FindingNiryoParticipantOctober 12, 2018 at 2:06 pmPost count: 54
Joint 4 Elbow rotate is over-heating and shutting down on V2. There is no remote reset, robot must be turned off and back on to clear the flag and resume operation.
#1 Providing DXL reset would allow errors to be cleared: protocol2_packet_handler.cpp int Protocol2PacketHandler::reboot(PortHandler *port, uint8_t id, uint8_t *error)
#2 Giving access to servo temperature in Niryo_one Studio through a parameter block would allow joints to be monitored and robot moved to safe position for cool-down cycle before over-heat limit.
#3 Expose PWM-Limit as torque control for XL430 (through setMaxTorque() setGoalTorque(), synchWriteGoalTorque())Edouard RenardKeymasterOctober 15, 2018 at 4:37 pmPost count: 152
Thanks for pointing this issue. We have experienced the same thing and have come up with a solution.
The actual problem is that the Dynamixel XL-430 need a 11.1V supply. We are using a 12V power adapter here, and this is what causes the overheating in the motor.
We have tried with just 11.1V and everything is fine: the motor can work indefinitely without overheating, even with some added payload.
(we have also tried your point #3, but this is not what makes the overheating. It really only comes from the too high power supply)
So, the solution :
Use a 11.1V power adapter. You can get on on the Internet, and for new orders we’ll change the power adapter that we ship with the robot. 🙂BManCanParticipantOctober 15, 2018 at 7:38 pmPost count: 23
Have you already started shipping new orders with the 11.1V power supplies? My Niryo One kit just arrived last week. I guess I will unpack it and check the power supply tonight.
Thanks.Edouard RenardKeymasterOctober 16, 2018 at 8:42 amPost count: 152
We have ordered new power supplies that we’ll ship with the next robot batches. If you’ve already received your robot you get the 12V power supply.
Note that the robot will work fine and the overheating on the motors will happen after you run it for about 1h (also there is an internal protection so the motor will not be damaged, instead it will just reboot)FindingNiryoParticipantOctober 23, 2018 at 4:12 pmPost count: 54
Joint 4 over-heating looks related to NiryoOne V2 Dynamixel xl430-w250 not reaching goal position
As noted Dynamixels at J4 & J5 do not achieve target position. It appears the motors may be expending extra energy without moving the joints.
With motor torque “ON” (not in learn mode) J4 warms to 65 Celsius in an air conditioned office; even when no movement is requested.
When joint position equals target (for example: Setting target to zero and forcing joint to be 0.000) J4 runs at 43 Celsius; even at 12V supply.shoune77ParticipantOctober 23, 2018 at 7:16 pmPost count: 3
With the 12 V power supply, would it be a good idea to put a diode or two in serial with the XL-430 supply wire, for a 0,7 or a 1,4 voltage dropout ?BManCanParticipantOctober 24, 2018 at 2:56 amPost count: 23
I have searched online and local shops for a 11 or 11.1V power supply without any success. My Niryo One was just purchased in September, so I am asking that you consider shipping the recommended power supply to customers who recently purchased the v2 kits. Having the servos overheat and reboot should not be considered acceptable.
Thanks.FindingNiryoParticipantOctober 24, 2018 at 1:31 pmPost count: 54
Confirmed: DXL430-W250 Joint #4 (elbow rotate) continues to apply power without moving joint (causing over-heat) .
After move request from NiryoOne Studio Dynamixel stops short of goal (as reported here: NiryoOne V2 Dynamixel xl430-w250 not reaching goal position)
Voltage applied to the motor is proportional to the distance from target angle; measured approximately 46 V/radian.
The joint stops moving typically (0.033 radians offset from goal) with 1.5 volts still being applied to the motor.
Modifying the Dynamixel PID parameters should allow XL430 to reach goal position and stop over-heating without motion.FindingNiryoParticipantJanuary 22, 2019 at 3:00 pmPost count: 54
Have any changes been made to the PID control values to improve the positioning accuracy and/or reduce the temperature build-up?
Trajectory planning includes velocity and acceleration which can be added to the Dynamixel PID feed forward.
If the Dynamixels can not be relied upon to reach their target position then perhaps a software undershoot/overshoot strategies should be explored.
Worst case a deadband should be implemented to change requested position once motion has stopped (with loss of positional accuracy).
You must be logged in to reply to this topic.