Zero Position and Quick Master

Hello folks,

Some time ago I used zero position mastering on one of my robots without completely understanding how it works.
This was done on a R-30iB/M-1A.

After realizing I should have used quick master (I have the inspection data sheet) I started over by: entering the reference position counts for all joints from the IDS, setting $REF_DONE to TRUE, $MASTER_DONE to FALSE, moved to the scribe marks, then used quick master and calibrate.

My problem now is that when I jog all the joints to their 0 position, joint 4 is off by about 6.5mm.
Did my use of zero position master affect the reference or encoder counts in any way? Have I effectively rendered the $MASTER_COUNT on the IDS useless?

Quote from zieterbock

joint 4 is off by about 6.5mm.

Correction, joint 4 is off by 6.5 degrees. My mistake.

Quote from skalactik

tried changing the $DMR.$MASTER_COUN values manually with the ones on the data sheet ?

Yes, same result as just using quick master.

Quote from USURP_RUR

Quick Master is for integrators.

I am an integrator.

Out of the eight robots I accidentally used ZERO POSITION MASTER on, then corrected, only one is giving me problems.

Other than using SINGLE AXIS MASTER, I'm not sure how to proceed.
Would like some input before doing something like INIT START or reloading software.

I do not have a backup.

Robots had arrived from FANUC and were installed into a frame, they were without power for some time.
I assumed the encoders lost their counts during this time, thus the need for remastering.

Currently what I've done and not liking it; out of hundreds of robots, only one is zero position mastered haha.

Working with FANUC to find out why this happened since it appears to be an encoder issue originating from their end.

Will post here if a solution or explanation is found.

Thanks for the feedback, guys.