Advanced Search

Ich habe aber noch was gefunden im Linuxcnc G-Code fehlt die G43 !
Das ist de Werzeuglängen Offset
zweites Bild wie es sein mueste
drittes Bild was vom Program erstellt wird 

king
regarde
Juergen

Ja Firefox-ESR Startet Webseiten lassen sich öffnen…

king
regarde
Juergen

ok, thank you!
I tried lcec_configgen and the result is xml file with 2 lines which, based on the error message, seem insufficient (please see attached image). note that it was before I tried adding cia402 module. Is that a mistake and is cia402 module mandatory? because the way I understand it now is that cia402 helps connecting HAL pins, but the VFD should be discovered over ethercat even without it. is that correct? one more thing is that cia402.comp in the comments contains suggested HAL file lines and there lcec_conf is loaded in line 50 before cia402 in line 52:
github.com/dbraun1981/hal-cia402/blob/main/cia402.comp#L50
so I think that it should work like it is, the problem most probably is the XML file
I was expecting at least some kind of output from "ethercat slaves" command - there was totally nothing. is that ok? I was expecting some message that shows discovered slave device or a message that no devices were found

The provided xml is useless for our purposes.Its for twincat on windows You could try typing: lcec_configgen to generate an xml file
cia402 is a standard for drives (and vfds). It essentially uses the cia402 state machine to safely power drives on. so for ethercat drives, we load the lcec driver then for each drive we load the cia402 state machine. Then at the end of our servo thread cycle, we write cia402 back to the ethercat layer and  then write everything back to the ethercat bus

zjjxxsh wrote:

I turned off all the turbo mode, C-state,etc. But the latency test shows the jitter is 50-60us. Is it enough for Ethercat control? Is there anything else I need to do with.
 

Your PC sounds fine. Just make sure coalescing is turned off on your NIC. type man hm2_eth for some details on this.

BSAT wrote:

I have used the Teknic software and run an auto tune and it helped but the problem still exits.

I havent changed the step input resolution settings(which is 3200 steps/rev) because other folks have no problem nor should I. I can see a tuning change being needed per machine but not a reconfig of input vs resolution.
 

Do the other folks you're talking about also have 1100MX's?  Just making sure it's an apples-to-apples comparison.

Assuming you and they have the same pitch ballscrews, the next thing to check is the actual gcode.

Is this a CAM-generated toolpath with lots of linear segments, or are the facets present on G2/G3 moves?

Are you using G64 in the problem code?  If so, what is it set to?

Hello!

My experience with LinuxCNC goes back for 15-16 years and now I am taking my first steps in getting it to work with EtherCAT. I am doing a retrofit of Biesse CNC machine and customer wants to use VFD for spindle that is made by ABB (model ACS380). Since by default that VFD has 2 digital inputs and 0 analog inputs, I could only get "start/stop" function to work so they got EtherCAT module (FECA_01 with firmware 1.30) for that VFD that I am now trying to get to work.
I have (at least I think so) successfully installed ethercat libraries as per first post by RodW in this thread:
forum.linuxcnc.org/ethercat/45336-etherc...-how-to-step-by-step
I have got XML file from ABB for the VFD drive (they provide ZIP archive with large collection of them and I picked the one that had matching firmware number of the EtherCAT module in filename), I have attached it.
Attached is also a screenshot that shows that EtherCAT service is active, those few lines in HAL file that I used to start lcec module, the thunar window to show actual location of XML file (to show that it matches the line in HAL file) and the error message of LinuxCNC.
Is there anything obvious I am missing?
In many forum posts I see that people also use cia402 module. Is it required here? I have not found any explanation of what exactly it is used for.
Thanks in advance!

Here is a little test program that slowly toggles all step/dir outputs so both states of all outputs
can be checked with a voltmeter:

 

You will need to

chmod +x 5i25stepgens.txt

then

sudo ./5i25stepgens.txt

and all step/dir outputs should toggle at a ~ 1/3 hz rate
 

Should be OK, but not the only important thing for EtherCAT, so in short: you can not know without actually testing it.

I'm using Simens IPC127e. It's a Siemens's computer working on Intel E3940/4GB and Intel i210 network card. This computer was used for industry control, it should working well with linuxcnc. 

I turned off all the turbo mode, C-state,etc. But the latency test shows the jitter is 50-60us. Is it enough for Ethercat control? Is there anything else I need to do with.

Btw, I have no Idea on turn-off the OnBoard Audio, how should I do.

Können Sie bitte einmal ein normales Terminal öffnen — ohne Takshak — und dorteingeben? Startet Firefox dann sauber, liegt der Fehler bei uns und nicht bei Ihrem System: unsere AppImage-Umgebung reicht dann eine eigene Bibliothek an Firefox weiter. Das wäre für uns die wichtigste Information.

You might try measuring at the 7I76U with the plugs removed
though 4.12V sounds OK under load. (its really marginal with no load however)

I can post a program a bit later today that will toggle all the step/dir outputs
slowly for a more complete test.

Vielen Dank für die schnelle Antwort…
Ok da ist was mit Firefox nicht so wie es sein soll…
Ich hatte zwa ein full Upgrade gemacht…..
Aber der Fehler blieb….
Aber erstmal Danke.

King regarde
Juergen

Hallo Juergen,

vielen Dank, das Terminal-Protokoll erklärt alles.

Beide Meldungen stammen von appimaged, das in Ihrem Downloads-Ordner liegt — nicht von Takshak. Der Ablauf war:

1. Takshak bittet den Desktop, das Handbuch zu öffnen.
2. Ihr Firefox-ESR stürzt beim Start ab: libatspi.so.0: undefined symbol: g_once_init_leave_pointer, „Couldn't load XPCOM". Das betrifft jede HTML-Datei auf diesem System, nicht nur unsere. Zum Prüfen bitte im Terminal einmal firefox-esr eingeben — der Fehler kommt auch ohne Takshak.
3. Weil Firefox fehlschlägt, reicht der Desktop die Datei an den nächsten Eintrag für text/html weiter, und das ist bei Ihnen appimaged. Dieses meldet dann „Not running on one of the supported Live systems" und scheitert am systemd-Dienst.
Bis Firefox repariert ist: das Handbuch steht auch online auf takshakcam.com in allen sieben Sprachen zur Verfügung.

Zwei Punkte gehen aber auf unsere Kappe, und die beheben wir: Takshak hätte Ihnen melden müssen, dass das Öffnen fehlgeschlagen ist, statt stillzubleiben — und wir werden das Handbuch künftig an einen festen Ort kopieren, statt es aus dem AppImage heraus zu öffnen.

Regards,
Georg

I don't usually clutter up threads like this just to say it looks great, but... looks great! I still need to get the spindle motor cooling shroud finished and installed on my machine... really easy to leave things unfinished once you can make chips, so good on you.