[R-390] Gettering old tubes

[2002tii]

Jerry wrote:

>He plugs in 10 at a time and he "cooks" them for about a week at rated
>filament voltage and reduced plate voltage.  These tubes span from the 1940-
>1970's.  His Sencor tester, which is supposedly very sensitive to gas, has
>found fewer than 1 in 100 to have shorts or gas.
>
>Does this really improve the tubes??  YMMV

No, if by "improve" you mean react accumulated gas with the remaining 
getter flashing.  It does give one an opportunity to spot gross 
failures that occur shortly after commissioning, or obvious parameter 
outliers.  I built a system something like this that also monitored 
noise when a client was building high-performance tube equipment in 
the '80s.  For most tube types, they ended up discarding from 40 to 
85% of the tubes they received from suppliers during initial QC 
checks.  We tried ALL of the suppliers, and bought up lots of NOS 
tubes.  It was such a problem that I even visited a number of the 
manufacturers to see if we could persuade or help them make better 
products, and drafted a business plan to build a tube factory (we 
concluded that it would not be profitable to produce a quality 
product).  Ultimately, they dropped the tube equipment lines because 
it was so hard to find good tubes.  As far as I can tell from the 
tubes I encounter today, tube suppliers' claims have gotten wilder 
and many old brand names are once again in use (no connection with 
the original manufacturers, though -- in fact, most of the names are 
owned by distributors, not manufacturers), but the tubes themselves 
have not improved since the '80s.

No commercial tube tester I have ever seen can determine anything 
useful about the gas in a tube except for gross failures that can 
usually be spotted visually by fogging of the envelope.  You can get 
some idea by doing a noise analysis, particularly if you 
simultaneously sweep the tube operating point.

Best regards,

Don


Copyright (c) 2011.  Not for redistribution.