TODO. I think the acceleration settings need to be adjusted down as I'm getting some ringing when transitioning from X to Y moves.

The testing below I arrived at the following:

• X Steps per Inch: 2037
• Y Steps per Inch: 2040
• Z Steps per Inch: 5080
• X Backlash: 0.0075
• Y Backlash: 0.002
• Z Backlash: 0
• X Acceleration: TBD
• Y Acceleration: TBD

The defaults I found where 2038 for X and Y and 5080 for Z. 0 for backlash.

I found a 3 steel 5 sided box box under the welding table. I used a large kreg C clamp to clamp it down to a clean part of the spoil board at the front right and it was surprisingly sturdy.

This will be the platform in which I use digital calipers to measure distance traveled over 5 inches with an accuracy of 0.0005“ And a dial indicator where you can inference down to 0.0001”

• Mounted the box with the open side to the front
• Mounted calipers off the left side of the box with a spring clamp clamping the bottom jaw to the box allowing the top jaw to move freely.
• Extended the caliper to near it's full extent.
• Jog the router so that the mounting plate is to the left of the calipers and would push the calipers closed if moved to the right.
• Slowly jog the spindle to the right with the pendant until it starts to close the calipers.
• Zero X axis and zero calipers.
• go to MDI so that you can enter Gcode.
• G1 X-1 F80 Move the machine to X -1 at 80 IPM
• Extend the calipers to near it's full extent again so that it's reading >0
• G1 X0.0000 F30 Move the machine back to 0 slowly. I noticed that if I didn't specify 0.0000 it would be ok with stopping at +-0.0005 as reported in Mach 3.
• Verify that the calipers read 0. If there is variation try to fix the measurement rig so that you get consistent results I was able to get it to be consistently at half a mil. 0.0000 - 0.0005
• After you're confident in the setup lets do the x distance measurement.
• G1 X5.0000 F30 This will move the spindle to the right and close the calipers by roughly 5 inches.
• Record the measurement ie 5.005 so it moved 5 mil more than it should have. Do the test a few times to make sure you are getting consistent results.
• Go to the config menu in mach 3 and pick motor tuning.
• Look at the steps per inch in the bottom left of the window 2038
• Determine the distance per step 5.005 / 2038 = 0.0024558390578999
• Determine the number of steps to travel desired distance 5 / 0.0024558390578999 = 2035.964035964038. Round that to a whole number, 2036.
• Update the steps per inch in Mach3 and test again. However I found that this number over corrected and needed to land on 2037 for best results.

Backlash is the term for the slack in the gear teeth, or belts when changing direction. I.E the difference measured when when travelling to a point when going to that point from the left and from the right. To test this we will use a dial indicator. It would be possible but tricky to do it with calipers.

Attach Magnetic indicator arm to the steel box. and arrange it so that the spindle mount will be able to depress the indicator probe when moved to the right.

• Use the pendant to slowly jog the spindle to the right until the indicator has started to be depressed
• Zero x axis and zero the indicator
• Now test for repeatability by going to -1 and to 0 a few times
  • G1 X-1 F80
  • G1 X0.0000 F20
  • If there is variability of more than 0.002 try to see if you can find the cause. It may be in the measurement rig.
• Move to the right, but don't crash the indicator. then back to zero from the other direction.
  • G1 X0.3000 F20 move 3/10 of an inch.
  • G1 X0.0000 F20 move back to zero and observe how far it is from 0
  • repeat the sequence. to to -1. then 0 then .3 then 0 and make sure all numbers are consistent.
    • G1 X-1 F80
    • G1 X0.0000 F20
    • G1 X0.3000 F20
    • G1 X0.0000 F20
• I observed a backlash of 0.0075“
• Under config open Backlash. and put the measured difference into X backlash and check the enable option.
• It says a restart is needed so click on config, save settings. close and re open mach 3, reset, and try the test above and see if it corrects for the backlash.

I used the same methodology as above I just needed to re-clamp the steel box so the open side was to the left.

For this one with the limited move I used the indicator to measure distance too but the machine was spot on for both travel and for backlash I didn't need to change anything.

I've seen ringing a few times on profile cuts on direction changes. In the motor tuning section there is the ability to lower the acceleration settings I still need to come up with a method to tune that. would like to be smarter than doing a bunch of test cuts but I'm not sure.