Ultrasonic in distance measurement

1. Jul 6, 2010

newyorker

Hi tehre, this is my firs post here!

I'm designing a complex system using a 40kHz ultrasonic transducers in distance measurement for waste container level detection.
Now I have a big problem because the hot season change the air properties in temperature and humidity.
Anyone may help me where I can find tutorials on how handles this problem?

Thanks to all,

NB

2. Jul 6, 2010

Hello newyorker.Whether there is a problem or not I think depends on the level of accuracy you need.I suggest that the first thing you do is to collect data which shows how the speed varies with atmospheric conditions.Try googling NIST and search their site.

Last edited: Jul 6, 2010
3. Jul 6, 2010

skeptic2

I remember reading about Polaroid's development of their ultrasonic module and how if the reflecting surface is not flat and uniform, echos from different parts of the surface can arrive out of phase and the echo won't be detected at all. To prevent that, the ultrasonic transmitter steps quickly through four different frequencies.

4. Jul 7, 2010

newyorker

Yes, really a frequency sweep could be a good solution, but my transducer is built in an alluminium metal can, because I need it waterproof since I work for the waste field. My transducer has a very narrow bandwidth, 2 or 3kHz maximum, so I cannot follow this way.

5. Jul 7, 2010

newyorker

But anyone knows where I can find C++ library with specific math functions that take in consideration the influence of temperature and humidity in ultrasonic distance measurement?

NB

6. Jul 7, 2010

7. Jul 7, 2010

newyorker

Not high accurated in truth. I have to measure the percentage of empty level of a waste container, but the temperature and air density when sun shining have a huge variation.

NY

8. Jul 7, 2010

skeptic2

How are you measuring the distance? Are you enabling a counter which counts an oscillator until the echo is detected? Are you using a microcontroller? Would it be possible for the microcontroller to read a temperature indicator and adjust the measured distance based on the temperature?

9. Jul 8, 2010

newyorker

I'm measuring a distance in a waste container. need to achieve the distance between the roof (where the device is fixed) and the trash on the bottom. The device is mounted outside the tank, so it's exposed to the atmospherics conditions and all job is done via embedded microcontroller card, and I have a temperature sensor, but not inside the tank.
Actually the measure is temperature compesated, but when the air in the tank raise above 122 °F occours a strange phenomena and the temperature compensation is not enought. I think the influence of air humidity, so the air density, change a lot in my measurement. Since, actually I haven't any RH sensing, I'm looking for a measurement method to prevent this trouble.

.... mission impossible?

NY

10. Jul 8, 2010

skeptic2

Which direction is the error? Does it show the distance to be less than actual or greater than actual when the temperature is over 122 F? If you are counting pulses to measure the distance, could the pulse generator be changing frequency with temperature? According to the link I provided, air pressure has no effect on the speed of sound. Would it be possible to locate everything outside the tank except the transducer?

11. Jul 8, 2010

newyorker

Dear skeptic2, thank U 4 your posts...

Unfortunately I haven't any target located outside the tank to use as a fix reference....and a frequency variation is not possible, because the transducer has a very narrow bandwidth (4..5kHz only).

Anyway the problem is the following: when temperature is quite cold (less 100/122 F), the received signal is very good and the target echo signal is very clear. But when temperature rise over 120/122 F the received signal has a big offset, that hide any other and this make difficult to detect the echo signal.

I'm trying to reduce the emitting signal, only when temperature is high, but I haven't reach a good goals yet... :(

NY

12. Jul 8, 2010

skeptic2

How do you determine the distance from the echo time of the ultrasonic pulse. Are you counting pulses or are you integrating a fixed voltage for that period of time?

I don't understand what you mean by offset in your sentence "...the received signal has a big offset, that hide any other and this make difficult to detect the echo signal." Why are you trying to reduce the emitting signal? (reduce the amplitude or frequency?)

13. Jul 8, 2010

newyorker

The distance is determined by counting the time between the echo and beginning of 40kHz emitting busrt, through a derivative function.
Not simple to explain...hmm...I'm trying to reduce the emitting signal, by lowering the voltage amplitude applied to pulse tranformer that generates the 40kHz carrier, because sometimes I receive reverberation echo from the tank walls. But it is only a test.

Regarding the sentenced offset, is related to signal scope sampled to the A/D port of the microcontroller.

Example of good signal
|
||||........................... burst
|
| +
| ++
| + +
|===== ===== echo signal

Example of dirty signal
|
||||........................... burst
|
| +
| + +
| ++ +
|++++ +++++
|
|=============== echo signal with dc offset for temperature effect

14. Jul 8, 2010

skeptic2

Thanks. So due to the offset, you may be picking up noise that precedes the arrival of the echo resulting in a measurement of too short a distance?

15. Jul 8, 2010

newyorker

Not exactly.
By this posting not possible to load image, and the sketch did not represent the phenomena properly.
Really the noise is summed to the scope as a DC offset, so this hide a part of the echo signal. Since I use a derivative function I can detect the echo signal until the noise is not too much.
So the problem is that with temperature increasing, increase the noise and I cannot well understand wich is the physics phenomena that gernerate it, but I have only some ideas, because this happens in the tank only, outside it the target detection is very good.

Physics misteria..........

16. Jul 8, 2010

skeptic2

Would it be possible to locate all the electronics outside the tank and run a cable from the electronics to the ultrasonic transducer inside the tank?

17. Jul 8, 2010

skeptic2

Another thing that might help is to ramp up the receiver gain with time. That way, when the target is close, the echo doesn't overload the receiver and when the target is far, you get noise immunity during the time you're waiting for the echo.