[time-nuts] Helium and MEMS oscillators don;t mix well
kb8tq at n1k.org
Thu Nov 1 15:49:43 EDT 2018
If you take an IC package up to fairly high pressure for a few hours (and it’s
defective) you can pretty well fill it with helium and have it well above 1 ATM internal
pressure. It will then leak out (maybe over a few years) until it gets back down to one
In some really rare conditions, you can pressurize it to high enough PSI to inflate the
package when the pressure goes back to 1ATM. Even then the oscillators generally still work ok.
This is inflation not really a function of helium. It’s just a function of what a lot of PSI does to
a sealed package. You could inflate it with air and see the same thing.
Getting helium in a properly sealed package *without* a lot of pressure *and* very
high concentrations (like 100%) …. very difficult.
> On Nov 1, 2018, at 3:27 PM, Wayne Holder <wayne.holder at gmail.com> wrote:
> The oscillator mentioned in the article is a SiT1532 made by SiTime
> <http://SiT1532>. It's sold in a chip scale package that's only 1.5mm x
> .9mm, which means it'a no much more than a chip of silicon with some solder
> balls attached. The data sheet indicates there is a "polymer" coating on
> the back side of the chip, but the working surface would be in the bottom
> where the solder balls are. There is a rectangular protrusion shown on the
> "Dimensions and Patterns" section (page 12) that's right over where the
> MEMS mechanism would sit that might be some type of seal, but there is no
> descriptive text.
> The curious thing to me is that some iPhones are said not to recover from
> exposure to helium but, as an essentially mechanical device, I can think of
> no reason that the SiT1532 would not recover from exposure to helium after
> the gas had migrated out. I wonder off the iPhone could be damaged by an
> oscillator failure, o one that's running off frequency? The devices sell
> for about $1.25 at Mouser and I have a tank of helium in the garage, so I'm
> thinking about doing an experiment. The only problems is finding a way to
> solder wires to such a small part? Might have to make a PCB, instead.
> On Wed, Oct 31, 2018 at 12:51 PM Mark Sims <holrum at hotmail.com> wrote:
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to
>> and follow the instructions there.
> 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