Utrasonic Speed dector

[flyone4]

I need to make a prototype  that will indicate, through both leds and audio tones, whether a person within its target area of 10 ft is walking or running.  Unit will transmit a ultrasonic wave into the target area and will detect the Doppler shift of the reflected waves to determine the velocity of the person.  A walking person is less then 5 ft per second.  can have a power supply of 120 volts  and 60 hz.  and less then 15 watts of power consumption.    Does anyone know were i can find a re fence to use in order to make something like this  and were would be the best place to buy the parts.  I know i have to use a transducer in order to make it work and the Doppler effect. But any tips or a way to find a re fence model or design that has all ready been done so i can follow it thanks please help.  This is for a school project.

 

[LostViking]

I used to work for a company making current profilers. Sort of the same thing and they used piezo electric transducers.

You will need to transmit a pulse at the target. Don't really need to know the doppler shift since you are not trying to measure the speed exactly, just determine if they are running or walking. What you do need to do is measure the time of flight. This is the difference between when you transmit and receive. You do not transmit continuously, but rather in pulses when you want to detect a target.

Since this MUST be a uP design you would then compare the time against limits for walking and running by taking several reading and looking at the change in position (i.e. 1st target was 5 feet away, 2nd target was 3 feet away). This is much easier than working with doppler shifts. You just measure the difference in distance as a function of time. If transmitting every second...moved 2 ft in 1 second...moving 2ft/sec.

Now if you do want to know how fast they are moving you can sample enough times to get that too.

So, you need a ucontroller to run the thing. Memory can be in the uC. You will need a transmitter circuit, I'm guessing 12Khz is fine for that distance but there may be something off the shelf on a different freq so check around. Of course you will also need to be able to receive. That's really all you need. The rest is software.

If you wanted to get fancy you can start transmitting a digital pulse string and reading that back. That would allow you to distinguish between multiple targets in the same zone. It gets very complicated to do this kind of thing though. Most techniques are patented and closely guarded by the companies.

Sounds like a fun project. Keep us updated!

 

[flyone4]

Thanks for the help were are in the early stages but you gave us a good idea to go off  of. Right now were just buying parts and  trying to get some ideas down we bought the transducers you talked about and your right thats what we needed. For now the teacher is letting us take it slow but in two weeks from today we have to start building it and showing progress to getting it working.  We have 12 weeks to get it  done we have about 9 weeks left now. And start  building it in 2 weeks were testing our transducer today.  We could use your help a  lot in the next week once we start trouble shooting to see what bugs we run into thanks  for you reply talk to you later.

 

[LostViking]

Shoot the q's and I'll answer what I can.