After I had been 'in country' a while I was lent to Marble Mountain to help resolve a recurrent problem with their radar. The Short Take-Off and Landing (STOL) Marine airbase at Marble Mountain had a GCA unit also. The base was used by helicopters and smaller planes, although on one occasion a commercial troop transport pilot mistook Marble Mountain for nearby DaNang and landed an airliner on the STOL runway, which was on a berm about six feet above the surrounding landscape. It was quite a feat and the story of how they got it out of there is now a Youtube video.
I went to the control tower at Phu Bai to hitch a ride to Marble Mountain. Since my MATCU colleagues controlled take-offs and landings, when a pilot requested permission to take-off, the tower operator asked for his destination and if it was Marble Mountain the tower asked if he could take a passenger. One said 'sure' so I ran out onto the runway and got into the aircraft and off we went.
I arrived at the Marble Mountain MATCU in the afternoon and was introduced to everybody just about quitting time. So after dinner I went with my new acquaintances to the Non-Commissioned Officers (NCO) club. Toward the end of the evening I offered to buy the last round of drinks but somebody else demanded they buy. So I insisted that I would buy a round the next night.
In the morning I reported to work on the GCA system. They were having a problem with their radar crystals frequently burning out. A radar works by sending a very high powered pulse of electrical radiation out the antenna into the surrounding air. The curvature of the antenna focuses that radiation into a flat paddle-shaped lobe. When the radiation strikes some reflective object a tiny amount of the energy bounces back toward the radar. That tiny amount of energy has to be collected by the same antenna and sent to a sensitive receiver that is tuned to the precise frequency of the pulse.
The problem is that the high powered pulse going out the antenna would overwhelm the sensitive receiver. So radars have two small crystals, charged to a high voltage, one on each side of the entrance to the receiver. When the outgoing blast of energy is sent out, its energy causes an arc to short circuit the gap across the crystals and that arc protects the receiver. What was happening in this radar was that the crystals kept burning out and thus the receiver didn't detect the returning radar pulses. Their technicians didn't know why the crystals continued to fail.
I immediately began checking the circuitry associated with supplying the high voltage to the crystals and soon noted that a voltage in a nearby circuit was incorrect. I traced this voltage anomaly through the circuits and determined that the voltage was anomalously low in a distant and seemingly unrelated circuit inside the radar. I opened a metal box that covered the circuitry involved and spotted some indications of burning. I quickly realized that a resister had been overheating and warping.
Normally a resister in this overheated condition would simply crack and break the circuit. But I suspected that this overheated resister warped until it pressed against the cold metal box which kept the resister from becoming hot enough to crack, thus the burn marks. However, it changed the electrical properties of the resister enough to lower the circuit voltage, which dragged down the voltage to the crystals so that the outgoing pulse didn't create a protective arc.
Because the circuitry would function normally when the radar was first turned on, it was difficult to detect. It was only after the circuit had time to heat up and the resister warped against the metal box that the voltage across the crystals decreased and didn't arc. It was the kind of intermittent electrical problem that was difficult to figure out. I only found it because I knew the crystal circuitry well enough to know where to look for the anomaly.
Thus I had resolved the problem and repaired the radar in time for lunch. When we got back to the unit after lunch the officer in charge told me to contact my unit in Phu Bai. I called them and was told to return to my unit as soon as the problem was repaired. I asked if it was okay to return the next morning, hoping to go to the NCO club with my new friends that night. I was told no, return that afternoon.
That upset me because I had promised to buy a round of drinks to pay my share of the entertainment, plus I wanted to celebrate having found and repaired the problem on the radar. But I obeyed orders and hitched a ride back to Phu Bai that afternoon. I arrived to find there was a minor problem that I easily took care of. But my new friends back in Marble Mountain would go to the club that night probably thinking I had stiffed them for the drinks.
I found out the next day that this was the least of their concerns. The NCO club I was supposed to be at that night buying drinks had suffered a rocket attack. Multiple rockets had slammed into the area around the club and many patrons were injured. I was unable to learn whether any of my new acquaintances were among the injured. But it seemed likely that my being ordered back to Phu Bai may have been a lifesaving decision.