| The following are examples of engineering projects I accepted while working for ETM Electromatic: Customer: System Type: Lockheed Skunk Works Tower-mounted 35 GHz Transmitter USC 50 KV Cathode-Pulsed Gyro-Klystron test set Navy, Aegis Program Simplified Driver (SDR) test set, 5 units Navy, Aegis Program Driver-Predriver test set, 15 units Navy, Fleet Microwave Coordinated Engagement Capability (CEC) test set |
 |
| ETM hired me as a technician in 1982. I did test work in QC for a month. Then Ken Lillis got me checked out dialing-in what we called cookie-cutters: 12 KV 1.5 amp beam supplies for Teledyne. Each came with a heater supply and two collector regulators. From the rear you saw three 10 KW ceramic water cooled tetrodes, on gold alodine boxes. Line them up like soldiers and it's a sight to behold. Our founder, Chuck Hayse boldly wired an op amp to a transistor to a monstrous tube. And it worked. So he did that everywhere. That's what I stumbled into. ETM has formal test procedures. The modules are standardized. It was pretty easy to just chug right along. I did get shocked on a test set for the Patriot tube. I got across some transformer terminals and did a 24 hr. stint in the corronary care unit. But generally, life at ETM was great. We got a new building and grew from 20 employees to 120. We did two runs of five 30 KV test sets for the F-16 fighter radar tube. Lots of cookie cutters went through. We did MBWO test sets. Some tubes got solenoid supplies. Our chief engineer Gunnar Wik and his confidante Al Johnson ushered in the first of a line of 100 KW, 25 KV test sets for hughes. They'd both worked for Cober, our competition. Camaraderie grew with our technical staff buildup. We all enjoyed growth through adversity. Engineers came and went. Test teck's saw every kind of engineering boondoggle possible. We learned avoidance of wrongdoing, vicariously, if not personally. Finally it only made sense that we be doing designs too. The following are excerpts from my experience in that capacity. My first full-blown engineering project was a 35 GHZ tower-mounted transmitter. |
| 35GHZ Tower Mounted Transmitter |
| The customer was secretive about what it was for. For each TTL level pulse he provided to an input jack, he expected to see a 500-Nanosecond 75-Kilowatt-peak 35-GHz-pulse: at up to 5 KHz. It had to weigh less than three hundred pounds. They provided 208 VAC, three phase power. They also provided a magnetron which they liked for the job. They drew a sketch of the enclosure. I layed out the panel so metering, protection settings, and controls were accessable behind a small access hatch in the enclosure. Skins were of perforated aluminum. |
| Line-type modulaton was presumed from the start. I went with a ceramic ITT zero-bias thyratron; featuring independantly adjustable reservoir voltage. The charging-choke/pulse forming network (PFN)/ pulse-transformer system was designed and produced by Magna Stangenes. A 10KV raw supply feeds through two paralelled 4PR-400 tetrode regulator tubes into an output capacitor which feeds the charging choke for the PFN. Thyratrons are traditionally driven by pulse transformers. I opted for direct drive using a 1KV FET. Resistor-coupled, Thyrite and zener clamped, The FET did fine. The thyratron grid takes +500V to fire. In the first instant of conduction, the grid bolts up to 10 KV for 200 nanoseconds. Then it's like a 150 V arc drop. I made a little motherboard for the thyratron, switch card, and discreet components. |
| I expected to run the thyratron reservoir control pretty low to accomodate the 5KHZ rep rate. A Sola constant voltage transformer stabilizes the process. I went too far once while reducing pressure.The anode metal was slightly blued. At 5 KHz, you have to be careful when adjusting gas pressure. Ions linger, it jams on, you back off. With dropping gas pressure, you reach a threshold where ions recede from the anode. Like a sheath of vacuum between the plasma and anode, the region looses the negative impediance characteristeric, coming to resemble a vacuum diode. Voltage drop and current through the region translate to unwanted anode dissipation. A custom heatsink was installed to make it more forgiving while being adjusted. It stabilized nicely. It did require a five-minute warmup period, which was not a problem for the customer. The magnetics worked out great. But the original mica-capacitors in the PFN had corona problems requiring replacement. |
| SN 703 |
| Arm-to-arm AC shock: Report to sick bay for mandatory 24 hr. observation period. The heart does odd stuff afterwards. Take it seriously. I felt fine--but my heart was stopping all over the place--once for 11 seconds. |
| I used a mechanical current sense relay for inverse clipping overcurrent. Adjustable comparators handled 8 other functions. Over and under indicator LEDs were below both corners of each digital panel meter. A cal door opens to reveal the 8 threshold adjust pots, with 8 switches for switching "setpoint" voltages to the panel meters. I did frequency limiting instead of having a fault. It starts dropping pulses if exceeded. A meter cal. cover unscrews revealing 8 meter cal pots. The front panel hinges down to access internals. Deitz pressure switches stand vigil in anticipation of fan failure. |
| Three adjustable threshold comparators sit poised to cut back heater voltage to the magnetron with increasing cathode current Three variacs on the panel allow user tayloring of cutback characteristic. |
| All senior people maintained CPR certification after the incident. By the time I left, every one of them had gone out on a stretcher. My boss was saved by CPR, thanks to Jerzy Spekowski. |
| Gunnar Wik had me use a 45 KV thermionic diode as a 2.3 pf vacuum-cap, for viewing divider duty. It's fine. |
| Expert: Ex=has-been, spurt=drip under pressure: a has-been drip under pressure. |
| Over/under trip levels/thresholds. |
| Heater Cutback System |
|