Advanced Search

Hallo,

ich habe nun wieder mal meine alte Fräse abgestaubt und in Betrieb genommen.
Diese läuft unter Linux-EMC2 (Schrittmotoren, Parallelport, Breakoutboard und 3x Endstufen fü die Schrittmotoren X, Y, Z)
Als Rechner hab ich ein mini-ITX Mainboard mit Paraleller Schnittstelle. LinuxCNC müsste so denke ich schon älter sein. Kann aber gerade nicht schauen.
War ein fertiges ISO mit ich glaube EMC 2.7.14. Einrichten der Maschine war glaube ich nicht so einfach, aber hatte es hinbekommen.
Bis jetzt hatte ich immer die Projekte mit QCad --> ESTLCam (Demoversion) --> Linux-EMC gemacht. Das hat auch immer super funktioniert.
Da ich auch FreeCAD Anwender bin und die neue Version 1.0 raus ist, würde ich FreeCAD als CAD-CAM verwenden wollen da frei. Das kann auch Gcode für LinuxCNC generieren.

Jetzt meine Frage:
Ich hab mir das ältere ISO linuxcnc-2.7.14-wheezy.iso heruntergaled und in Virtualbox installiert. Im Wizzard nun sim.axis-->axis mm ausgewählt und es wurde ein Icon LinuxCNC-HAL-SIM-Axis angelegt und ich kann LinuxCNC starten. Default wird ja axis.ngc mit den "LinuxCNC" Schriftzug geladen. Auch kann ich eine Referenzfahrt machen.Ich möchte nun unter der Simlation meine eigenen Programme testen. Also aus FreeCAD den Gcode erstellt und in LinuxCNC geladen. Das hat auch soweit funktioniert. Nur kann ich bei meinen Projekten keine Referenzfahr machen. Fehler lautet hier:
Keine Referenzfahrt bei geschlossenem gemeinsamen Endschalter. Aber warum geht bei den default geladenen "LinuxCNC" Schriftzug diese Referenzfahrt? Wo muss ich was einstellen damit es auch mit meinen Gcode die Simulation funktioniert? Wie gesagt mir get es nur um die Simulation um den Gcode zu testen.

MfG. 

in this case you can try in your terminal:

to verify your installation

if you are on the dev branch: (git checkout dev)
you can also set the gowin path in 'riocore/toolchains.json'

 

Sorry in advance for the noob question, I am finally setting up some hardware, generated a simple 3 axis step/dir machine.
When I try to compile I get      make:gw_sh: No such file or directory
                                                make***[Makefile:34: impl/pnr/project.fs] Error 127

I figured I hadn't set the path to the /home/user/gowin/IDE/bin folder but I double and triple checked and it seems correct.
I'm trying to program a tang nano 9K
 

the pull up/down resistors are there to have a defined state when starting / flashing,
otherwise it could happen that the spindle simply starts or the motors start uncontrollably

As there can be differences from board to board (BOB's), it can be installed using a jumper.

I have plugged in the TangNano and not soldered it on

Is there any way to improve the loading speed of G-CODE, reduce the memory footprint, and operate the response speed of G5X? Updating the plot in real time is too much of an impact on the experience, and the GUI is unresponsive.

@meister

Tangbob board:

What is the reason to add the 10k resistors as pull-down or pull-up resistor?

The FPGA is not configured as open-drain or open-source output.

The standard output configuration of the FPGA would not need the resistors.

Correct?

I would like to leave them out, but the cannot be added later when the FPGA has been soldered. That's why I am asking.

Correct.

But, with the mouse, I can select a MDI command from the list above and execute it.

Spending time looking a Youtube clips and reading documentation.
This is a whole new ballgame for me.

Should I give up on the keyboard and implement a virtual system to replace it.
Still not sure of terminology yet.

As a comment, and room for improvements, if one so wishes.

The stepspermm variable is set up as a pdos. Meaning that this value
will be updated every servo cycle (miilisecond). That is not very optimal,
the steps per mm scaling factor is a constant. Set it once and it stays there.
It is not like one will change the scaling value during operation.

The options I saw when I put it as a pdos was to either make it a sdos or a constant in the MCU program.
As a constant in the MCU program it becomes kind of inaccessible, but apart from that it will work fine.
Making it a sdos adds a bit of complexity when giving it a value. There is no interface for sdos to hal
that I know about. One can give a value to a sdos in the ethercat-conf.xml config file, but that brings
its own unnecessary complexity. One need to enter a line like the last line below

To end up with "40 06" one brings up a decimal to hex converter and converts 1600 which gives 640.
Shift place on the two bytes 06 40 => 40 06 which one then enters in ethercat-conf.xml.
The code in the MCU stays the same, still Obj.StepsPerMM1 (or what is is called). It is just not updated
every serocycle but once in the beginning.

Just a bit of background and rambling.

You mean you cannot enter MDI commands in that entry? (sorry the image is from an older gmoccapy version) 

die fehlermeldung sagt dir doch genau was los ist... dir fehlen diese 2 ordner in deinem config ordner.... erstelle diese und fertig

yes.

Amateur hobbyist on a 3-axis mill with RPi4 B in a single person shop.

I have just got LC 2.9.1 Gmoccapy working after using AXIS for a couple of years with just mouse and keyboard (on LC 2.8.4).
My interest is to regain MDI queuing and Run-from-Here that were lost in AXIS 2.9.1.

But, on my current system, when on-screen jogging is enabled by selecting "Keyboard Shortcuts" on the Systems Page, I cannot enter MDI commands from the keyboard.

Is it possible to setup Gmoccapy to allow both on-screen jogging and keyboard entry ?

John.

now we have modifier's for all pin's:seems more complicated at first glance, but is much more flexible