passes; versus an even number through the other core window.

A: 40 turns paralelled with 39 turns. He'd seen it on a schematic and looked to see what the author meant.

I was instituting a lot of changes for optimization. But none aimed specifically at a root cause for catastrophic loss. We kept watching the cathode voltage, with a storage-scope through the retrofit. Harv' caught the cathode voltage jutting up, in unison with mechanical crowbar contact closure. It seems B+ grounded first so B- flew up. The surge didn't present enough energy to hurt an output window. We reconfigured the mechanical crowbar anyway, for saftey's sake. Isolated contacts now discharge B+ to B-, through respective dumping resistors. This eliminates that reason for "jutting-up" voltage phenomena.

Then I caught it red handed: cathode voltage crowbarred to zero, stayed there a millisecond, then jutted up, dumping the full raw supply cap-bank into the cathode of the tube. Follow-on current would also continue for 30 long milliseconds, until contactor interruption.
     I got the rep. for the ignitron manufacturer on the phone. At his behest, we beefed up the holding circuit. doubling capacitance for a one-second-discharge duration. Inductance in the holding anode didn't help. We settled at 20 Amps initial, using carbon comp. current limiting. Nothing helped. We cooled the case more. We heated the anode more (by changing a resistor mounted to it).
     Fit to be tied, I looked to our competition for answers. The Cober repetitive-firing crowbar-driver is electrically intimidating, with a couple of KV on the final thyratron anode. I couldn't rig up a repetitive firing circuit on the fly. Then the guy from Richardson said I needed "an ignitron with no de-ionization baffles." That sounded right, for enhanced ionization. But it showed up without a holding anode. They didn't make one with a holding anode and no baffles.
     So they founded a product-line more to my liking, "the hot setup." An NB after the type-number denotes "no baffle." But another year had passed.

I called the Hughes CWI program manager, Bill Wilkins, to tell him we finally got-in a patch for his curse. He sounded bummed. He thought Harvey had blessed the units, years earlier, when the flap blew over. Two weeks later Bill Wilkins was gone from Hughes. Originally the OEM source, Hughes never produced one tube on our equipment. They lost the contract. Everyone, even Harvey, got laid off.

Ron Carter and Ken Wonder at Crane NWSCC got SN 335. They ended up scrapping it. Ken Wonder took a photo of me ripping an anode modulator switchbox from the twisted wreckage of the cabinet at the scrap-yard. My baby (sob).

Econco bought SN 336 for $5,000. I warned now-owner Bud Butterfield about the ignitron issue. ETM retrofitted it with a grid modulator too.

Artificial Lightning Inc. got SN 586, the newest Hughes CWI unit. Being "Tesla Coil" enthusiasts, they bought it for the parts. An anode switchbox performed well upon initial inspection, as a driver for a Tesla Coil.

SN 335 did some duty as a testbed for a Hughes product called the Crossatron. The device basis is ionized gas plasma and thyratron-like conduction. Also having interruption capability, Crossatrons can be stacked for cascode switching and linear regulation. Centered around a cold cathode, magnetic cross-field makes the ionization susceptible to quenching. Handling over 1,000 Amps and 20 KV, high input impedance gives way to effortless triggering. Not symetrical or linear, turnoff takes -1,500 V, and 600 amp's peak (An SCR driven capacitive discharge would do it). I wonder how power dissipation manifests when used as a pass-element in a regulator.

Ron Carter had final say over scraping SN 335. He said it was like a trainer unit. Everything had been buggered and fixed, to the point where every connection was fatigued. There is a point of diminishing returns. Each time you open something to modify or repair it, life is shortened. There is an insidious nature to the degradation taking place. A box never opened, may never need be opened. A box opened ten times will probably need servicing incidental to rough treatment. The process spirals out of control with increasing rapidity until scrap time.