When I run this the cnc stops when touching the touch plate and does nothing anymore, it freezes. I have to reboot to get access again. I checked with Gemini, and it adviced me to use the following instead of the G38.2 line:
G38.2 Z-110 F20
G92 Z0.5
Now this does work. So I was wondering how come using the “P” command makes it freeze? I have recently installed the jackpot v3 with the latest fluidnc firmware and lowrider settings. Any idea what could be the cause? Also probing with the WEBUI works normally.
It touches the plate and pauses waiting for your input to resume from the webui, that is normal. At the very top you should see a play/resume button. This is to give you time to remove the probe before it moves again.
As mentioned, it’s just waiting for you to attach the probe and resume. I thought I’d add a bit more explanation.
The P parameter of the G38.2 probe is what defines the thickness for the touchplate. The V1E touchplate is 0.5mm thick. When this parameter is included, it sets the Z zero position.
This will technically be slightly off. When the probe is triggered, the machine still needs to decelerate. So, it will actually move down a bit further.
I was also using the F200 feedrate. But had the impression that everytime the endmill touches the touchplate I could hear a little crack on new endmills. The very tip of the single flutes broke off. I’m now down to F20 and so far this seems to work fine.
Wondering whether anyone had similar experiences…
Ok, I was not clear, the freeze is after it touches the touch plate. So it is stuck between the workpiece and the bit. It does not go up again, it is not waiting on a pause. As soon as the bit touches the plate it is unresponsive. It is also not waiting for input, as the buttons are still showing play.
The pause should be after G1 Z10 F900 but it does not execute this… I’ll try to capture the logging later, but it did not show a clear error to me.
I’m coming from a Marlin perspective, but this is normal after a G38.2 Z(n) command. The unit waits (pauses) until the next command is issued. The next commands are usually G92 Z(thickness of probe), followed by G0 Z(n), where n is the height you want to raise the bit away from the stock after probing.
I usually enter all of these commands individually, not from the project g-code, but my workflow is:
Home X, Y & Z individually from the GUI (automatically sets X=0, Y=0, Z=200)
Move X and Y from the GUI so that the bit is above the probe location (I don’t probe at (0,0), usually at (10,10)
Attach the probe
G38.2 Z0 ; (probe Z in the direction of 0)
G92 Z0.5 ; (set Z = thickness of probe)
Move Z up 10mm from GUI (Or G0 Z10 ; (move Z 10mm above surface))
Remove probe
Start g-code from SD card
Some of the commands are slightly different in GRBL, as you are supposed to be able to combine the G38.2 and G92 commands into one command line (I think G38.2 Z0 P0.5), but I’m not a GRBL expert).
Maybe try running the individual commands from your starting g-code from the GUI instead, at least as a troubleshooting method, then put them back into your g-code after you get things working.
thanks, but the next line is already G1 Z10 F900 As you can see in my original post. The thing is that it does not execute this line anymore. It becomes completely irresponsive. I have to reboot to be able to do anything again.
the suggested troubleshoot let me to this question. When I remove the “p” command in the G38.2 command it works fine.
also as Jason said, the “p” command is more precise than G92 Z0.5
if I find some time this evening i will to execute the command separately in the command line and also try to add a short pause after the G38.2 command. I will also try to capture the output log which hopefully helps.
What version of FluidNC are you running on the Jackpot 1? What version of FluidDial?
If you are having to restart the controller to get out of a freeze, then something in the P command is corrupting the controller. Whether that is in the gcode, in the controller firmware, or in FluidDial is the question to be determined.
But a simple work-around is to not use the P command, and use G92 instead.
I’m running FluidNC v4.0.1 with WebUI 3.0.0.6 Fluidnc on the Jackpot v3. I have not used the fluiddial yet for the probing. I only probed and started projects with the WebUI. Also I use the config file 11-01-2025 RyanZ. To my knowledge these are all the latest versions. Correct me if I’m wrong.
BTW I used to run the cnc with marlin with the SKR pro 1.2 but I did not like the changing of the settings etc. And wanted to switch to fluidnc, as this is were most of the new development is done.
Also, indeed I can use the work-around fine. However, if I do not know why this fails, who is to say what other commands will fail next when I’m in the middle of a real project.
Ok problem solved! I installed 3.9.9 and it is working correctly now!!! Thanks @vicious1
I initially installed throught the webinstaller and just choose for the latest version.
Now we are on the topic of starting scrips, i was wondering why the standard starting script retracts only 10 mm after probing (Z10). This does not give me enough space to easily place the dust shoe. Or am I doing someting wrong? I changed it to Z20 and this solves it for me.
