I can't respond to the physics but I can tell you to use the 74HCT4046 or better an 'old' signetics NE561 or 2.
Bob, thanks for your reply. After reading your suggestion, I ordered 3 of the CD74HC4046AE from Digi-Key, but only now noticed that they were the CD74HC4046AE, not the CD74HCT
4046AE. Do you know if there will be a significant performance penalty because of my mistake?
I am very dubious about the whole physics of the project.
You are only using one phase locked loop and I don't see where that gets you. Drive the transducers with a differential signal not a single ended one.
Normal wind vectors are measured by three othagional paths and a time of flight, with temprature measurement sepratly as an adjusting factor.
Thanks for your advice on driving the transducer. I've attached an updated block diagram which I believe corrects that particular error. I would be very grateful if you could take a look at it to check if I altered the diagram correctly.
Much of the inspiration for this project comes from this https://docs.google.com/a/google.com/viewer?url=www.google.com/patents/US5877416.pdf
patent. According to the patent, creating the vertical standing wave is very helpful in increasing the amplitude of the signal and thus helpful in increasing the signal to noise ratio. I also worry that if there is no standing wave the reflections from the cavity walls will start to mess with the signal. That's why I have that single PLL controlled by the transducer directly above the signal transducer. I think it will compensate for temperature effects on the speed of sound when setting the frequency that will make a vertical standing wave (at 72 F the 4th harmonic of the cavity is 40 kHz, but the harmonic frequency changes with temperature).
One of the primary objective of this anemometer is for it to be small, so I decided to forgo the 3D wind vectors to make the system more suited to for use on a mobile robotic platform. Once again, thanks for your advice.