24 Jul
2005
24 Jul
'05
5:40 p.m.
The fail rate can be reduced significantly by bringing
the
filament supply up slowly, over a couple of minutes, and then
applying the bias supplies for a period, the longer the better,
before the HT is applied.
It's not unusual to see lifetimes of 150,000+ hours on
transmitter valves using these techniques and that's on valves
power cycled a few times each day.
For transmitting tubes (valves), yes, careful power-up is important, but
for computer rated minis, not so much. If anything, the a slow power-up is
probably protecting the power supplies more than the tubes.
Computer rated minis fail, like most receiving tubes, due to a lack of
emmision. Weak filaments that lead to burnouts and loose elements that
lead to shorts generally get weeded out quickly, as the tubes die
prematurely.
The margin checking the /FSQ-7 did, as well as most tube computers, is an
old radioman's trick - reduce the filament by 10 precent. A strong tube's
gm will not change by much, but a weak one will take a dive in gm.
William Donzelli
aw288 at osfn.org