I’m thinking about buying a few to try some basic room navigating stuff… just need to know what they are accurate to… the web page says 1.5 to 115 inches… so does that mean its only good at 1.5 inch increments??
I’d like to be able to place it at ground level and have it basically go in a circle(measuring every 4 degrees as thats the limit of the optical shaft encoder) and then look for gaps to move through.
from my personal experience the ultrasonic sensors are good for finding large objects(walls, Atlas balls). I belive that the ultrasonic sensor works from 1.5" tp 115 on any increment. to calculate the values given use the print to screen function and watch the values given. Start the sensor touching the wall and slowly back away at equal increments. In my own experimentation i found that the value change is most easilly noticed in about 1/2" increments.
Well, I am looking for accuracy of something close to an inch. I do plan to do a bit of experimenting with it to see what all I can get out of it. The ideal goal would be very rough 3d composites from the front only kinda thing.
IE, if I put a ball infront of it it gets results in say a 10x10 grid with the numbers inside being the ultrasonic results then I could at least look at it and go “Oh, that could be a cup.” etc etc…
Just wanna see how well it works. using optical shaft encoders to get at least somewhat accurate readings on where the sensor is pointing.
One thing to keep in mind on the ultrasonic sensor is that the beam width is quite wide - nearly 15 degres. That means for any direction, you get the distance to the nearest surface within 15 degrees of the sensor pointing angle. This sensor is not very good for mapping of any sort. See the good writeup at http://www.robotbuilder.co.uk/Resources/Articles/138.aspx
I’ve got the Sharp GP12D2 IR ranging sensors working just fine on my Vex robot – use them as analog inputs. They are good out to about 4’ and have a very tight beam. see http://www.acroname.com for details.