0M coordinate rotation- something to watch for...

[tc429]

Today had a little toolchanger issue...a repaired/reloaded machine had ran toolchanger test cycle all day, they loaded up a part program, ran it thru several times fine...these are little horizontals with the plastic tool pots/axis motion to put away/pick...

(I'd rewrote the macro long ago so it would allow for bent tool racks, as the original program used incremental steps on x, same y/z for all the tools... a minor crash would bend the rack a bit and one end or the other would always drop tools until the steel was fixed...just wrote one where you punched in first and last tools, macro divided up the differences and kinda followed a off axis line for the bent rack, has worked fine for a long time, fabricators finally got a break from always having to cut/tweak the mounts...)

Anyways:

they ran the part program partially thru, stopped/reset it, homed it up, loaded a part hit start and wrecked a toolchange pot...

looked and looked (I'd modified the t/c macro, assumed I'd messed something up), couldnt find anything...single blocked thru fine, so ran the part program/stopped it, reset/restarted, it tried to go to the wrong position again- I'm talking over 2 feet with only a 11" start dimension, was supposed to only move about .08"

looking thru the program, noticed he had a G55, then used coordinate rotation to orient the toolpath, THEN called the tool...and the tool sub runs G54, and I didnt put a safe start block in there that cancelled rotation...I never realized till today that (at least with this edition firmware) that rotation active and calling another coordinate system really screws with the math...recalling G55 after G54 still moved the rotated position even farther away...kinda compound rotation?

todays lesson, always put a G69 the block before calling G54~G59...even if the last part program ran didnt finish and cancel the coordinate rotation, resetting/homing, even running another program wont clear it... funny thing is, the macro I edited was one I wrote 14 years ago to the day, and had never had a issue before... needless to say its got a safe start block including rotation cancel now...

anyone else ever seen issues like this? I know some of our early firmware 6 and 0 lathes had issues with stacking offsets, but I'd never seen coordinate rotation issues before- and weve probably had 30-50 of these machines for a couple decades...