| The other run of units Chuck asked me to do was the Driver-Predriver test set, 15 units total. The specification was worded in the same way as the SDR spec. Almost everything scary about the SDR spec was "rubber-stamped" into the Driver-Predriver spec. Output ratings dropped to 25 KV for the beam supply and to 1.5 KV for grid-swing. With the two specs basicly the same, I rubber-stamped SDR modules into the block diagram. Chuck's ignitron crowbar stayed in. Chuck designed the original Driver Predriver tube while working at Teledyne MEC. It runs 15.5 KV, thus falling within the 17.5 KV limit put fourth by Lawrence Labs. Plenty of ETM 2062 test sets run those tubes fine, sporting ignitrons. The spec resembled a call for one of ETMs oldest lines, the 2552, but with double the average current. It was bordering upon the average power level where ignitrons invoke helplessness. In my estimation, 15.5 KV was well inside the safe operating area. There was no "one-hour no-spurrious-firings" spec. There was no "five microseconds from bias-undervoltage to crowbar" spec. |
| Driver Predriver test set |
| Jake and Tuan Tren deserve full credit for successful execution. The line was expedited essentially without me. There was an instance where one of the units wouldn't pass a foil test. Tuan was on it and believes he found the problem: an open resistor on the crowbar driver board. The failed resistor was interesting (it only provided referencing for the circuit), it was surperfluous to functionality. The rash of wrath really bore the earmarks of my nemesis. Another ignitron was playing peekaboo with me. I'd been duped, by that tired old gag, again. I should have added repetitive firing. |
| For 25 KV output, we needed a 35 KV raw supply. We specified our transformers at 10% impediance to protect our rectifiers. Gating our pass-tubes for arc-current interruption appeared tenuous, as nearly 40 KV would instantly land across the pass tube. External electrode spacing doesn't permit this. Continuous 40 KV applied to that creepage path distance would generate popcorn noise, like an open-air Geiger Counter. 20 KV spark-gap bypassing protected the pass-tubes in both lines. Jake found he could remove the gaps in the 25 KV line. The pass tubes spit a couple of times and cleaned right up. A loud report is heard at the socket with each internal arc. Splattered metal burries gas atoms on the inside walls of the tube envelope, according to popular theory. |
| Jake proved me wrong with my contention ignitrons were "too slow" for my fast bias undervoltage fault. We'd redone the crowbar-driver-board for our effort. We'd eliminated the scourge of "shutdown via ESP." And it seems we'd shaved a microsecond off firing times too. Jake dragged the test jig over and invoked bias-undervoltage upon a driver-predriver grid modulator: cathode supply discharge was completed in 4.2 microseconds. Jake is really a cartoon character trapped inside a human body. These are the times it comes in handy. Who else could bask in such buffoonery without even a hint of competitive spite? We were, in the words of my boss, "the dream team." I thought Harlem Globe-trotters was the better analogy: we accepted no difficult challengers. How could we fail? Reckless abandon was our bounty. |
| Engineering is a process by which reality is created. Engineers create a tubular fog-bank of reality around themselves while progressing. It's the art of conservation while moving full-boar forward through obstructed space in minimum time. Alas, some engineering thrusts get subverted, ending up wedged into a corner. A case in point is Harrison Labs, a division of Hewlett Packard. It all started when Brian Dickerson wired 480 VAC to the wrong end of a HP/Harrison Labs power supply. He also used too-long of screws in the terminal block on the floor. The thirteenth unit into test for the month, it was cursed. I got a spectacular display of white-hot metal balls shooting all over the cabinet interior when I hit the "on" button. The building main blew, the whole place went dark. The HP solenoid supply was blown.This threatened to block shipment of a beam supply, to Ratheon Goleta. Chuck dug in and "fixed" the errant solenoid supply. To his dismay, that HP design has been frozen since the dawn of solid-state. It still uses all germanium devices. Chuck got it going by substituting 2N2907 transistors for the duds. It was eventually rejected at the customer's facility and repaired by HP. I felt sorry for the HP technician still carrying the torch for germanium; perhaps, to never know better. It had to be "HP policy" holding Harrison Labs back, capping progress, and casting them all as dimwits. It's an odd case where HP's acquisition may have relegated a once-thriving engineering community to obsolescence. Or was it capitalism? |
| Everyone got to eat their' words on these two projects. Chuck got to eat some humble pie when all of the sudden his favorite 50-KV-rated ignitrons wouldn't hold off 25 KV for ten seconds. On-site acceptance testing ground to a halt for three Driver Predriver test sets. A couple of units in test at our plant were faring no better. I was on vacation. Those with remaining doubts about triggered-vacuum-gaps were overdue another dose of ignitron reality. Like clockwork, it hit. Then the ignitron needed more heat on the anode. It needed more cooling at the case . . . modifications went in one after the next, as the speculation went on and on. I returned from vacation to find our computer wiz-kid, Steve Warner, looking glum as a monkey in the rain. He was being sent like a sleepwalker, to institute no-longer needed changes. The rash of wrath had vanished as quickly as it had come. Not lost on anyone, an SDR unit at CPI-Varian had just finished a flawless week-long 47 KV heat-run, with tube load. This was in stark relief to the usual backdrop of people struggling with finicky ignitrons. The high voltage business is rife with poetic justice. I imagine the cosmic background radiation just flaired up. Lawrence Labs would have used two ignitrons, in series, for 25 KV. But who do you listen to? They're rated 50 KV. . . and then the vendor can't help you. |
| . . . Not checking vacuum interrupter contacts for radiation emission? I saw one peg a survey meter. Point-source field-emitters (burrs) can liberate energetic electrons capable of producing X-radiation. Why mess with lead underware. Go for the Personal dosimetry alarm badge, or the Victoreen 440 meter. |