ergoCub 1.1 S/N:001 – Left hip roll goes in HF while walking

[isorrentino]

Robot Name 🤖

ergoCub 1.1 S/N:001

Request/Failure description

The left hip roll joint goes into hardware fault as soon as the robot starts walking.

Here the video.

IMG_3240.MOV

Detailed context

Here is the log file

log_ergocub-torso_yarprobotinterface_1154.txt

cc @DanielePucci @GiulioRomualdi

Additional context

No response

How does it affect you?

No response

[DanielePucci]

Most likely, the motor was burned after we performed one-foot balancing experiments since that motor was pretty hot.

This is another element that confirms the importance of the temperature protection feature that is currently being implemented by iCub Tech at the firmware level, and that hopefully will be deployed onto the robot soon.

CC
@ale-git @valegagge

[GiulioRomualdi]

Hi all, by any chance, if the motor is burned, is there another one to replace it?

[maggia80]

@GiulioRomualdi yes we have two more motors left. The problem is that we have zero mech guys till Gen 2nd...

[mebbaid]

We were trying an integration experiments with HSP maniupulation team using the walking controllers, but this issue is blocking.

[fbiggi]

LLM1 motor replaced

I deliver the robot to for testing to @AntonioConsilvio

@Fabrizio69 @maggia80

[maggia80]

@AntonioConsilvio we should change the current limit for this motor type MOOG_C2900576_ROTOR (see cub-tech-iit.github.io/documentation/ergoCub1/ergoCub1-joints) to:

 <param name="motorNominalCurrents">             5000                     </param>
        <param name="motorPeakCurrents">        10000               </param>
        <param name="motorOverloadCurrents">    15000           </param>
        <param name="motorPwmLimit">            18000               </param>

[sgiraz]

A PR on robot-configuration with the changes mentioned above will follow soon.

cc @AntonioConsilvio

[SimoneMic]

Hello,

Has this issue been solved?
Can we use the robot in its current state or do we need to take some further precautions?

Thanks!

[AntonioConsilvio]

Hi @SimoneMic, As long as the temperature reading on the robot is not fully functional, the current limits will not change as the lower limits would cause the hip roll i2t HW fault during walking.

So, the usable current limits are still the same (devel).

[sgiraz]

Hi guys,

A new PR on robots-configuration w/ the new current limits, is coming soon.
Once the PR is ready to be reviewed, I would ask you to help us to validate the new limits.

cc @AntonioConsilvio @GiulioRomualdi

[sgiraz]

Hi @GiulioRomualdi,

To definitively close this issue we'd need someone who does the robot walking.
Can you assist @AntonioConsilvio w/ this task?

cc @DanielePucci

[AntonioConsilvio]

Hi! We tested these current limits on ergoCubSN002:

<param name="motorNominalCurrents">             5000                     </param>
       <param name="motorPeakCurrents">        10000               </param>
       <param name="motorOverloadCurrents">    15000           </param>
       <param name="motorPwmLimit">            18000               </param>

The robot works but the joint tracking is less accurate, negatively affecting the walking.

So more studies can be done on this to verify its feasibility, however I am proceeding with the closure of this issue as the burnt-out motor has been successfully replaced!