[Amps] leaking tubes TSPA

John T. M. Lyles jtml at lanl.gov
Thu Sep 21 12:35:59 EDT 2006


Tom et al,
Just to give you an example of how tube envelopes can and do leak, 
our RCA/Burle 7835 triodes (3 MW, 200 MHz, pulsed) have a 'legal' 
warranty of ~50 hours. This 'warranty' is only valid if we keep the 
appendage ion pump on top of the tube connected to a power supply at 
least every 60 days. So in storage we have ion pumps running on all 
tubes. We have large klystrons from CPI - that we keep the ion pumps 
connected at all times, even for spares. These large tubes (bigger 
than a person can lift) will slowly degrade into worthless 
counterweights if they are kept in storage and not pumped 
continuously or at least at regular intervals. Trying to recondition 
a tube that is very gassy, without a lot of damaging arcs, is risky 
and time consuming. Functioning power supply crowbars are, of course, 
mandatory to meet the manufacturers warranties as well.

73
John
K5PRO
at a particle accelerator


>
>Myself, I'd be very reluctant to pay anything for an as-is
>power grid glass tube without knowing for sure it runs. The
>reason is the seals always leak some air, especially if the
>tubes have been used or if they are in a humid environment.
>
>Years ago I was at Dayton with a friend, and he bought a
>pair of  unused 3-400Z tubes from a guy. I bought two the
>same fellow said were just pulled from operation. The tubes
>had the same date code, they were 5 or so years old.
>
>His unused tubes were both so gassy they were unrecoverable.
>The used tubes with slight envelope discoloration were
>perfectly good. The difference was for several years the new
>tubes slowly gathered air, while the used tubes were
>operated and the zirchronium getter on the anodes did its
>job keeping them gas free.
>
>So again the fair thing to do is make it clear they are
>untested, and just take whatever you can get. How they look,
>unless something you can see is broken or melted, has little
>to do with if they actually will work.
>
>73 Tom
>


More information about the Amps mailing list