[time-nuts] Grinding crystals...

[Magnus Danielson]

On 06/21/2013 06:11 AM, Jim Lux wrote:
> On 6/20/13 4:57 PM, Gary wrote:
>> A common scheme in metal deposition measurement is to measure the
>> frequency of a crystal prior to starting the deposition process, then
>> monitoring the frequency shift of the crystal as the metal is sputtered.
>>
>> I was told crystals are tuned this way at the factory, but don't know
>> this for a fact.
>
>
> it's also how particle counters based on "piezobalance" work. The
> particles hit the crystal and stick, lowering its frequency.

Which is how one mechanism for frequency drift works, and hence is 
avoided. Whatever junk is on the crystal, it get's thrown off by the 
heating and operation, and that balance take a long time to shift, until 
you turn it off and it deposit back on. Then as you turn it on you 
experience the same thing again. By cleaning the crystals and even bake 
them out, this can be significantly reduced. BVAs for instance is baked 
out with a turbopump operating, and at the end the "stove pipe" is 
pinched of as in normal vacuum processes.

Cheers,
Magnus