Bad readings from the maxbotics sonar

[GuyM142]

We've managed to wire the sonar to the robot and get reading from it with the LV example,
The problem is that the readings are not accurate. (For both digital and analog)
We actually have 2 of them and both of them aren't accurate. (The readings are jumping or gradually rising to what seems random numbers).
We could read right distances once but it went bad again.
What can we do?

[Herbblood]

We had the same problem and found there were some errors, but we took some data to find a number ours 1.242 I think, which would lead the number read to be closer to the actual distance. Some of the problems with the sensor may be that it is being affected by temperature which changes the speed of sound or then "beam" may be picking up other things in the same area.

[GuyM142]

Originally Posted by Herbblood

We had the same problem and found there were some errors, but we took some data to find a number ours 1.242 I think, which would lead the number read to be closer to the actual distance. Some of the problems with the sensor may be that it is being affected by temperature which changes the speed of sound or then "beam" may be picking up other things in the same area.

How did you get to that number?

[JesseK]

We found that there is different backscatter in the beam depending on if you're sensing over carpet vs. sensing over linoleum.

Using a narrow-beam sensor or mounting the sensor > 18" form the ground solved the more general backscatter problem for us, mostly.