SAW filter can be used in place of SAW resonator?

[neazoi]

Hello, I was wondering if a SAW filter can be used as a resonator in oscillator circuits, like a SAW resonator does?

BTW, the filter I have available is the TA171E1 171MHz

Thank you

 

[vfone]

Any filter (that produce a phase change when the signal pass through) can be used as a resonator in a oscillator.
Check the net for "SAW Oscillator" and you get a lot of examples.
From my experience, those oscillators do not provide very good frequency stability, stability which is very dependable by the SAW type.

 

[jiripolivka]

In an oscillator one needs a single-pole resonator only to generate a single frequency. Using a filter means that oscillation may take at more than one frequency which is not desired.
SAW resonators and filters are generally lossy and not stale enough in an oscillator. I would advise to use a coaxial or ceramic (coaxial or dielectric) resonator which offers a good stability. Coaxial and DR oscillators are generally used in good commercial products.

 

[neazoi]

In an oscillator one needs a single-pole resonator only to generate a single frequency. Using a filter means that oscillation may take at more than one frequency which is not desired.
SAW resonators and filters are generally lossy and not stale enough in an oscillator. I would advise to use a coaxial or ceramic (coaxial or dielectric) resonator which offers a good stability. Coaxial and DR oscillators are generally used in good commercial products.

Thank sou. It is a problem of the availability of resonators in the frequency I am interested in (actual center of 172.3MHz). I have only found some ceramic filters at 171MHz with BW of about +-8MHz, so they can be oscillate within their bandwidth range. The original circuit that was drawn by someone else, used an 173.3MHz SAW resonator, pulled down to 172.3MHz. One of the pins of the resonator is be grounded, the other connected to the base of a BJT.

 

[vfone]

17.2 MHz was a very common crystal in some of the old PC mother-boards, providing one of the clocks.
Try to find one, multiply x10, and using maximum 3 transistors you get a rock-stable signal on 172 MHz (adjustable up to +/- 30kHz).

 

[neazoi]

17.2 MHz was a very common crystal in some of the old PC mother-boards, providing one of the clocks.
Try to find one, multiply x10, and using maximum 3 transistors you get a rock-stable signal on 172 MHz (adjustable up to +/- 30kHz).

Hm... you are right. The butler oscillator seems suitable...

 

[biff44]

yes you can use one, but as mentioned above the frequency will not be very accurate since a saw bandpass filter has a wide bandpass!

If i were doing it, i would add a series 1-pole LC filter to it to stabilitze the frequency to one point. The saw filter's added dPhase/dF will stabilize the signal's phase noise somewhat. In essensece ALL resonators have fairly flat passband inserton loss, and it is there rapid change in phase vs frequency that makes the oscillastion stabilize on a single frequency.

 

[neazoi]

yes you can use one, but as mentioned above the frequency will not be very accurate since a saw bandpass filter has a wide bandpass!

If i were doing it, i would add a series 1-pole LC filter to it to stabilitze the frequency to one point. The saw filter's added dPhase/dF will stabilize the signal's phase noise somewhat. In essensece ALL resonators have fairly flat passband inserton loss, and it is there rapid change in phase vs frequency that makes the oscillastion stabilize on a single frequency.

No, here I am talking about the solution mentioned in the previous post about the multiplied crystal oscillator. I think that the butler XO will be suitable for this one.