I changed the configuration of my tool mount yesterday and went through the process of making sure everything I touched was tight and homed correctly. I also took measurements of where my spoil board and standard 3/4" stock heights were. I updated my software with these new values and noticed it was cutting about 7mm deep or 3mm deeper than expected so I repeated the process and adjusted my Zero for the 3mm.
Running the same program then resulted in the bit not touching the worksurface at all. Today I tried again and got a cut but it is short of the full depth it should be cutting by 2-3 mm.
My Z had no issues previously like those above. Can someone point me towards a direction for diagnosing what I must have done wrong?
It sounds like you are getting an inconsistent result with respect to cutting depth. Typically, this is caused by losing steps. It takes more work to lift the Z axis, so usually lost steps happen when the Z axis is lifted, resulting in deeper cutting over time.
There are a variety of reasons for losing Z steps. The grub screws (as Barry suggested) are usually the cause. If you are sure the grub screws are not the issue, take a look at any way the Z axis might be rubbing or binding. Given you wrote, “making sure everything I touched was tight,” examine anything that you changed. For example, on my machine, the clearance of the router is very tight and requires precise rotation. If I’m off, the switch rubs or worse. Or examine the brass t-nut for the lead screw. It should be almost floating. Tightening the t-nut down could cause the lead screw to bind. If you have not done it recently, lube the lead screw. If you are sure the router is not impinging on the core in any way, pull it off the machine and move the lead screw by hand to check for how easy or hard the screw is to turn over the range of movement.
There are other reasons the Z axis can lose steps. They are unlikely given your scenario of having a working machine with the only change being tightening things up. Things like wiring issues, stepper current issues, feedrate issues, etc. But if you don’t find any binding or rubbing and the problems continue to happen, post back and we can explore other possibilities in depth.
Thank you for the info but I notice the Z depth is off immediately after starting the machine.
IE, It homes XYZ and then travels to it’s point of origin where it either digs into the stock or is not close to it. It then begins the job at that height and the rest of the cuts are evenly too high or too low.
I do not see anything rubbing or binding, all movement is smooth on the Z axis when moving it with the controller or manually when the steppers are disabled.
If your error is repeatable and “wrong,” this implies that your g-code is sending your bit to the “wrong position.” To sort this out, can you provide:
How you are authoring your g-code file.
How you do the initial setup of your machine.
Attach the g-code file to a post so we can take a look at it.
That’s where I was thinking as well. Fusion 360 is creating the Gcode using the guffy1234 Marlin 2.0 Post processor.
The initial setup of the machine is done in Octoprints GCODE Scripts, Before print job starts
G28 Z X Y ; Home @TPLINKON 192.xxx.x.xxx (dust extractor) @TPLINKON 192.xxx.x.xxx (Spindle)
G0 X8 Y440 Z50 ; 1" 3/4" mdf, 2.5" bit
M140 S0 ; Set bed heat to 0 degrees
Attached is the Gcode file that is giving me problems. Thank you. 2 800 x 300.gcode (1.1 MB)
This line establishes that wherever the tip of your bit is when this file is run is the Z=0 position. The first pass of cutting is to a depth of -4mm, and the file continues to make passes until it reaches a depth of -19.3mm (a bit over 3/4").
Looking at your Octoprint g-code, this line puzzles me:
G0 X8 Y440 Z50 ; 1" 3/4" mdf, 2.5" bit
How is the 50 derived? It doesn’t match any calculation with the bit and the stock. And in typical setups, the length of the bit with vary depending on how far it is inserted in the collet, so I’m not sure how you can include a fixed number. I suspect this is the root of your issue.
When I’m cutting I either use a touch plate to establish the Z height of the top of the stock, or I use a touch plate to establish the spoil board height. If I do the second, I author the CAM using a bottom corner in Fusion 360.
The 50 comes from manually jogging the bit to the top of the start position as all my stock remains the same size. I do this again any time I mess with the collet.
The 50 is a strange number that does not make sense to me. It is a bit under 2" so that means you are lifting your router almost 2" from the Home Z position. Is the home Z position roughly 1.25" below the level of the surface of your spoil board? Do you have to reedit the “50” every time you change a router bit? If this is your procedure, there are approaches to setting up your job that don’t involve you editing the 50.
Assuming the 50 is not at the root of your issues, and that there are no mechanical issues, then the next thing to check would be the Z homing. For example, I’ve seen electronic noise interfere with homing. Gut your g-code file to remove all cutting moves, then run your sequence a number of times to see if the bit is just at the top of your spoil board when you run your Octoprint g-code and your file g-code.
After noticing the tool is no longer dropping while testing parts of the start script and gcode file. The machine is running as I would like it to. Maybe some type of interference but I won’t know till it decides to happen again. Thanks all for your help.