[time-nuts] Helium leaks

[jimlux]

There was discussion last October about the very tiny MEMS oscillators 
failing from He leaking in.  By now, there's several videos online 
showing folks taking that MEMS oscillator and putting it in a chamber 
and demonstrating the failure.

So here's something interesting - There are a remarkable number of 
places with high atmospheric helium concentrations. And it causes a 
variety of mystery failures.

The one of interest here is that helium can leak into a physics package 
with Rb or Cs, and that changes a) the total pressure (which causes line 
broadening, apparently) and b) the thermal conductivity of the vacuum.

The latter is of some concern to me, since I have some CSAC devices in 
space that may have been exposed to a helium rich environment during 
storage. The failure mode on a CSAC is that more heater power is needed 
to keep the vapor pressure up, due to increased heat loss through the 
"not as good a vacuum" in the physics package.

Has anyone on the list ever had a CSAC in a He-rich environment? Noticed 
any performance differences?

I've asked Microsemi - their first response was that it's not something 
they tested for and that the leak rate is very low.  I suppose it's 
probably in a hermetic can (which are *tested* with Helium leak tests)

Jim