OK well that helps some.
This is odd. Do I understand correctly that for example 1) if you command the sled to move down the motor turns clockwise, but 2) if you command the sled to move up the motor also turns clockwise?
That sounds like it might be an issue with the kinematics. Do the values for the machine dimensions in your settings look reasonable (nothing is negative or miles apart)? What are the coordinates the machine thinks itâs at?
Machine dimensions look fine. I think you might been right about it running in one direction no matter what when it thought it was out of the workspace area. Sorry Iâm still learning the software 
Just to clarify. Does anyone have a video of just the motors running so I can compare sound?
And, is it normal for the heatsink to get pretty hot just moving say 40 inches without any load on the shaft?
One other note, I am using an off brand Arduino. So there is that possibility, which is why I plan to pick up an official one.
You can try these settings that seemed to help another user with a soemwhat similar issue:
It gets warm yes, but not super hot.
1 Like
Where do you enable custom settings?
Found it.
At Kp = 5 it actually sounded better, still jumpy. At Kp=3 is pretty amazing.
So maybe its a physical load related issue?
Also, is it normal after going into TEST MOTORS function. Coming out and doing a simple move command it starts fast for a good 1-3 seconds (seems variable in time), then slows down to a slower speed?
Let me add to that. Coming out of test motors function (its last move was clockwise, and it rolls a little past when motor turns off). and clicking manual move down, the motor will spin counter clockwise just a bit, then reverse and go clockwise in its slower speed. Looks like maybe its using full speed to catch up to where it thinks it should have stopped.
1 Like
Yes, I donât think test motors updates the internal state of where the sled is. So your guess is right. Hitting the stop command, even if nothing is running, should always cause the machine to recalculate itâs internal representation of its position.
Wow Kp at 3 is working for you. Hmm. Yeah, part of it might be the lack of resistance. But I wouldnât have guessed that much. The rest of us are able to run the calibration process without issue.
There is clearly some oddity with your setup. Where it is I donât know.
If you have an oscilliscope and you are inclined to play with things. You can play with the following command.
B16 L1 S100 F150
That will cause the left motor to run at voltages starting at 100 incrementing by one each 200 milliseconds and ending at 150. To test the right motor it would be R1. S is the staring value F is the finishing value. The valid range is -255 - 255.
This bypasses the PID control completely. You should be able to check the M+ and M- and they shouldnât fluctuate. If possible it would also be interesting to know if the PWM looks approximately correct. I donât know how it could be possible, but maybe the off brand arduino isnât getting the PWM right? That would explain some of this.
2 Likes