wyojustinParticipantFebruary 11, 2019 at 1:15 pmPost count: 16
I just updated the SD card with the latest image (2.1.0). The new image will not calibrate shoulder or elbow, but just chatters away until the 30 seconds times out. I rolled back to original SD card,and found cal procedure worked as expected.
Is there a way to modify the calibration constants so I can relax them a bit in the 2.1.0 code? I’ve already tried tensioning the belts (which resulted in a crack and a repair).
Fix:Edouard RenardKeymasterFebruary 12, 2019 at 1:57 pmPost count: 239
I guess your issue is coming from the belt or pulley from the axis 3.
What changed from 2.0.0 to 2.1.0 (Raspberry Pi image):
– Torque on axis 2 stays ON during all calibration (in case the torsion spring makes the arm lean forward)
– Speed on axis 3 during calibration is reduced a little bit, but the torque stays the same.
So, what happened here is (my guess): your axis 3 belt is not tensed enough, but just at the limit so that when the speed is increased a bit, the calibration will still be done. The axis 2 should not have a problem, I think it’s already correctly tensed.wyojustinParticipantFebruary 13, 2019 at 11:49 pmPost count: 16
No joy. First thing I did this morning is run the 2.0.0 code. Axis two chattered. I tried tensioning both axis two and axis three. Axis two still continues to chatter and worse, when I tightened the screws I stripped one (another appeared to be stripped when it arrived). Anyway, it looks like I need a new motor for axis two (shoulder axis). Can you point me to a part number?wyojustinParticipantFebruary 14, 2019 at 11:58 pmPost count: 16
Some Joy!! (Copied from slack channel)
Ok, here is was my path from many failures to success:
1. Effed up N1 by emulating a joystick in python and hitting the “speed up” button. Also double speed restrictions (less restrictive). That sped things up, but the motors seemed to stall out. Changing everything back, did not fix motor stall problem.
2. Upgraded firmware to start all over. Now N1 would not calibrate as more tension is required on the stepper axes.
3. Tightened timing belts, cracked N1 tensioner. Fixed with 3D printed/glued part.
4. Tensioning belts did not fix calibration on new 2.1.0 sw. Rolled back to 2.0.0.
5. Stripped motor threads. Screws from either end of the motor were butting up against each other with the motor mount screw not having enough purchase. I backed off the screws on the back side to make room for the mounting screws. I didn’t have the right size screw, so I fixed with a screw AND nut. New system has plenty of bite.
6. Tightened the SNOT out of all timing belts. 2.0.0 works! Woot!
7. Tested 2.1.0. WORKS!!!!
One suggestion @Edouard Renard, make screw mount screws 2mm longer and through screws 4mm shorter (or thread opposite site and use same size screws on each side).
Spoke too soon. New 2.1.0 code has the same stalling problem as 2.0.0. Now I need to resolve the stalling problem as it has infected both versions of the code. I will start a new thread if I can’t find the solution.
- You must be logged in to reply to this topic.