Electronics > Electronics
Fuzzy sonar spinner.
nanob0t:
--- Quote from: Admin on April 17, 2007, 08:25:22 AM ---FYI - its sharp IR, not sonar, that is spinning on the servo :P
--- End quote ---
I don't know what you are refrencing. On my robot I used IR, but it was still the same concept of an array.
Yes, and no trig is necessary. The only math you will need to do is determine how many places you want to scan and divide 360 by it. This will give you the degree the servo will need to rotate to take a measurement. If I wanted to scan 4 times in a 360 degree array. 360/4 = 90, so I would have to find a pulse for the servo equal to rotating it 90 degrees whatever. Then at each spot you have the robot pulse out to the sensor and wait for input, which goes through a calibration equation to output your distance.
Fixing it to the robot requires lots of trig, though. You need to find the radius of the wheel and do some stuff with funtacious radians, then compare it to the chassis of ther robot, and if your robot isn't a perfect square, you get to deal with ellipses! What fun! ;D
Stick with the servo :D
Admin:
--- Quote ---
--- Quote ---FYI - its sharp IR, not sonar, that is spinning on the servo
--- End quote ---
I don't know what you are refrencing.
--- End quote ---
sorry, that message wasnt meant for you but the others before you that called my three sharp IR on a servo 'fuzzy's spinner sonar' :P
--- Quote ---Fixing it to the robot requires lots of trig, though. You need to find the radius of the wheel and do some stuff with funtacious radians, then compare it to the chassis of ther robot, and if your robot isn't a perfect square, you get to deal with ellipses! What fun!
--- End quote ---
i meant that the robot can stop, spin in a circle and map, choose a new heading, then move again. spinning while moving makes it much much harder i agree . . .
nanob0t:
--- Quote from: Admin on April 17, 2007, 09:21:47 AM ---i meant that the robot can stop, spin in a circle and map, choose a new heading, then move again. spinning while moving makes it much much harder i agree . . .
--- End quote ---
Hmmm, I have not had good experience with sonar and such. I don't understand how it could get accurate measurements while rotating in a circle, as the pulse would be sent at one point and be recieve at another point, although rather insignificant due to the speed of sound.
I figured that you would calibrate it so it would momentarily stop in order to get accurate measurements. If you wanted a precise map of the room, you would just increase the array and the amount of stops it takes in the room. Instead of taking just 4 points, it could take 36 at each 10 degrees. Then just maximize it's efficiency, so it can rotate as quick as possible.
Even when rotating, you have to do some trig, so you can calculate how long it takes the robot to move in a circle around the robot's radius. Any more and it would be mapping two points twice, and it would get a repeated image.
ed1380:
Fuzzy? I though Admin's bot was called Stampy
Admin:
Ive built 30+ something robots in the last 5 years, but only named like 15 of them (6 of those had the same name), and only documented about 10 of them . . .
http://www.societyofrobots.com/robots.shtml
Stampy
http://www.societyofrobots.com/robot_sumo.shtml
Fuzzy
http://www.societyofrobots.com/robot_omni_wheel.shtml
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version