Advanced Search

Yes, now it works fine.

Hello,

I have completely reinstalled LinuxCNC 2.9.3 and would like to work with QD Dragon in the future. Previously I had used LinuxCNC 2.9.2 with ProbeBasic, which worked quite well.
The machine works with LinuxCNC 2.9.3 and QTDragon_Hd and the first workpieces have been machined. Unfortunately the functions GoHome, ToolSensor and Go to Sensor are causing me problems, see log files.

It would be nice if someone could help me with troubleshooting or explain where I am making the mistake

Best regards from Germany / Lübeck

Translated with DeepL.com (free version)
--- QtVCP Screen Started on: Mi, Sep 11 2024 10:31:35 ---
--- QtDragon_hd Version 1.5 on Linuxcnc 2.9.3 ---
Mi11 10:31: Machine OFF
Mi11 10:31: Hard limits tripped
Mi11 10:31: Tool 0: No Tool
Mi11 10:31: Machine ON
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:31: Hard limits tripped
Mi11 10:33: Bewegung Linear in Zeile 0 würde von Achse/Gelenk Z den Anschlag Negativ überschreiten
Mi11 10:33: Bewegung Linear in Zeile 0 würde neg. Anschlag von Achse/Gelenk 2 überschreiten
Mi11 10:33: ungültige Parameter in linearem Kommando
Mi11 10:33: Bewegung Linear in Zeile 0 würde von Achse/Gelenk Z den Anschlag Negativ überschreiten
Mi11 10:33: Bewegung Linear in Zeile 0 würde neg. Anschlag von Achse/Gelenk 2 überschreiten
Mi11 10:33: ungültige Parameter in linearem Kommando
Mi11 10:33: Cannot touchoff with no tool loaded
Mi11 10:33: Loaded tool 1
Mi11 10:33: Tool 1: 3mm end mill Sample Tool
Mi11 10:34: Touchoff to sensor started
Mi11 10:34: Touchoff routine: Touchoff failed: 1st Probe down failed: Tastspitze schon ausgelöst bei Start durch G38.2 oder G38.3
Mi11 10:34: Bewegung Linear in Zeile 0 würde von Achse/Gelenk Z den Anschlag Negativ überschreiten
Mi11 10:34: Bewegung Linear in Zeile 0 würde neg. Anschlag von Achse/Gelenk 2 überschreiten
Mi11 10:34: ungültige Parameter in linearem Kommando

---

2024-09-11 10:34:03,709 - DEFAULT - INFO - Logging to: /home/rainer/linuxcnc/configs/QTDragon/qtdragon.log
2024-09-11 10:34:03,721 - DEFAULT.QTVCP.QT_ISTAT - WARNING - filter Extension Parsing Error: 'NoneType' object is not iterable
 Using Default: ALL (*)
2024-09-11 10:34:03,721 - DEFAULT.QTVCP.QT_ISTAT - WARNING - Valid Extension Parsing Error: 'NoneType' object is not iterable
 Using Default: *
2024-09-11 10:34:03,722 - DEFAULT.QTVCP.QT_ISTAT - DEBUG - Machine is METRIC based. unit Conversion constant=0.03937007874015748
2024-09-11 10:34:03,722 - DEFAULT.QTVCP.QT_ISTAT - DEBUG - TRAJ COORDINATES: XYZ
2024-09-11 10:34:03,722 - DEFAULT.QTVCP.QT_ISTAT - DEBUG - DEFAULT_LINEAR_VELOCITY = 360.0
2024-09-11 10:34:03,722 - DEFAULT.QTVCP.QT_ISTAT - DEBUG - MIN_LINEAR_VELOCITY = 30.0
2024-09-11 10:34:03,722 - DEFAULT.QTVCP.QT_ISTAT - DEBUG - MAX_LINEAR_VELOCITY = 4000.0000200000004
2024-09-11 10:34:03,888 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: M70
2024-09-11 10:34:03,908 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: G49
2024-09-11 10:34:03,928 - DEFAULT.__MAIN__ - DEBUG - COMMAND=
2024-09-11 10:34:03,929 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: G10 L20 P0 Z0
2024-09-11 10:34:03,949 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: G91
2024-09-11 10:34:03,969 - DEFAULT.__MAIN__ - DEBUG - MDI_WAIT_COMMAND= , maxt = 30
2024-09-11 10:34:03,969 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: G38.2 Z-10.0 F40.0
2024-09-11 10:34:04,170 - DEFAULT.__MAIN__ - DEBUG - MDI error= Tastspitze schon ausgelöst bei Start durch G38.2 oder G38.3
2024-09-11 10:34:04,170 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: G90
2024-09-11 10:34:04,190 - DEFAULT.QTVCP.QT_ACTION - DEBUG - CALL_MDI Command: M72
2024-09-11 10:34:10,493 - QTvcp.QTVCP.QT_ACTION - ERROR - CALL_MDI_WAIT Error: Bewegung Linear in Zeile 0 würde von Achse/Gelenk Z den Anschlag Negativ überschreiten





