sdr22

Patriot II

The Patriot tube got the first deluxe-treatment I saw at ETM. Ken Lillis' 48-relay panel was in the test-set (as opposed to the usual 18-relay panel). It was the pilot-program for a line of computer-interface that used stepper-motors for control actuation. I got my introduction to ETM's tube-type grid-modulator. Penta-Labs PL-264 pentodes served as beam and collector pass-tubes, and as grid-modulator switch tubes. Low ripple was achieved (200 millivolts on the 12 KV beam supply). Computer monitoring was facillitated by a Datel 12-bit hybrid A/D converter sporting multiplexed inputs. A floating 5-channel A/D converter board was linked to ground-level via fiber-optics. It monitored cathode referenced perameters. I'd been with ETM for about a month when I started testing the unit.
The grid switching section had an intermittent problem that eventually revealed a bad switch tube. It was tough to troubleshoot. We saw quirky readings at the socket of the series switch tube. Such measurements are made difficult by the reference point flying dynamically all about, to the tune of 1-KV P.P.@ 100 KHz, squarewave. Ken suspected that floating the modulator-chassis and grounding its output might simplify testing. I was in the process of stuffing a rag into the mechanical-crowbar when I contacted a live transformer secondary. I was burned badly. My arms fell like sandbags, unresponsive. But my legs still worked, so I backed away. The wound on my forearm was perfectly cauterized, glistening with solder. It was about 3/8" deep, 3" long. Burned flesh covered the transformer terminals.

I felt okay as I peered into my arm. Ken and another teck named Jim Blaha came around behind the unit. I turned my wound toward them. They freaked out and grabbed me by both arms. They walked me up to the front office. I could hear sirens in the distance. They laid me down, removed my clothes, and covered me with blankets. I still felt fine. I was somewhat annoyed by all the hubbub. Ambulance drivers strapped me into a gurney and rolled me out. En route to the hospital, they were expecting to jab me with a huge needle. I hated that idea. So they held off with the needle. At the hospital, they tried endlessly to get me to accept that needle. My answer was always NO.
I was in Stanford Hospital's Corronary Care Unit. Skin-patch EKG connected me, via strap-on transmitter, for remote monitoring. I could walk around the hospital freely, while under observation. I was fine. Finally the nurse told me I would invalidate my insurance coverage if I disobeyed doctor orders. I relented. Then she had trouble sticking me right. Finally she went in at the wrist. I felt nauseous. I started reclining the bed. Then the door flung open. A bunch of doctors rushed in. They were surprised I was awake. They asked me how I felt and I told them about the nausea. One of them said "your heart is stopped." I guess I nearly blacked out. I got a copy of the EKG, it was only 11 seconds. But over the next 12 hours, my heart did that a few more times. I nearly blacked-out each time. Each grew shorter Then the stoppages went away. I guess they know what they are doing. The burn took three months to heal. There was numbness for years afterward, just past the scar. Feeling has come back now. The scar-skin is still slightly tender. I'm fine.

So tell everyone: when someone gets an arm-to-arm AC shock, they better report to the hospital for a 24 hour observation period. I know this sounds extreme, but I guess it's right.

The Patriot unit was well received. But those PL-264 tubes were dogs. By the time I was done I had replaced every single one of them. One was bad new, got rejected, came back bad again: it was a fiasco. It took us years to get that last one back. Then a bridge rectifier blew in the bias supply for the beam-box. That fried the very transformer that had shocked me before. So I added a fuse (and I put heat-shrink tubing on the new transformer terminals). I now believe tube arcs pose a problem in the design. It needs an anode resistor. That little tube doesn't have the interelectrode spacing for 15 KV, which lands across it on crowbar firing. The poor little thing is surrounded by big capacitors. The grounded supressor-grid makes the arc-path shorter yet. The unit blasts with a sharp "crack" when it does arc. I was young yet. I'd only begun to notice flashes at sockets coincident to such reports. Such would blow out diodes in our cookie-cutter relay logic. Ken's 48-relay-panel is a nightmare that way. And the complexity-level is intimidating. Fortunately, our standard 4CW10,000 tube never arced much after initial debarnacling.

John's Patriot test-sets were the first of our units to ship with stock floating oscilliscopes, for grid viewing. The 50 KV Patriot test-sets by Cober at Hughes also came with floating oscilliscopes. Someone in that program must have specified "a round of 'scopes, on the house." I liked the feature so well I gave it to Crane in their SDR and Driver Predriver test-sets, though it wasn't specified. Margins were robust. Skimping is simply inappropriate for some systems. Such was the case. It met the Crane grid-view-jack specification: the grid-view jack was internally terminated to an internal oscilliscope. Ground-level access wasn't via the BNC they had specified: instead the grid view was HPIB-488 accessible, as were all other sub-system modules.

John might rightly feel harried by such an environment. He tried working for Varian. I saw him there and he said that it was stressfull too. He retired to Indiana. By the time the last PL-264 came back, John had moved and retired. ETM should probably send it to Richard. ETM ordered a bunch of those PL-264 tubes with water jackets. They worked out fine paralelled in a mobile CFA power amp. But cathode pulse duty is easy on tubes: the worst that normally happens is arcing drives the load like any other pulse would: no harm is done. But if load-arcing causes switch-tube gassing (due to instantaneous anode-heating), you may get switch-tube arcing with load arcing. Successfull operation of the series-resistor mitigates damage. Splattering metal may subdue residual gas. Tubes do clean up. Crowbar circuits supplement where energy diversion is desirable.