Advanced Search

It's pretty rudimentary. There are these basic PDOsso basically one send a wanted position to the drive and read the current position back. That's all.
Following error is calculated in linuxcnc.
The cia402 module basically just passes position after scaling to the drive. And reads back actual position.
It also does the state machine walking, but that's in the beginning, then that idles.
 

Spindle RPM is so high, backlash will be totally negligible.

Your motor and/or geartrain might not be designed for, or capable of, instant-reversing.

The motor could break, the gears could shatter.

(Steel can shatter; I've seen old worn-down scrap rails, stored beside the edge of the track, shatter like potterewy in placed where stresses, presumably due to thermal exansion, snapped them)

Thayloreing wrote:

I do the ST Linkv2 procedure and it does everything right,
 

hi
by correct procedure you mean the one described by @Scotta to convert a STlink V2 into a CMSIS-DAP probe.
Pyocd will not see it as STlink, but as CMSIS-DAP

 

Some easier way to do this ?
This code github.com/LinuxCNC/linuxcnc/blob/master.../hallib/xhc-hb04.tcl
creates hal pins pendant_util.X can do something with these pins?

Moin,
willkommen im forum.

für den Raspberry gibt es auch einen Parport adapter, so könnte man ein altes alte Mach3 BOB vielleicht weiter nutzen:

byte2bot.com/products/parallel-port-raspberry-pi-hat

aber es kommt auch immer darauf an was man vor hat, es gibt da einiges an hardware die man nutzen kann: Mesa, Remora, Rio, GPIO.

Gruß Olli

hmm, I am not seeing 6065 in my drives manual.

0x603F reports any errors including following errors which are described as Closed-loop mode tracking error over limit

In practice, a drive error is thrown and this pin can be referred to to determine the error . I have seen a few following errors here.
 

English is not my native language, so sorry for possible mistakes.

I write a c++ program that I run by adding loadusr myprogram to custom.hal(I'm using AXIS gui). This program do some stuff and closes with exiting from AXIS. I catch unix signals like SIGTERM and SIGQUIT to gracefully terminate this program(closing file descriptors, sockets and so on). Everything worked fine on LinuxCNC 2.8.4 but stopped working when I switch to LinuxCNC 2.9.2.
I can catch and handle signals when I directly run my program from terminal, but I can't catch signals when I run it via loadusr. I try with SIGTERM, SIGINT and SIGQUIT signals, nothing gets caught. Looks like in LinuxCNC 2.9.2 AXIS does't sending any of this signals to close it's loadsur subprocesses. Does it use any other signals to close its subprocesses? Or do I need to use another way to catch the termination of the program?

You have to use an OR comparator, to connect two inputs to a single output.

@Rod,

In the mean time, have done some reading about the cia-402.

• The

  0x6065

  : following error window is not used or set by linuxcnc.
• Velocity mode in combination with a tuned linuxcnc pid could be a try out.

@Tia,
if you want to access my machine remotely i can give you access
We could verify iff all is installed ok. Like the libclothoid is in place etc.
But i think you already used the ./installer. It can be the installer failed at some step.
For the rest i don't know much about the cia-402.

No double posting. Thank you.

I have already configured in hal cycle start pause stop external buttons
and spindle start stop external buttons and when i start linuxcnc it reports that
these signals already used and can't used by XHC-HB04.
How can i use both the external buttons and XHC-HB04 together without conflict ?

Hallo und guten Tag, bin neu hier und habe eine Frage zu deskproto.
Auf welchem System benutzt du es ? habe es über Wine auf Zorin installiert, Startbildschirm von Deskproto öffnet sich auch, habe es lizensiert - jedoch kommt dann recht schnell die Meldung "deskproto antwortet nicht" mit den Optionen "Beenden erzwingen" oder "warten". Unter warten geht es leider gar nicht weiter sodaß mir nur bleibt das Program zu beenden- irgendeine Idee warum das so ist ? besten Dank und Gruß Marc

@Pcw,

Thanks for your effort. Not everything is installed by the makefile.
I think you need to look at the ./installer file, wich install's locally :

libclothoid3d
libscurve

segment.h is part off libclothoid3d and is installed in /usr/local/include

Then finally there is done a ldconfig.
I hope it works for you then.

 

Also i saw, that in the Identity Mode the Axis soft limits work fine, but in TCP-Mode the maschine runs into the joint hard limits.

For non trivial kinematics proper limit checking prior to execution time is currently not possible because the planner is totally ignorant of custom kinematics. Your best option would be to check for limit violations in cam or running your gcode in a simulation configuration.
Limit checking does work during execution (ie when the joint position actually exceeds the limit) but that will cause an immediate (emergency) stop.
Sometimes it may even be necessary to loosen axes limits to be able to reach the full work space.

Trying to modify qtdragon a bit and add hal pins for slides. Found an example on the forum for analog, but I need to do for encoders. Tried to redo the code, at first glance everything seems ok, values change, sliders react. But the problem appears if we change the value by encoder, say by 50%, then press the button 100%, and if after that we reduce the value by encoder again, it immediately jumps to 40% (I specified a step of 10%). As far as I understand the problem is that absolute values are used, but it is necessary to use relative values. But I am completely stuck on how to implement it in the code.