Best regards from Germany / Lübeck

Translated with DeepL.com (free version)               

Ich denke auch es ist ein EMV Problem….
Ist die Schrimung des Kabels auf Erde gelegt bzw. Masse ?
Die letzten 60 cm sollten auch geschirmt sein .
Auch könne ein Netzteil das Problem sein Schaltnetzteile arbeiten mit Frequenzen.
Oder ist es ein Längen Problem des Kabels.
Die Schirmung würde ich als erstes überprüfen die Letzten 60 cm Schirmen.

MfG
Juergen

juliankoenig87 wrote:

Wenn es bei geringen Geschwindigkeiten über größere Verfahrwege funktioniert, dann passen ja die scales.

Hast du mal die Absolutwefte gescheckt, wenn du höhere Geschwindigkeiten nutzt?

Ich vermute jetzt erstmal ein EMV Thema.

Woran sind die GMS angeschlossen? Mesa vielleicht? Und sind das differentielle und mit geschirmten Leitungen?

Julian
 

Hallo Julian und danke für deine Antwort.
Die GMS sind an einer Mesa 7i85 (ohne S) angeschlossen. Die Kabel sind die Originalkabel von den GMS incl. DB9 Stecker die ab Werk dran sind, also alles geschirmt, nur die letzten 60cm im Schaltschrank nicht, aber die Leitungen sind getrennt von den Netzteilen und den 230V Netzkabeln und eine Hochfrequente Spindel oder Frequenzumrichter gibts an der Maschine nicht, deswegen denke ich, dass das kein EMV-Problem ist.

Welche Absolutwerte und wo gecheckt?

Hier mal ein Screenshot vom HAL-Scope wenn der Fehler im Eilgang kommt:
 

Gruß
Olli

According to the log your graphics card is the reason why Probe Basic doesn't start.

Probe Basic requires a OpenGL 3.2 and pixel shader 1.5 capable graphics card.2024-09-10 16:14:08.613 ( 5.246s) [ A8C6040] vtkShaderProgram.cxx:438 ERR| vtkShaderProgram (0x86b5790): 0:1(10): error: GLSL 1.50 is not supported. Supported versions are: 1.10, 1.20, and 1.00 ES

/usr/bin/linuxcnc: line 974: 4169 Segmentation fault $EMCDISPLAY -ini "$INIFILE" $EMCDISPLAYARGS $EXTRA_ARGS
4132
4165
Stopping realtime threads
Unloading hal components
Note: Using POSIX realtime

i think what they referred to as follow error is what i was looking at as an indicator to test with if that value is set high then i can set it low to force thing to kick an alarm. that would show me which drive is doing the most correction i think so i can focus on the task in the right place and basically narrow things down some.

on small bench top or desk top size machines i have had good luck with 283 oz/in motors and those are what i have all around on this build. i was looking at bigger ones and if it would actually be an issue with 1 axis like the X i could go up to a 425 oz/in and keep the same size saft and frame size. so that could be a work around depending on where things land with testing. the drives are even the same too.

if this were open loop things would be more out in the open as far as lost steps are concerned.

added edit here. i did see about setting the dip switched to default and setting the micro stepping in the drive. that was said in the one video i found for this particular drive. so that is not an issue either. 

Is the stepgen getting its input from the g-code from the ngc file/mdi commands?

and where does that take place?

Thanks

Try this fix:

hello rodw!!! Im istall www.linuxcnc.org/iso/rpi-4-debian-bookwo...23-11-17-1731.img.xz and I download a forum.linuxcnc.org/ethercat/53785-instal...thercat-repositories and install, then I go forum.linuxcnc.org/ethercat/45336-etherc...-how-to-step-by-step ... So, I got here! ethercat slaves and from here I gave the command and there was no error or anything appeared... and from then on... nothing else...

Turns out the issue was no M2 or M30 at the end of the file. I was just using the % markers at the beginning and end which is the wrong way to do that.

Deleted var file and it worked again.
cheers rod

Please follow the steps here
forum.linuxcnc.org/ethercat/53785-instal...thercat-repositories

Then go back to the first post on this thread.

SteveatSteelCut wrote:

update I turned ohmic probe off and it worked again.
If I try and use ohmic probe I get the same error.

Try deleting the VAR file anyway. I will look at the ohmic hal file but there should be nothing in it to do with offsets etc.

There should be a .var file in your config folder that linuxcnc saves data (variables) in. Normally you never have to worry about it but sometimes when hacking into a config, it gets out of step with reality. Please delete this file and report back if it is still a problem

update I turned ohmic probe off and it worked again.
If I try and use ohmic probe I get the same error.