Motors Burning Out on RD4B

Hello all,
On my RD4B, I have a problem with burning out. When I lift up, the robot works fine for one or two hours when I drive (I’m not driving continuously, of course:). Then, the motors start to burn out whenever I lift two cubes. I am not sure if it is mainly a battery issue or something else. Right now I have a RD4B with each stage powered by 2 motors and rubberbands. All the four motors are plugged into a power expander. I have thought about putting two of these onto the cortex. Would that help at all? Thank you!

Also, I forgot to add something. My arm has slew rate control implemented in it. The bottom arm works with trim, and the top arm tracks the bottom one’s angle with PID.

It shouldn’t be a battery problem if the motors are burning out. If the motors are getting hot and lose the ability to lift two cubes, it all comes down to tuning. Off the top of my head, my assumptions are that

-you’re using a 7:1 torque ratio
-you’re using a tuned elastic assist (maintains height without motors mounted)
-you’re replacing the battery after every 6 minutes or earlier
-the screws along the linkage are properly tightened
-your axles aren’t bent
-your axles aren’t digging into the tower
-motors are properly tightened
-your lift does not arc when going up or down
-the drivers do not quickly swap between raising and lowering the lift for an extended period (quick, continual routines on skyrise would count)

if any of these are wrong, then that would contribute to the burn out. Either way, it sounds like you’re doing a lot of mechanical things correctly. You’re probably pushing the robot too hard.

So right now I actually have a 9.375:1 ratio (15:1 on speed). I think I will try to add more rubberbands, but I have concerns about being out of size when the lift pops up at the bottom. For batteries, I guess I should replace it more often. Screws are tightened, axles are fine, motors are tightened. What do you mean by arcing? As for swapping between raising and lowering the lift, I would think that the slew rate control would help compensate for that. Would it be better to split the arm power onto cortex and power expander batteries?
Thanks once again!

Do you mean the lift motors work well for two hours? or just the drive motors? Which motors are overheating?

You said the you are using PID control, what is the value being sent to the motors when the lift stops?

Actually, jpearman’s comment would make a difference.

If your drive was overheating, there’s not much you can do than increase torque and lower weight. If your lift was overheating, then that’s another story.

Splitting the power between the cortex and power expander shouldn’t help you with overheating. With overheating, too much current is going through the motor. So you aren’t low on current or battery power, it’s the other way around. In fact, I have a feeling that if you split the power between your power expander and cortex, I have a feeling arcing will be more of a problem (if it was a problem)

By arcing, I meant one side of the lift is raising faster than the other, causing the lift to veer to the left or right. This is prominent in scissor lift systems, but can also be seen as other lift systems get taller and taller. In this case, if you split the lift system between two different battery sources, if one dies down, then your PID would work at a lower power to compense, or the PID doesn’t work fast enough and your lift begins to arc an create frictional issues

On the other hand, your gear ratio is fine, and it’s good that your motors and axles are correct. However, you want to be careful with the screws you use as joints. By “tightened”, you want them to be loose enough to easily spin in your bearings, but tight enough that lift wobble is minimized. Places you want to check are where you mount the intake to the last stage of the rd4b, and screws through gears (if you use high strength). Another thing you need to check is the rubber bands. It is possible that you’re over-rubber banding the lift. If you’re not careful with where you place your rubberbands, then you can create excess friction, making your motors work a lot more, especially with PID continuously correcting your lift. For example, if rubberbands are mounted directly on joints, mounting rubberbands on only one side of one side, or if you don’t compensate for the backlash your motors experience, then all of these are stressing the motors more than it needs to be. It’s critical that you experiment with different angles to see where elastic is uniform.

Edit: Then again it is possible that if you tripped all 4 motor ptcs you might trip the ptc in the cortex, which could explain why the lift fails, but you still need to resolve the problems with the lift motors first with either scenario.