Advanced Search

Did you check the Beckhoff EL5101 latch documentation?
EN_LATC / LATC_VAL look like the terminal’s native C/index latch mechanism to me, so the EL5101 does seem to provide native index support in hardware.

The real gap looks to be on the LinuxCNC side: mapping that cleanly into the expected index-enable semantics.

From what I have seen in the LinuxCNC forum threads, there are bits and pieces of reference material, but apparently no complete end-to-end example covering EL5101 + LinuxCNC spindle index semantics + threading/orient.

Already replied on the other thread:
forum.linuxcnc.org/9-installing-linuxcnc...nc-2-9-8-mesa-7i76eu

Did you try to lower the jog velocity to something really slow and jog?

Do you have a copy of the hal file?

Também estou com o mesmo problema.aparentemente precisa editar o arquivo .hal no LinuxCNC. Conseguiu resolver?

This section of the hal file is broken in at least 3 ways:

loadrt [EMCMOT]EMCMOT servo_period_nsec=[EMCMOT]SERVO_PERIOD num_joints=[KINS]JOINTS
#loadrt [KINS]KINEMATICS
loadrt hm2_eth board_ip="192.168.1.121" config="num_encoders=1 num_stepgens=4 sserial_port_0=2"
#loadrt hostmot2

1. loadrt [KINS]KINEMATICS is commented out
2. loadrt hostmot2 is commented out
3. loadrt hostmot2 is in the wrong order (it must be loaded before hm2_eth)

It should look like what pncconf creates:

loadrt [KINS]KINEMATICS
loadrt [EMCMOT]EMCMOT servo_period_nsec=[EMCMOT]SERVO_PERIOD num_joints=[KINS]JOINTS
loadrt hostmot2
loadrt hm2_eth board_ip="192.168.1.121" config="num_encoders=1 num_stepgens=4 sserial_port_0=2"

is it possible to set it up as a output to a 5v input? I can trigger it that way?

@andrax

Sittner's repository does  not support 

syncToRefClock

If you want to test Sascha Ittner's code with pll
Set 
refClockSyncCycles="-1"

My code is a different repo

github.com/grandixximo/linuxcnc-ethercat

Sasha's code is missing some drivers, he has done a huge rework, he also removed the grind, but used another method, I am still undecided on how to merge his code properly.

Hello all,
I am a newbie and I might have posted this question on the wrong thread before so I apologize for my lack of knowledge and experience. I have got my board and pc communicatiing and I think I have fixed most of the error but until i get this one fixed I won't know. I have an error which seems to be misidentifiying my z axis. I have a xyyx router setup hopefully to get y  squaring. The error I am getting says an undefined kinematicsSwitch and it leaves my setup with undefined limit on z. All help will be greatly appreciated. I will post my ini and hal files for you to look at.
Thanks again,
George

        Thank you for for your assistance, i VERY MUCH appreciate it.I am having problems with my hal files and my ini files no matter what i do i keep getting the same error message kinematisSwitch not defined. If I could get you look at my ini and hal files I would be eternally gratful. I have got past the communication errors and my pc is able to communicate with the 7i76eu I think I have most errors fixed now but unitl I get this current error out of the way I don't know if ther are anymore. I am not sure if this is the right place to ask, so please excuse my inexperience. Thannks again
George

grandixximo wrote:

@TAKUYA
@Atsu

please check the value of the parameters lcec.0.pll-step
You can do so with this command

halcmd show param lcec.0.pll-step 

This parameter will be set at 1000 automatically if you have RTAPI_TASK_PLL_SUPPORT

also I have updated again the repo

if you could be so kind as to download, build and test again, thank you

github.com/grandixximo/linuxcnc-ethercat

Thank you! I pulled the latest changes and can confirm that it is now working. The dc-phased pin turns true, and I can see the pll- pins getting values other than zero. I did couple restarts and so far there is no grinding noise. I will continue testing to see if its gone for good.

I've checked the site buildbot.linuxcnc.org/ again. It's back up and running. A while ago, the site was showing an error and couldn't be opened. The inactivity is probably temporary.

Thank you all for your responses.

you are missing the go package.

"sudo apt-get install golang"

then "make" and "sudo make install"

Small progress update: the stub has basically turned into a virtual drive backend for validation.
I’m already exercising a fair amount of the core CiA402 runtime behavior in AXIS SIM with it.
Still a lot left to test before going back to real EtherCAT hardware, but the direction looks good.

You might also want to mention what tools/holders, speeds, and feeds you're using, as well as how far off the dimensions are.

That said, my first reaction to the above is deflection- the gouge looks like where the tool ended the cut, possibly dwelled there a bit before/during retract, giving it more time to cut. If it is deflection, there's a good chance you'll feel ridges if you run a fingernail along the floor of the machined surface perpendicular to the direction of cut.

Adding a finishing cut might be all you need- I generally leave 0.1~0.2mm of stock during roughing operations for the finishing cut(s).

Generally increasing rigidity should help as well (shorter tools/toolholders/stickout, bigger tools, better worktholding, working closer to the table), but I've always thought the columns on benchtop mills look too small, so you may be limited in options if that's the root cause. The good news is that so long as it cuts OK, taking a light finish cut should still get you on dimension.