[time-nuts] BVA resonators

Murray Greenman murray at rakon.co.nz
Sun Dec 9 12:59:16 EST 2007


In the BVA device, it's not so much that the 'top is domed', but that
the resonator is plano-convex. I stand to be corrected, but I believe
plano-convex crystals are used in high quality oscillators because the
shape leads to a lower level of spurious responses, hence fewer 'dips'.

BVA (and QAS, which is a similar technology) devices have very low
ageing, very high Q etc, partly because of the construction, and partly
because of the processing skill of the manufacturer (a lot of the
processing techniques are closely guarded secrets).

The main applications for these devices lead from the very high Q
properties of the construction, which gives the devices low ageing
(handy for a spacecraft application), but also extremely good short-term
stability. The 1s tau AV can be better than 1e-13.

There are plenty of devices with low ageing and good long-term stability
(Rb oscillators, for example), but short term stability is a very rare
commodity. Some of the newer atomic sources (the Hydrogen Maser is a
good example) exhibit very good short term stability, but if you don't
have $200k to spare, you look for something else. A BVA or QAS device
can be phase locked to a Cs source etc for long term correction, but the
best short-term performance is still going to be as a stand-alone
device, where it out-performs most other technologies.

So who needs high short-term stability performance? Anyone making high
performance OCXOs, that's who! If you are intending to measure the
frequency vs temperature performance of a 1e-11 level OCXO, you won't
know if the perturbations are temperature related or reference related
without high short-term stability. Similarly, measuring ageing of a
device with ageing spec better than 1e-9/year will be impossible on any
time scaler shorter than months unless your reference is short-term
stable to 1e-13.


Murray ZL1BPU





More information about the time-nuts mailing list