Forum Post

[OpenArm Troubleshooting] Base Mount Shift: Calibration Reference Mismatch (Intermediate)

OpenArm still homes but fixtures drift 1-2 mm after a base move? Use fast checks to confirm frame shift before production scrap grows.

Forum / Posts Index / OpenArm

Post

A subtle OpenArm precision problem appears after the robot base or mounting plate shifts slightly. The arm still powers up and can home, but task-space references, fixtures, or known poses no longer line up with what the old calibration assumed.

How are you diagnosing OpenArm cases where the base mount moved and the calibration reference no longer matches reality?

Please share how you detect mount shift, what checks distinguish a base-reference problem from encoder or tool issues, and how you decide whether partial recovery is safe or a full recalibration is required.

If you reply, include one exact symptom after the mount change and one exact geometric or fixture-based check that exposed it.

Related troubleshooting path: End-effector Z check fails after table move · Repeatability worse after fast recalibration

Module: OpenArm · Audience: builders-integrators · Type: question

Tags: openarm, base-mount, calibration, reference-frame

Comment 1

Intermediate context: In our cell the first symptom was a clean home but a hard-stop fixture touch in +Y was off by about 1.8 mm. A 3-point fixture touch check exposed it in under 2 minutes, while joint repeatability stayed under 0.05 mm, so encoders were likely fine and the base frame was not.

Comment 2

Intermediate context: We separated base shift from tool drift by swapping end effectors and rerunning the same known-pose cube-corner check. TCP behavior stayed consistent but every waypoint was translated by almost the same vector (about 1.6 mm, mostly +X), which pointed to mount shift rather than tool or wrist calibration.

Comment 3

Intermediate context: Our rule: partial recovery only if a rigid-frame fit on 6 fiducials leaves less than 0.2 mm residual and less than 0.15 degrees yaw error. If not, do full recalibration plus collision-model revalidation. We restored trust with 30 dry cycles on two fixtures before releasing back to production.

Comment 4

Intermediate context follow-up: Follow-up question: after your frame correction, how much residual error remained on your worst fixture point (mm), and did that value stay stable after a full warmup shift?

Quick Symptom Selector

Pick your closest symptom to follow the right troubleshooting path.

Not selected yet.

Quick FAQ

How do I quickly confirm base frame shift vs encoder issue?

Run repeatability first. If joints repeat tightly but fixture points move with a similar vector, suspect base/reference shift.

python tools/check_fixture_points.py --fixture A --points p1,p2,p3
python tools/check_joint_repeatability.py --cycles 20 --joint-all
When is partial recovery acceptable?

Only when residuals stay within tolerance after warmup; otherwise do full recalibration plus collision-model revalidation.

Can I copy these commands as-is?

Use them as a checklist template first. Confirm interface names, fixture IDs, and safety conditions in your own cell before execution.