Centroid Community CNC Support Forum

I have been trying to get screw mapping for the z axis on my mill, the maximum axis physical travel is 510 mm, I have homing configured in such way the axis moves UP (positive) towards home, and therefore the travel limits are set to be (0 & -507)mm.
Up to this point I have no backlash compensation set yet, but I have already measured it and I know the physical backlash is 15 to 20 microns (around half a thou for our imperial friends )
I did the travel measurements for both Forward and Reverse movement and got the tables implemented in CNC12, The results were spot on in terms of travel but there is an issue now with what looks like a massive backlash measured at the dial indicator, after inspecting this more it appears that the software is forcing a movement of about 115 microns (4.7 thous) evry time a direction switch is requested, so I tried to use this number in lash compensation and the effect was doubled so I changed it to become negative lash value and it improved things and became reasonable.
I’m not comfortable leaving this huge lash compensation value as I know the recommendation is to keep it below 2 thous for practical machining reasons.
I researched this a lot here on the forum and the best I found was couple of old documents in this post:
https://centroidcncforum.com/viewtopic. ... omp#p51477
Long story short from what I understood is that for the REVERSE compensation table the home position must have Zero compensation value (right now I have 115 microns which is exactly the massive lash I’m measuring), and to do so I should subtract the correction value measured at home location from all entries in both the forward & reverse table, Is that correct? Now If I do that I’ll end up with non-zero compensation at the home location in the FORWARD table, is that acceptable?
I know I can just go and try that but I don’t want to keep messing with the machine and cause some unexpected moves and damage something, especially that the zero location is right at the home position and can easily travel into the limit switch and damage it.
Below is my current compensation table, I hope one of the experts here can take a look and tell me what to do. I’ll upload the machine report zip file to the thread.

cnc-z.tab file content:

CNC10 TAB FILE VERSION 3
AXIS Z
UNITS MM
INTERVAL 25
ENTRIES 21
+0.00000000 +0.00000000 +0.11500000
-25.00000000 -0.05000000 +0.08500000
-50.00000000 -0.08300000 +0.05400000
-75.00000000 -0.09400000 +0.05000000
-100.00000000 -0.10400000 +0.05200000
-125.00000000 -0.09000000 +0.06200000
-150.00000000 -0.07600000 +0.08000000
-175.00000000 -0.05200000 +0.09100000
-200.00000000 -0.03400000 +0.10900000
-225.00000000 -0.03300000 +0.11500000
-250.00000000 -0.02600000 +0.10900000
-275.00000000 -0.01500000 +0.11600000
-300.00000000 -0.00900000 +0.12100000
-325.00000000 -0.02000000 +0.11500000
-350.00000000 -0.05100000 +0.09700000
-375.00000000 -0.06600000 +0.08300000
-400.00000000 -0.07400000 +0.08300000
-425.00000000 -0.09600000 +0.06300000
-450.00000000 -0.11300000 +0.04500000
-475.00000000 -0.13500000 +0.02100000
-500.00000000 -0.16200000 +0.00100000

and related.
https://centroidcncforum.com/viewtopic.php?f=60&t=3230

cnckeith wrote: ↑Tue Jan 31, 2023 5:30 pm and related.
https://centroidcncforum.com/viewtopic.php?f=60&t=3230

Keith,

Thank you for your quick replies, the two threads you shared are both great ones that I already learned lots from while I was searching this topic, but they still don't address the issue I have, at this moment I want to know if I should sbtract the 115 microns from both Forward and Reverse tables so that the Reverse table has zero adjustment at the home location, and whether it is ok to have a non-zero adjustment in the Forward table at the home location?

Move your axis an inch or 2 and then move it back to zero. Does it come back to zero or is it off by a certain amount with the compensation table populated.