measuring distance with low resolution

[qahramon0786]

I have a problem with getting exactly time of flying with high resolution! I'm using dspic for getting peak of ultrasonic signal! peak of ultrasonic signal is jumping one period ! How can i solve my problem !
At the moment i have two version! first one i used ultrasonic reciver and second one i used elctronic micrphone with bandpass filter!
Second version gives me very high resolution but my filter not helping me from other signals

 

[misterT]

How can i solve my problem !

What problem?

 

[qahramon0786]

problem with measuring distance with high accuracy! accuracy must be less then 0.5 mm

 

[misterT]

Second version gives me very high resolution but my filter not helping me from other signals

Solution for you: Use the second version.
I don't understand what are these "other signals" and what is the problem with them.

In general it is difficult to get good precision with ultrasound, but you can get ok accuracy if you take the average of many measurements and calibrate your system properly. Still it is very difficult to get 1mm accuracy.

Could you describe how your system works (or is supposed to work) at the moment?

 

[qahramon0786]

in my first version i used ultrasonic receiver and amplified with 5000 times and after that i rectified it and then i used filter. that my version gived me one period error it means 4.28 mm error .
in my second version, i used electronic microphone as a ultrosonic receiver. received signal amplified 5000 times and that signal is converted to digital signal with dspic(500 ksample/s). that version was gived me 0.686 mm error, but it has many feedback, for example microphones band!

 

[misterT]

in my first version i used ultrasonic receiver and amplified with 5000 times and after that i rectified it and then i used filter. that my version gived me one period error it means 4.28 mm error .

If you know the error there should not be any problem to compensate for it! This is called "calibration".
https://en.wikipedia.org/wiki/Calibration

 

[alec_t]

How do you plan to get 0.5mm accuracy? What ultrasonic frequency are you using? If it's, say, 50kHz and we take the velocity of sound as ~ 300mm per mS then one wave period corresponds to 6mm. If you only detect the wave peak then the distance resolution by my reckoning will be 6mm (or 3mm if you use reflected ultrasound).

 

[misterT]

How do you plan to get 0.5mm accuracy? What ultrasonic frequency are you using? If it's, say, 50kHz and we take the velocity of sound as ~ 300mm per mS then one wave period corresponds to 6mm. If you only detect the wave peak then the distance resolution by my reckoning will be 6mm (or 3mm if you use reflected ultrasound).

I agree. If you need better than 1 cm accuracy, ultrasound is a wrong solution (for hobbyist).

 

[qahramon0786]

i think everyone knows about Mimio Xi interecative whiteboard, in that documentation wrote 0.3 mm resolution! i opened and learned . there is a dspic 40mips processor and 500 ksample/s ADC. i want to build my version(for interest). my tranmitter work on 40Khz, if i got half perriod error duriing in receiving ( 1/80000*343m/s=0.0042875m=4.28mm). At the moment i'm not understand recived and transmit signal! there are not similar! i send only one period of signal from tranmitter but in receiver i got many period(Using ultrosonic reciver). but if i used electronic microphone, tranmitted and received signals are very close to each other.

 

[alec_t]

At the moment i'm not understand recived and transmit signal! there are not similar! i send only one period of signal from tranmitter but in receiver i got many period(Using ultrosonic reciver)

The multiple periods are due to ringing of the receiver in response to pulse excitation. Ringing is undamped (or lightly damped) natural oscillation.

 

[qahramon0786]

what kind of solution you can prefer me?

 

[qahramon0786]

I agree. If you need better than 1 cm accuracy, ultrasound is a wrong solution (for hobbyist).

what kind of solution you can advice me?

 

[misterT]

What kind of application is it? What are you measuring and why? Why you need at least 0.5mm accuracy?

 

[qahramon0786]

What kind of application is it? What are you measuring and why? Why you need at least 0.5mm accuracy?

i want to build whiteboard like a Mimio Xi! 0.5mm accuracy need to write something on board

 

[misterT]

i want to build whiteboard like a Mimio Xi! 0.5mm accuracy need to write something on board

I don't think Mimio Xi has anywhere near a 0.5 mm accuracy. I don't even think that any smartphones have that kind of accuracy.

edit: They say that Mimio Xi has 100 dots per inch resolution, but I doubt it is that accurate.. even the tip of the pen is larger than one dot.
"100 dots per inch. Signal tracking frequency 87 times per second."