Uranium glass seals in a vacuum tube under UV light.
This classic DEC PDP-8 was recently donated to the Rhode Island Computer Museum by Boston Financial and Equity Corporation and came from Sonny Monosson's American Used Computer. It was powered on Saturday for the first time in probably 50 years.
I have been working on a Wavetek 1080 Sweep Generator from the early 1990s. Traced a -18VDC power supply fault to a shorted capacitor. Not the one in the middle that burned, the one on the left that is shorted. Wavetek put both of these polarized Tantalum capacitors in backwards.
I have been working on the DEC PDP-9 at the Rhode Island Computer Museum for several months fixing numerous hardware faults. We finally found the last failed component and were able to boot ADSS from DECtape, making it the only running PDP-9 on the planet.
We found a broken W640 Pulse Amplifier in the magnetic tape controller and replaced it with a spare. Still not working. We found that there is no pull-down resistor on the READ RQ I/O signal so it is always active. We added a 10k resistor to -15V and now it is passing some of the diagnostics. It failed the data-break test. That is why we started all of the diagnostics months ago.
I am working on the Rhode Island Computer Museum's PDP-9 today. A few months ago it stopped running. I have found several broken FlipChips so far and have the Processor working OK. I am chasing problems in the I/O Controller and the DECtape and Magnetic Tape controllers now. This machine usually runs for 6-9 months between failures, and each failure is just one transistor or diode. This failure is very different. Lots of broken parts and we are hunting for more.
For Thanksgiving, here's a wild turkey running across the parking lot of the Computer History Museum. I saw this turkey last year; the museum isn't a normal place to see turkeys.