I can tell you that if you go through a pile of output tubes with the idea that you will select out pairs that pass currents within "X" ma (where X is a pretty small number, 1-2-3-4 ma) of each other, and THEN test them for Gm (and insist they pass) within say 10%, your "matched pair" yield will fall by more than 50%.
When I was a kid in school in New Jersey, we went on a field trip to the RCA tube plant in Harrison, NJ. I was well into building tube stuff back then and it was pretty danged interesting, a HUGE factory from the 1920's or before, although it was also clear that it would get unbelievably boring working there every day. The tubes (only all-glass 7/9 pins and perhaps compactrons were being made, no octals that I could see) were made on these 7-10 foot across carousels. The raw materials were the bases, just a glass disk with 7-9-12 pins stuck thru with the glass-to-metal seals made, and the upper part of the envelope which looked kind like a slender and straight sided wine flute, with a 4-5" tube attached, obviously to hold and evacuate the tube.
The carousels advanced kachunk-kachunk maybe 15 seconds dwell at each station, with 30-40 stations around their circumference, and at every station, a wire or two or three got bent, a pre-fabbed grid on two support wires got jammed into the base mica, a sheet of plate-metal was bent into a cage....and on and on it went. It was mesmerizing. At the end, a torch sealed up the base, heated and stretched out the tube, and a coil flashed the getter. The machines were marvels of dozens if not hundreds of actuators, grippers, little arms moving in and out, and it must have all been programmed with some kind of chain, inside the thing.
Topic: Tube Current Draw (Read 11856 times)