[time-nuts] Thoughts on Cs tube failure modes

[John Miles]

Hmm!  I appreciate the tip.  I'll confess it didn't occur to me to try to
force more current through the ion pump.  The supply I was using was limited
to about 300 uA, but I took your advice and tried a beefier one, set to 3
kV.  As I increased the current limit from 0 towards 5 mA, the ion-pump
current fell back to ~25 uA once I reached 1000 uA.

At the rate I was adjusting the current-limit control on the Glassman
supply, this occured about 10 seconds after power-up.  It seems stable now
at 25 uA, several minutes later.

I didn't leave the hot-wire ionizer energized for more than 10-15 seconds,
all told.  I'll let the ion pump run for a few hours, and then try the
hot-wire ionizer again.  Should I bother with the other supplies (Cs oven
heater, EM, mass spec), or is it reasonable to recondition the tube using
only the ion pump and hot-wire ionizer terminals?

Keep in mind that I don't have a 5062C mainframe, just the Cs tube and some
bench supplies.  I'm essentially recreating Tom's experiment from
http://www.leapsecond.com/pages/cspeak/ but without benefit of an actual
clock mainframe.  So I don't have a filament shutdown/recycle controller in
the picture.  Obviously that would be needed if I were to actually build a
clock around the tube, but for now, I just want to see if the tube is
functional at all, and determine its figure of merit.

-- john, KE5FX



> John,
>
> What you describe with the ion pump curent is normal if the tube has been
> powerd off for a long time.
>
> Th mainframe normally would turn the oven and ionizer filaments off when
> the ion pump current pegged.
>
> After minutes (can be quite a few minutes) the ion pump current will drop
> as it pumps the surge of outgassing from the heated filaments.
>
> The cycle repeats and can take a couple days on a stubborn tube.
>
> You bypassed that circuit so I'm not sure how much gas you introduced.
>
> If after a couple days the current is still pegged try connecting an
> external high voltage supply of around +3000VDC that can provide at least
> 5ma. (turn unit off)
>
> Let it run overnight and if the current comes down reconnect the internal
> power supply and turn the mainframe back on.
>
> You probably will see it peg again and the see the oven shut down (it
> happens fast!), just leave it on and it should start cycling and
> eventually the filaments will reach the oven set points and have
> outgassed enough so that the protective circuit will not trip.
>
> Then you can try to see if the rest of the tube has any life. (you may
> have to reduce the oven set resistor to 130 ohms as well as reducing its
> companion overtemp resistor. This is the value the Navy uses to get the
> last bit of life out of the tube, don't do it if you can get the rated
> beam current at the original value!)
>
> If you email me I can give you my evening phone number if you need some
> more info.
>