Y Axis G54 Location Moving Troubles

I'm currently running Axis 2.7.0 and have recently started to have an issue where while machining a pocket my G54 is moving in the Y direction each pass I step down. It's very consistent and is about .010" per pass. I've ran this part before and didn't have any issues.

I program with Fusion360 and reprogrammed and posted the gcode and still have the issue. I'm not sure if this might be a cumulative error although I don't believe so as it's only in one axis and one direction. The other two aren't stepping. Anyone have any ideas where to troubleshoot. At first I thought it was the motor so I replaced it with a new one (Nema 42 with a larger holding strength) and it still has same issue. I also went through my ball screws and now have less than .001" lash in each. This leads me to believe it's something on the software/computer side and not mechanical.

Also using Geck g201x's for all my drives. Any help would be huge as I'm dead in the water til I sort this out. The last pocket I made was 1.25" deep and is supposed to be .700" wide and ends up being .760" wide at top and down to correct size at bottom.

I had a thought and I wanted to mention it in case someone has an idea if it may be the case. I've noticed in previous projects I sometimes get an arc error (start - finish location off). For this pocket I did Fusion360's Adaptive Clearing which runs the fishtail arcs over and over again. Could it be possible that this is why I'm getting the error? Thought being if it is getting error I could see it "losing place" although the part of this idea that doesn't seem logical is that it should know that it's off so G54 shouldn't change. Just a thought.

this movement appears to be consistant , and as you say it also appears to happen at the stepdown .
so this can be a number of things , from missing steps to a faulty ballscrew , and could also be attributed to running the axis at too high a velocity , so try reducing say by 10% . another try is to reduce microstepping if you have it higher than half step .
also make sure your power supply can supply enough current to satisfy all drives maximum amps i.e
if each stepper has 2 windings that say take 2A holding that's 4a per driver x 3 ( xyz) =12A , and that your stepper driver current setting is correct ,

sorry not to be able to pin this further , but if you try the above it will soon be obvious .
I doubt it's gcode , unless you have an errant setting in your tooltable , make sure you set or reset your tooloffset to 0
for testing

At first I thought it was the motor so I replaced it with a new one (Nema 42 with a larger holding strength) and it still has same issue. I also went through my ball screws and now have less than .001" lash in each.

You might try setting up a dial indicator on the problem axis at a known location and reading it before and after running the part to see if in fact steps have been lost.

You stated you ran the part before without problems, are you running the same file or did you create a new one?

Does the display of the part on linuxcnc look correct?

Rick G

After many many frustrating hours I've now got it running correctly. I'm not sure why but my Y axis geckodrive was the culprit. I had another installed for a 4th axis I'm building and I suspected it might be that or breakout board so swapped wiring and it now works. I'll be calling gecko to see if they have any ideas. Thanks for any info provided.