In this answer they talk about
SAW (Surface Acoustic Wave) and
SDL (SAW Delay Line). I understand SAW can be used to create filters, oscillators and delay lines.
I would like to know what typical properties of SAW are, and following from that, what areas they are used in. For instance, they're used in HF, but do they also exist for LF like audio, or isn't that possible?
@JustJeff already mentioned that SAWs are narrow band.
In this answer they talk about
SAWs, like you said, can be used to create filters, oscillators, and delay lines. The idea of SAWs is that you produce an acoustic wave that is then passed down a surface (hints the name) and then is transferred back to an electrical wave after passing along the surface.
A good comparison is an audio PA system, you produce an acoustic wave using a speaker, it then travels through air (versus a surface in the SAW), after the propagation time through the air a microphone can pick the signal back and turn it into an electrical wave.
The problem with creating a SAW in a LF form is that the size required to produce the LF acoustic waves is rather large and could never fit in a package like what the HF ones are available in.
There are lots of similarities here, for example, the speaker and microphone both have a relatively narrow frequency range that they are able to reproduce. You may not initially think it is narrow, but 200 Hz to 18 KHz is a common range. Adjust that to the MHz range and less then a 20 KHz bandwidth is pretty small.
I am currently working on a project that uses the fact that the propagation time changes if different chemicals are applied to the surface. This change can be measured by either a frequency change if in an oscillator configuration or phase shift if used in a delay line configuration.
To talk a bit more about how they are used to create filters, oscillators and delay lines... The can be used to create filters in 2 ways, the first is that the properties of the SAW itself have a fairly narrow pass band. This is very nice to get essentially a very high order band pass filter in a small package. They can also be used in delay lines to produce what you might be used to seeing in digital filters. You can use the delayed signal to either feedback into your system of add to your system. You can think of this like adding a sample from x ns ago to your current sample.
They can also be used in an oscillator configuration. These basically can be used for anything that you need a frequency reference for, many times used in PLLs or mixers of some sort.
One cool use I have seen of SAWs is for passive user authentication. This is done by placing reflectors or transducers at different locations on the path of the acoustic wave. The base device then could provide some sort of source, whether it be a single frequency or frequency sweep that then passes through the SAW and then based off of the locations of the reflectors or transducers a unique key can be returned to the base.
Not sure about nowadays but SAW filters were very commonly used as the IF filter in TV receivers for at least a couple of decades. A neat application I read about was for contactless torque sensing - a SAW is bonded to a shaft in a way that makes the resonant frequency shift with increasing torque, and a nearby coil detects the resonant freq.
The basic theory is that you excite transverse acoustic waves on the surface of a piezoelectric material. The acoustic waves travel much faster than sound does in air owing to the solid medium, but much slower than light or electricity. You inject the signal at one end of the device, the waves propagate across the surface, and are picked up at the other end.
The SAW device I worked with was a gas sensor. The surface was coated with a metal oxide that would adsorb the gas of interest. The concentration of the gas would establish an equilibrium mass loading on the surface which I think changed the speed of the waves. This meant the delay across the device would be sensitive to the gas concentration. When the delay was used with a feedback amplifier, an oscillator was formed, the frequency of which would then depend on the gas concentration. From there, measuring the frequency let you measure the gas.
There may be other transducers to couple the electric and acoustic signals, but the one I know of was the 'interdigital filter'. The size of the transducer depends directly on the acoustic wavelength of the operating frequency on the chosen material. I recall a quartz substrate, with interdigit spacings somewhere in the millimeter range, and an operating frequency somewhere between 150MHz and 200MHz. Contrast this to the same frequency as RF, where the wavelength would be 1.5m to 2m.