Crystal Oscillator architecture limitations

[malaganas]

Hi,

I have been reading that different XO architectures are limited by the oscillation frequency of the crystal i.e. the famous pierce/collpits oscillators are limited until 200MHz crystal frequency.

In what sense is that correct? Is the limitation factor the amount of negative resistance provided by the core circuit or is it something else that I am missing?

In this case for high frequency crystals (>200MHz) what architectures are more suitable and why?

Regards

 

[chuckey]

Crystals actually mechanically wobble at their oscillation frequency, perhaps there are constraints that cause them not to wobble any faster then 200MHZ.
Frank

 

[D.A.(Tony)Stewart]

Both work well in GHz resonators with better phase noise if low ESR resonator using Pierce mode.

MEMS resonators are common in GHz range with ZnO film over ceramic with Al electrode.

 

[Warpspeed]

Crystals actually mechanically wobble at their oscillation frequency, perhaps there are constraints that cause them not to wobble any faster then 200MHZ.
Frank

This is quite true.
The "traditional" quartz crystal resonator is a very thin wafer, the whole of which bends and wobbles at its own mechanical resonant frequency, somewhat like the way a tuning fork vibrates.
There is a mechanical limit beyond which the whole thing cannot be made small enough, or thin enough to vibrate any faster.

The solution to that is overtone operation at a harmonic of the basic resonator mechanical resonance. that gets you up much higher in frequency, but you require an oscillator circuit designed for an overtone crystal.

Faster still ?
The trend now is towards surface acoustic wave resonators, where the whole mass of the thing is not involved in the wobbling vibrating action. A surface wave only is involved which allows frequencies well up into the Ghz region.

Its still a quartz resonator (of sorts) but the mode of operation is vastly different to the original mechanically ground wafers of quarts used in crystal oscillators around a hundred years ago.

 

[malaganas]

Hi,

Thanks for the replies.

The trend now is towards surface acoustic wave resonators, where the whole mass of the thing is not involved in the wobbling vibrating action. A surface wave only is involved which allows frequencies well up into the Ghz region.

Its still a quartz resonator (of sorts) but the mode of operation is vastly different to the original mechanically ground wafers of quarts used in crystal oscillators around a hundred years ago.

Could you give some more details about them? Are they modelled same way as a quartz crystal?

Both work well in GHz resonators with better phase noise if low ESR resonator using Pierce mode.


View attachment 121149

You mean to say that pierce oscillators have not really limitations due to the crystal frequency, right?

Crystals actually mechanically wobble at their oscillation frequency, perhaps there are constraints that cause them not to wobble any faster then 200MHZ.
Frank

Not really sure about this, nowadays there are quartz resonators that exceed a lot this limit, for example normal AT cuts with series resonance of 500 MHz.

 

[D.A.(Tony)Stewart]

Series mode is preferred for harmonic tuned UHF due to less sensitive to stray capacitance since low impedance. Choice of amplifier is possible to suit frequency in GHz range using GaAs etc, to improve gain if needed, but the architecture is not limited. Only the ESR of the resonator. Resonator must be larger to get low ESR, since ESR increases with each harmonic.

Thus resonator may be ~1cm instead of few mm. for High Harmonic Resonator.

Yes resonator uses same equivalent circuit with Rs,Cs,Ls,Cp