Advanced Search

I am very new to the system, i just bought a used CNC (with macch3) about 2 months ago and decided to upgrade to linuxcnc. i have a g540 diver and bought a mesa 7i92t. installed it and after 2 days i got it to where machine will turn on and respond to jog, however i cannot have Y and Y2 work together i can run them individually.

I have tried some stuff i found (including AI) and cant make it work i know its something simple for you guys but not to me.
and eventually i will have to figure my spindle out i have a 120v water cooled spindle with a VFD in one of my tries spindle kinda worked but now its no. any info or posts to help with that is also appreciated

I have the .hal and .ini attached if that helps figure out 

thank you in advance! 

   

Summarized Problem: Why doesn't the orientation of the spindle in the AXIS GUI reflect the actual position of the machine's spindle in space? Within the bounding box. The bounding box dimensions and size are correct. 
 
In our Linux CNC configuration when we turn on the machine and open AXIS, the GUI displays a 3D bounding box built by the min and max values given for the joints/axis in the INI file.  The GUI then puts the virtual spindle in the center of our workspace which is fine initially however our HOME position for our Bostomatic 3-axis milling machine is in the top front corner. I was expecting that when we press home all then the GUI virtual spindle will realize it is not in the correct position, however it does not, after it homes the virtual spindle is in the same position in AXIS  as when the application opened (the center). Where as our machine correctly went through the homing procedure where the soft-limit switch and home is shared and the actual spindle is now in that top front corner (which would be the top front corner of the GUI bounding box if it did what I wanted it to).  In this config the min and max values for the x, y and z axis are respectively (-9,9), (-6, 6) and (-6, 6). Which makes sense why AXIS is assuming that the origin, (the (0,0,0)) is the same at the home position.  So then I thought the solution would be to change the x, y, and z axes' mins and maxes to be: (-18, 0), (-12, 0) and (-12, 0) so that the (0, 0, 0) position is that corner where the home actually is.  This does put the virtual spindle in the right position. However this messes up the range of the axes because is assumes the position the machine is in when it is turned on IS on the edge of the bounding box and therefor won't jog in the positive direction because AXIS thinks it has no room but the machine indeed does. And therefor the GUI and the machine are still not on the same page just off in a different way.  SO, how do I get AXIS on the same page as the machine.  

I set configPDOS="false" and set debug level to 1 then started linuxcnc and did dmesg:
[ 5955.390588] EtherCAT DEBUG 0-3: Sync after 2916 ms: 10723 ns
[ 5955.391589] EtherCAT DEBUG 0-3: 9904 ns difference after 2920 ms.
[ 5955.391591] EtherCAT DEBUG 0-3: ref_time=812044903503955000
[ 5955.391592] EtherCAT DEBUG 0-3: app_time=812044909343879717
[ 5955.391593] EtherCAT DEBUG 0-3: start_time=812044909442878936
[ 5955.391594] EtherCAT DEBUG 0-3: cycle=1000000
[ 5955.391595] EtherCAT DEBUG 0-3: shift_time=0
[ 5955.391596] EtherCAT DEBUG 0-3: remainder=923936
[ 5955.391597] EtherCAT DEBUG 0-3: start=812044909442955000
[ 5955.391598] EtherCAT DEBUG 0-3: Setting DC cyclic operation start time to 812044909442955000.
[ 5955.393588] EtherCAT DEBUG 0-3: Setting DC AssignActivate to 0x0300.
[ 5955.396589] EtherCAT ERROR 0-3: Failed to set SAFEOP state, slave refused state change (PREOP + ERROR).
[ 5955.397588] EtherCAT ERROR 0-3: AL status message 0x001E: "Invalid input configuration".
[ 5955.399588] EtherCAT 0-3: Acknowledged state PREOP.

ethercat upload 0x60c2 1
0x02 2
ethercat upload 0x60c2 2
0xfffffffd -3

Don't know what this means; could you explain what interpolation time is?

Yes, that sounds like the standard Realtek driver issue (R8169)
You need to install the r8168-dkms driver

5I22-1.5 firmware for:

7I36 on P2 (PWM/ENC 0..3 sserial port 0 chan 0)
7I36 on P3 (PWM/ENC 4..7 sserial port 0 chan 1)
7I44 on P4 (sserial port 1 channels 0..7)


 
 

At its simplest
Then whenever the carousel.0.enable goes true, it will drive to that pocket.

All the subroutines etc are about the other things that have to happen before carousel motion, such as tool unload, spindle lift etc.

Ist ein Oszilloskop vorhanden? Falls ja, einfach ans Glasmaßstab anschließen, auf single shot stellen, Triggerspannung einstellen und die Achse joggen.
Ansonsten kannst du probieren mit dem Signal einen MOSFET und damit eine LED zu schalten (mit Diode in Serie am Gate, damit er auch geschaltet bleibt). Wenn bekannt ist, wo das Signal ist, kann man die Stelle anfahren und die genaue Spannung mit dem Multimeter messen. 

Can you start LinuxCNC from the command-line and paste the output here?

1/. The computer has a NEX 604 motherboard and uses a dual Realtek 8111E PCI Express Gigabit Ethernet chip.

2/.  

I'm guessing that the 8ms maximum tells you all that you need to know?

 

I think that this approach is probably valid, but will be necessarily comlex when building it from discrete HAL components.

You can look at the servo thread execution time with and without the added components to see if it is using a lot of CPU time, but it probably won't be.

Does Tormach include "halcompile" to create custom HAL components? If it does, then this is getting to the point where that might be worth doing.

linuxcnc.org/docs/stable/html/hal/comp.html

I always fancied the idea of being able to define limits as an STL file.
(STL because it's simple and easy to parse)

Are you basing your config on this wiki page?

wiki.linuxcnc.org/cgi-bin/wiki.pl?Simple_Remote_Pendant

Are you using the "-f" flag when loading the component?

That seems to have that effect. (But it might also be the default behaviour)

I've just put up a YouTube video with a quick demo:

unknown wrote:

Uninstall 2.9.5

Manually download the 2.9.4 deb files using gdebi.

Or 2.9.6 which has been created to fix this specific issue, but hasn't been moved onto the repository yet (I haven't been at home long enough to do it since the debs got built).

buildbot2.highlab.com/release/debian/dists/
 

"Dear friend,
I am working on retrofitting a grinding machine and I have prepared the following hardware for it: Mesa 5i22-1.5 + 2x 7i36 + 7i44 + 7i70 + 7i72 + 7i73. I would kindly like to ask if it would be possible to create firmware for this combination.
Thank you very much in advance."