The -- hp -- 428b verification procedure is steadily and slowly advancing. Built a little voltage divider box for the cal from a plastic mini Hammond box, 3 Chinesium binding posts and 2 hand-selected Vishay 1K 1% resistors that the -- hp -- 974A declared were 0,1 Ω from each other. (428b manual says one should select these carefully for similar values.) The -- hp --427a (hi, bd139!) said there was 22mV on the centre of the divider when used as directed in the oscillator circuit, which is in-spec ("under 50mV"). Oscillator frequency is now 40KHz, not 39980Hz (which was also in-spec, though), according to the -- hp -- 5221A (yes, Nixies at work here. Wonderful.), and the oscillator level is 8,02V, according to the Marconi VTVM, which is in-spec. My ScopeMeter was very finicky and did not want to auto-trig on the waveform; it took almost CRO amounts of fettling the controls to get it to display anything sensible.
All vintage instruments have had a modern sanity check made in parallel, because even if the person with two clocks always wonders which one is right, the observation that they roughly agree the next day too is somewhat reassuring. The -- hp -- 974a really is an excellent DMM, steadily getting a good measurement off of the DUT.
All in all, so far every measurement has been within tolerance. This instrument must have been put aside at a very young age, probably because it blew fuses, and then just been a Sleeping Beauty waiting for someone to take it to bits and figure out what was wrong with it.
This is bordering on obsession, but I've got my reasons, and I like to think they're on track for a group of T.E.A. afficionados...