Any tips on how to avoid this?
I am experimenting with writing a program to have the Rover make turns by spinning in place. Is that maybe the cause?
In general, what maneuvers might cause the tread to fall off?
Thanks!
First, can you send an image of the rover so we know the motor has been assembled correctly? It needs to be in configuration C only. The rover also uses all of the track segments equally on either side.
If the rover has been assembled correctly, then rotating on the spot on a surface with a lot of friction will cause the tracks to slide off.
Note that “skid steering” forces the tracks sideways. Normally this does not happen on smooth surfaces like a tabletop or floor.
Thanks. Here are two photos of my rover. I don’t know if you can see what you are looking for, whether it’s config C: snag.gy/wthFSC.jpg
Indeed I was doing on the spot rotating, and on carpet, so probably significant friction. So your explanation about skid steering and sideway force on the track makes sense to me.
What do you think of the photographs, does it look right?
Thanks!
Rover assembly looks good. Looks like you opted for the crimps rather than soldering the wires to the motor leads, so just check that they don’t fall off.
You can run the battery wire below the PCB to make it cleaner.
Note that even normal (military) tanks lose their tracks in those situations (not on carpet, but on high friction surfaces).
Now that you have the rover up and running, what do you plan to do next?
Youre right about the crimps because they fall off. They are a pain. I was following the instructions. On the other hand, I had the setscrew in the gearbox get loose, and I had to remove it to get to the setscrew, in that case the crimps were handy. But net net, I am ditching them.
I’m a software guy (as you can tell?). I am seeing how far I can get towards a simple autonomous rover. My first “mission” is can I get it to drive in a 1meter square pattern. I can already tell that I need sensors because the effect of running the motor is not deterministic on position.
What sensors do you suggest?
The crimps included were the best we could find for the motors (we don’t manufacture the Tamiya Twin Motor Gearbox and cannot modify the contents). With proper crimping, they should stay on, but soldering is always best.
Regarding the set screw - make sure it’s pressing against the flat part of the shaft.
Since the DC motors included are not perfectly identical, you’ll find that the rover does a gentle arc when both motors are operated at the same speed. To move in a straight line, you’ll need to add encoders:
robotshop.com/en/encoder-pai … arbox.html
You would have the microcontroller continuously read the encoder count to ensure they are identical, and adjust the speed of the motors accordingly.
The encoders can also be used to provide distance traveled, knowing the radius of the drive wheel.
An alternative would be a magnetic compass (to provide absolute compass values):
robotshop.com/en/magnetic-se … mpass.html
For obstacle detection:
robotshop.com/en/sharp-gp2d1 … cable.html
robotshop.com/en/lynxmotion- … using.html
Getting absolute positioning indoors without some sort of external means (line or sensors) is not easy, since there is slippage of the wheel.