Ethercat drive configuration woes

Yes I found that it easiest to do it headless via a VNC viewer I have two screens I set VNC with the pi on one and use the other one as normal you can copy and past etc makes it very easy

Ok, so with the EK1100, LinuxCNC won't start if Ek1100 is not connected. To get the system to work, the drive needs to be disconnected, and EK1100 connected then connect the drive after linuxcnc starts!

So EK1100 is '0'
EL1809 is '1'
EL2809 is '2'
drive is "3'

If I get time over the next few days, I will try to see if I can configure it step by step in the .xml as I first did with my mill 8 years ago. I think the setup has changed a bit since then, but I might try to configure it step by step and let the cia402 component do the rest. This should allow the drives to remain plugged in when I start linuxcnc (hopefully); otherwise, I will go shopping for a different drive

I've just done a fresh install to see if any of these issues are fixed, and I can report that it's now possible to get all 3 drives functioning, however they still don't enable if plugged in when LCNC is loaded.

I've not had a look to see exactly what's happening with status/control words, but if I have them connected when LCNC is loaded, then they don't enable. I tried disconnecting and reconnecting them once loaded, but they don't enable.
However if I leave them disconnected, open LCNC, then connect them, they enable, and all 3 work.

I've just spent some time going back over this, and the problem is still the same (I'd forgotten that I had managed to get all 3 drives enabled previously).

If I load LCNC with drives powered up and connected, opmode and statusword remain as 0 in Halshow, yet looking at the ethercat debug output, the value is shown as 0x0231, as per my previous screenshot in post forum.linuxcnc.org/ethercat/47534-etherc...woes?start=40#259032

Load LCNC with the drives disconnected and/or powered off, then connect the drives, opmode and statusword get reported correctly in halshow (they match the values shown in the ethercat debug output), and the drives enable as expected.

I forgot to ask, how could I go about debugging this?

I've got as far as understanding that lcec is the go between the Etherlabmaster, and LCNC. However I'm unsure how to go about trying to establish if the value not being transmitted is a lcec or etherlab issue. How could I go about viewing the various variables involved?

my 5cent ... if you slaves have a "strict" configuration, if you make some error on PDOs definition in xml file maybe related value stay at 0 without other messages or warning .... simply not work at all .... in that case if you try open a terminal and launch "ethercat slaves" command probabily you see your slave in PREOP or SAFEOP state .... so nothing work ....finally if your xml file contains "ek110" definition for sure linuxcnc not start (or start with some error and close) if you not connect it .....

regards

When you recently started where you left off, did you use the master from debs or the master from 1.5.2? 
db1981 suggested this back in december.

I'm struggling with statusword=0 too, and I wonder if I should go back to the 1.5.2 master, because I had these drives working several years ago. 

Reid

I used the master from debs, as I'm not sure how to get 1.5.2

My guess is some kind of issue between the master and lcec, as if you query the master via a terminal, it reports the correct statusword value, but that value doesn't make it to LCNC, so the cia402 module then doesn't do what it's supposed to in terms of triggering the process to enable drives.
However I'm not sure how to debug this any further, as Linux isn't my strong point.