Auto-Calibration Issue - Y steps not updating

[PannDemic]

So I just upgraded my heated bed, installed a ground flat aluminum heat spreader under my glass, and ran the auto calibration script. However on my first test print, at the Y tower the nozzle is nearly scraping the glass, the plastic was so thin I am pretty sure it did not actually extrude anything. On all of the other towers it was fine, maybe a bit thick at Y'.

I checked my EEProm settings and noticed that my Y end-stop steps was set to zero. I re-ran the calibration and still 0. So I decided to trick it, and placed a 0.1mm feeler gauge where the nozzle would strike at the y tower. Ran the code, still 0. It did mess with the steps for x and z though.

I am not sure what to do next for this issue. If I have anything with a bit of reach towards the Y tower it is nearly impossible to print.

Thanks

[PannDemic]

Is there some additional information I am failing to provide that may help me sort out the issue?
I am using the latest version of Repitier from the SeeMeCNC Github, and I am totally baffled at what it could be.
I am not really a programmer, so I am not sure if I could identify the issue if it is firmware related.

[DeltaCon]

What are your other endstop values?
Do the other endstop values change when using the feeler gauge on the Y tower?
Start calibrating endstops by hand using the endstop screws. You should be able to almost level them at 0.
I am not familiar with the V3 calibration script, but could it be that the Y tower is reference 0 to set the X and Z tower endstops?
Make marks at the X tower on both glas en heatspreader. If you turn the alu and glass 120 degrees, does the problem rotate along with the marks?

[mhackney]

The auto calibration picks one of the endstop offsets and sets it to 0 and adjusts the others appropriately. So it is normal that one of yours is 0.

When this happens, measure the Z height at each of the tower positions manually and tell us what you find.

Also, have you checked your tower perpendicularity as per my V3 build thread, glass flatness (the same) and the build notes that Steve posted mid week? You really need to make sure the bed is seated properly in the bed clamps and that they are all snug but not over tight. When you press at the base of each tower (where the probe will tap) there should be minimal bed flex and what little there is (if you push really hard) should be the same at all 3 spots. With an accelerometer probe, the bed should be as rigid as possible.

[geneb]

I ran it on Otto a bit ago - got a good cal, no low or high spots when printing Steve's calibration object. Kisslicer made a dog's breakfast out of it though - moves all over the place.

g.

[mhackney]

Yeah, not a good model for KISS. I've asked JJ to post their gcode version of the object. Since we are running stock V3s it should be fine on all of our machines.