Hacker News new | ask | show | jobs
by mikewarot 1970 days ago
I'll second that... one of the things that you might encounter is a tube that has been out of service for a long time will have quite a bit of gas leaked inside, and you might have to let the ion pump run for days to get the leakage current down to an acceptable level before turning on the rest of the circuitry. My friend actually ended up using a high voltage supply one particularly troublesome tube, but in the end we were able to get it going... it just needed a lot of ion pumping.

The bandwidth of the tube is VERY narrow (on the order of a few hertz, which means you have to be quite close on the frequency of the crystal oscillator before it can lock in.

You also want to be careful when adjusting for the Zeeman frequency to pick the right peak, or you will end up with locked to the wrong transition, and your clock will drift ,although in a very precise amount of drift ;-)

It's fun, once it works. It's easier the second or third time, of course.