raspberry_pi_camera_bot (experimental)

[michelvancan]

Hello,

I made a robot with Dagu Rover5 + Raspberry Pi + Dagu Mini driver. I found that the work on the experimental branch of raspberry_pi_camera_bot is very interresting (thanks). I have made the following modifications :

1) in mini_driver_firmware.ino :
- I modify waiting time for ultrasonic from 500 to 800
- I change the sign of the left encoder value in order to have the same sign for the two encoder when going forward.

2) in mini_driver.py
- In order to recognize mini driver I change : "lsusb -d 10c4:ea61 " to "lsusb -d 10c4:ea60"
- I change PRESET_ENCODER_TICKS_PER_REVOLUTION from 1000/3 to 256 (I don't know why but this value work!)

Now the command "bot.drive_straight(0.3)" work perfectly, but the command to turn the robot doesn't work (I need to put 190 in order to turn 90 deg!!).

Thanks for this work, Cordialement Michel Van Caneghem

[Alan]

Hi Michel,

Welcome to the forum. Sorry it's taken so long for me to reply to you.

That's really cool to hear that you got the Rover 5 working with the Mini Driver. I would have thought that there might be some power issues there. How exactly are you powering the Pi and the Rover 5? Are you able to run everything off 6xAA rechargeable batteries?

That's odd that you needed to change the lsusb command to detect the Mini Driver. I would have thought that they would all have the same VID/PID. Perhaps I'll need to check for a range to get that working reliably.

Interesting to hear tht you've got driving in a straight line to work well. I'd really love to know how well that works for longer distances? One of the many things that's been holding me back from releasing a Rover 5 version of out Raspberry Pi robot kit is that it can sometimes be tough to get it to go in a straight line. I've found that depending upon the surface type you're running on (smooth seems to be worst) it's possible to get slippage between one of the wheels and the tank track, making the encoder reading pretty useless.

Also, I noticed the strange requirement to practically double the turn angle also. Haven't really looked at this yet, but I'm assuming I must have done something dumb with the maths somewhere.

Regards

Alan