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

[jimlux]

On 11/1/18 8:47 AM, Bob kb8tq wrote:
> Hi
> 
> A lot of Radon and *really* poor ventilation….
> 
> There are a lot of ways for He to show up. In normal use, issue is hanging on to it.
> It tends to run away from its source very quickly. Maintaining a measurable concentration
> in something like a normal room …. not very easy at all.
> 
> Bob
> 


a couple interesting things to think about (I personally think the 
original story has some other confounding factor they forgot):

1) MEMS pressure sensors have been around for decades, and they're used 
with helium all the time.
2) MEMS accelerometers (which have moving parts, vs the pressure sensor) 
have also been around for a long time.  I've not checked, but I'll bet 
some are in hermetic packages which get He leak tested.  If there was a 
Helium problem, you'd have heard about it.
3) There *are* stories about trace contaminants affecting the 
performance of MEMS RF switches, specifically water vapor - it affects 
the stiction of the moving contacts.
4) What is the proposed mechanism for Helium affecting the oscillator?


Here's an article from 2006 discussing SiTime's stuff
https://www.rdmag.com/article/2006/04/new-paradigm-time-silicon-mems-resonators-vs-quartz-crystals

They discuss how hydrogen diffuses *out* of the area where the resonator is.

Now, it's possible that in the interests of saving fractions of a penny, 
Apple is using resonators that aren't packaged as well as the SiTime 
units (which are awfully cheap). (although the news stories say Apple is 
using SiTime's parts)

"Apparently, SiTime also is aware of this problem and says its newer 
devices are “impervious to all small-molecule gasses.” But they admit 
older parts were not immune."  I'd be interested in the context for that 
quote.

off SiTime's FAQ page:
How effective is the hermetic seal of MEMS oscillators??
One of the key elements enabling extremely stable MEMS resonators is 
SiTime’s EpiSeal™ process which hermetically seals the resonators during 
wafer processing, eliminating any need for hermetically sealed ceramic 
packaging. SiTime’s EpiSeal resonator is impervious to the highest 
concentration elements in the atmosphere, nitrogen and oxygen, and 
therefore acts as a perfect seal. Previous generations of EpiSeal 
resonators may have been impacted by large concentrations of 
small-molecule gas. Newer EpiSeal resonators are impervious to all 
small-molecule gases. Please contact SiTime in case you are planning to 
use a SiTime device in large concentrations of small-molecule gas, so 
that we can recommend an appropriate, immune part.






http://memtronics.com/files/Zero%20Level%20Packaging%20for%20RF%20MEMS%20Switches%20v7.pdf



ANother paper on packaging

https://file.scirp.org/pdf/JST_2013122009560886.pdf