Trouble with Z axis/touch plate calibration

I am not sure where it came from either. When I input the actual thickness and I go to Z0.0 it goes .007" into the spoilboard. If I subtract .007" from the actual thickness and input that then everything is good. I ran a few jobs after and it was dead on. I have no idea why the discrepancy. The calibration seems to be good.

You had mentioned lost steps/motion. How do I go about checking for that?

Since calibrating and setting up the touch plate as mentioned earlier, I have had one instance where when I ran Z0.0 the tip of the end mill stopped just below the spoilboard. I did not measure the distance but by the looks of it it looked like about .007" which is what I had adjusted for in a previous step. This happened after I HOMED the machine and restarted LCNC.

After I touched off the Z axis again, everything was fine. I am still doing research trying to see if I can find a cause or a way to reproduce. I wanted to check to see if there are any missed steps/motion. I found in my research to check some settings in the INI file. Not exactly sure what looks good or bad or suspect. Was hoping maybe someone can look over my Z-axis settings (below) to see if anything is questionable?

My setup:

• Nema23 on a leadscrew/linear rail Z-axis.
• DM860T drivers
  • Current: 
    • REF: 3.14A
    • Peak: 3.77A
  • Pulse:
    • Pulse/rev: 2000

The gantry and Z-axis is solid with no slop/movement. The Z-axis seems to run very smooth with no grinding or odd sounds. There is no visible signs of missed steps in its movement either. The one thing I have always noticed since I got this thing running is that the Z-axis seems to run slower than I would expect (same max speed both up and down). The MAX_ACCEL is set to 50 but it does not seem to move anywhere near that. I would guess that it maxes out at about 20 or so. Not sure if this is an issue or not.

[AXIS_Z]
MAX_VELOCITY = 0.5
MAX_ACCELERATION = 50.0
MIN_LIMIT = -5.251
MAX_LIMIT = 0.001

[JOINT_3]
TYPE = LINEAR
HOME = -0.2
FERROR = 0.5
MIN_FERROR = 0.05
MAX_VELOCITY = 0.5
MAX_ACCELERATION = 50.0
# The values below should be 25% larger than MAX_VELOCITY and MAX_ACCELERATION
# If using BACKLASH compensation STEPGEN_MAXACCEL should be 100% larger.
STEPGEN_MAXVEL = 0.62
STEPGEN_MAXACCEL = 62.5
P = 500
I = 0
D = 0
FF0 = 0
FF1 = 1
FF2 = 0
BIAS = 0
DEADBAND = 0
MAX_OUTPUT = 0
# these are in nanoseconds
DIRSETUP   = 8000
DIRHOLD    = 5000
STEPLEN    = 2500
STEPSPACE  = 2500
STEP_SCALE = 10178.175
MIN_LIMIT = -5.251
MAX_LIMIT = 0.001
HOME_OFFSET = -0.100000
HOME_SEARCH_VEL = 0.500000
HOME_LATCH_VEL = -0.033333
HOME_FINAL_VEL = 0.000000
HOME_USE_INDEX = NO
HOME_IGNORE_LIMITS = YES
HOME_SEQUENCE = 0
#******************************************

So I spent quite a bit of time today trying to figure out a pattern to my Z-axis issue. I thought I had something but it is inconsistent. So far every time that it seems to lose its settings is after a complete restart of the CNC and LCNC. But not every time. Here are a couple examples of what I observed today.

I had touched off the Z-axis and everything was spot on. I did a complete restart of everything. I immediately ran Z0.0 in the MDI and the end mill stopped .01" from the spoilboard. I did a Z touch off again and everything was spot on again. 

I then tried different things like just restarting just LCNC (multiple times) but nothing changed. I did a bunch of complete restarts but nothing changed. Also in between every restart I moved all axes around at different speeds as well as at the same time just in case it had something to do with too much power being drawn from the PS. None of these things made any difference. Then out of nowhere I did a complete restart and ran Z0.0 again and the endmill dove into the spoilboard .018". I then did a touch off again and everything was good again. 

I am not sure what to do at this point. I was hoping I could find a pattern. 

Any thoughts?