I am getting slight distortion (approx 1mm) in some 70mm circles and the circles a very slightly “rippled” around the edges. Cut was in timber (Paulownia [think balsa wood], with a 2.4mm straight bit, 2mm DOC & 8m/sec (from memory). Arc was enabled and disabled on two seperate cuts. The results were consistent with both cuts. Cuts run through Repetier.
The original crown test and test cuts were great, until I realised they were 50% small on X/Y and 80% shallower on X (steps/min settings). After correcting these, I had issues with Z acceleration speed causing the bit to consecutively cut deeper.
There is a slight angle to the sides of the pieces, which I attribute to the router not being perpendicular to the x/y axis.
Following advice in other threads I have checked:
Pulley grub screws (one was slightly loose, have corrected)
X/Y square (within approx 0.5-1mm)
Cursory inspection of tube / pulleys (no obvious flaws/issues).
During the process of elimination of the earlier faults, I made the following adjustments:
Tightened all the belts
Tightened the 130mm bolt on gantry
The machine was relocated to a new (permanent) table, but I spent the time getting the leg heights and squareness sorted.
Once I get home from work I am going to:
Replace the tension cable ties on the belts
Loosen off the 130mm bolt on gantry
Try to correct slight angle in router/bit
Install pen again, and rerun the Crown test
Try running the gcode from the TFT, just to eliminate the possibility of a bad USB connection.
I believe the fault is mechanical rather than gcode, however gcode is attached below.
Is there anywhere else I should be focus on if the adjustments above don’t work out?
Custom build:
SKR V1.3 with DRV8825 drivers running firmware from -
No dual end stops, but it may still be enabled in the firmware.
Can you upload a pic of the work and one of the machine?
If you started from dual endstop firmware, the soft stops might still be enabled, which would limit the work to only positive numbers in X and Y. The gcode looks like everything is positive though.
From the gcode, these first few lines look dangerous:
That will move the Z axis down to zero at 35mm/s. Which is too fast for the Z.
Later, the speeds on Z are F240, which is 4mm/s, which is good. Your plunge rate is 8mm/s.
The XY speeds are actually F900, which is 15mm/s, not 8mm/s, like you said in your post. The DOC is also 4mm, and you said it was 2mm. So maybe you’re just not setting the right things?
Bottom right was first, I adjusted some belt tensions and also noticed that even though my XY on gantry was square, the gantry was twisted on the frame, so there was a larger gap between the leg on one side than the other. I managed to jog that back into square then ran the second image.
On the new images no, I ended up having to do up a new gcode. My pen tool had a xy speed of 0.3mm/s and when i tried to update gcode it wouldn’t change the speed for some reason. I am going to attach the mill again tomorrow and try out both the test.gcode and the new file.
I am wondering if a poor connection between Repetier and the board may have been causing some of the issues on the cuts. I tried to use Repetier-Host today and it came up with communication timeout. I was using Repetier-Server (web) previously.
