[time-nuts] Helium and MEMS oscillators don;t mix well

Bob kb8tq kb8tq at n1k.org
Thu Nov 1 14:57:23 EDT 2018


A mil standard leak test works like this:

First you run a bubble leak, generally in something like hot synthetic oil. If the leak
is too big, you will not catch it with a helium test. 

Next you put the parts in a pressure vessel and pressurize it with helium (or possibly 
a gas mix). How high a pressure and how long depend a bit on just what you are testing. 

After they have soaked for a while, you de-pressurize and let the residual gas on the 
surface of the parts fly away. That does not take long at all. 

The go into a mass spectrometer and get pumped down while it looks for an outflow 
of helium. A flow above a certain level means it’s a failure.  What that level is depends
on the spec on the part and it’s size. 

Many thousands of parts get tested this way every (working) day of the year.


> On Nov 1, 2018, at 2:40 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
> kb8tq at n1k.org said:
>> Helium leak testing is a *very* common thing in the oscillator industry. I’d
>> bet it also is done in the  MEMS oscillator business as well.
> How does that test work?
> It seems obvious how to test stuff that is designed to hold Helium - put the 
> Helium in and sniff the surrounding air to see if any leaks out.  But how do I 
> test a small package to see if any leaks in?
> Maybe give it a chance to leak in for a while and then see if any leaks out?
> -- 
> These are my opinions.  I hate spam.
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.

More information about the time-nuts mailing list