II have been reading a bunch of threads where users are experiencing differences in X and Y cut lengths, or there are differences in size between inside (hole) cuts and outside (part) cuts with the same specified dimension.
I’ve compiled suggestions from several different threads into a single list for calibrating my own machine (LR3), with a goal of minimizing those differences and having as accurate a machine as possible.
I want to run this list by the Forum for a sanity check to make sure that I’m not missing anything. Hopefully this can also be a resource for others wanting to fully calibrate their machine.
Comments and feedback are appreciated.
BTW, I’m running Marlin on an SKR v1.2 Pro controller. The process may vary slightly for Jackpot/FluidNC machines.
Prerequisite: Adjust belt tension, ensure core bearings are contacting gantry rails (no core looseness)
Step 1: Square the machine
- Home the X and Y axis
- Move the machine in a rectangular pattern (as large as possible)
- Use a V-bit to create marks on painters tape at each corner of the rectangle (lower Z until the V-bit just touches the tape)
- Measure the diagonals of the rectangle with a steel tape and compare for different lengths
- Adjust the Y endstops as necessary (my endstops have threaded adjustment screws). If no adjustment screws, use M666 Y(n) to compensate for Y offset
- Repeat above steps until the diagonals are within +/- 0.5 mm (pretty much the measurement granularity using a steel tape measure)
- Save settings with M500
Step 2: Level the gantry
- Home Z axis
- Measure the distance between the lower gantry rail and the spoil board at both XMin and XMax (I have some hole gauges that are perfect for this)
- Adjust one Z endstop to get the difference between the measurements as close as possible (repeat previous 2 steps as required)
- Home Z, then probe using the Touch Probe (G38.2 Z0) to set Z=0 at XMin (G92 Z0.5).
- Move the core to XMax and repeat the probe (G38.2 Z-10), but do not use the G92 command
- Read the Z position on the display. This will show the offset difference.
- Use M666 Z(n) to compensate for a Z Offset between XMin and XMax
- Re-home Z and repeat measurements/adjustments to ensure the offset is as low as possible (< +/- 0.2mm).
- Save settings with M500
Step 3: Tram the router
- Attach Ryans’s printed Tramming Tool to the router collet
- Move the tool so that the pointer is pointing toward YMax.
- Move the Z axis so the tool is almost touching the spoil board.
- Rotate the tool 360 degrees, making note of where the pointer is closer and further away from the spoil board
- Remove the router from the mount and place a few layers of painters tape on the bottom mount opposite the lowest point (or on the top mount in the location of the highest point)
- Reinstall the router and retest
- Repeat until the tool can rotate fully with minimal variation in distance from the spoil board
Step 4: Measure belt spacing (Reportedly some belts from V1E have very minor errors in tooth spacing, resulting in slight errors over long distances.)
- Using a V-Bit, make marks in painters tape at YMin and YMax.
- Note the Y position at each location (ie. YMin = 0, YMax = 2250mm)
- Measure the actual distance between the marks using a steel tape
- If difference is >0.5mm, it may be necessary to change the steps/mm for that axis using M92
- To calculate required steps/mm: New Steps/mm = (Old Steps/mm (default 100) * Expected Length)/Measured Length
- For example, YMax = 2250, measured length = 2247. (n) = 100*2250/2247 = 100.1335
- Set new steps/mm using M92 Y(n)
- Retest to confirm correct measured distance
- Save settings using M500
- Repeat test/adjustment for X axis if required
- As per @Dreyfus and @CesarH , recheck square if steps are modified
Step 5: Perform test cuts
- Using appropriate feeds and speeds for the test material (MDF or plywood), create two square cuts of 100mm per side in EstlCAM (one as a hole, one as a part) using Climb Milling. Use a Finishing Pass of approximately 10% of bit diameter. The squares should be oriented such that the edges are parallel to the X and Y axis
- Create a second pair of identical cuts, this time using Conventional Milling.
- Perform the cuts on a test piece of MDF or Plywood (thinner material and a single cut preferred - use holding tabs).
- Measure the distance of both the hole and part cuts along both the X and Y axis.
- Distances should be identical for all measured lengths
- If the hole and part measurements are different, you may need to adjust the Tool Diameter in EstlCAM to compensate for the cutting width of the bit varying from the nominal width (for example 1/8" bit may be cutting at 3.120mm instead of the expected 3.175mm)
- If there is a difference between Climb and Conventional cuts, that may indicate excessive tool flex (reduce cutting depth or decrease stick-out) or looseness in the core bearings.
Please let me know if I have missed anything, or have made any mistakes in the above procedure.
