Ethercat protocol not responding

did you add the "dcconf" line to the xml again?

Hi db1981,

I did add "dconf" line in xml.

I noticed that the uplink signal is working fine but some how downlink is not responding.

Regards
Abhishek

Hi,

I did some research and I found 2 things which might be the causing the problem:

1. Enable notLRW: yes
     I have gone through some post in forum and I came to a conclusion that because of the above problem, Drive required different domains for read/ write PIDs.
  Unfortunately I could not find examples or docs to activate multiple domains on the master to fix my problems and I don't know how to proceed further if it is "the Problem"

is there a way I can set "Enable notLRW" to no?

I will attach a links to get a clear idea of above problem:
www.forum.linuxcnc.org/24-hal-components...pins-rw-doesn-t-work

2. My Kernel is  "uname -a"

Linux linuxcnc 4.19.0-22-rt-amd64 #1 SMP PREEMPT RT Debian 4.19.260-1 (2022-09-29) x86_64 GNU/Linux

is still 4.19.0 but I don't think it should give any problem.

if anyone has any advice/suggestion, please help me out

Regards
Abhishek

Hello,

i have checked a little bit deeper about your problem...

Unfortunaly at the moment it is not possible to adapt notLRW drives. To get this handled it is neccesarry to do complete redesign of lcec.main and lcec_conf. But i can't see a time window for this in the next few months.

Explanation:

all slaves i had contact with, did not need this feature. notLRW tells, that the PDO data is transmitted in own telegramms for Read and Write (LRR and LWR). This should normaly be handled by the master, but the etherlab master does not do this by itself.
This can be solved by creating different pdo adress domains for read and write at ec conf state. Actual lcec_mains creates only one fixed domain, to blow this up a huge change in code is neccesary that sadly has to be backwards compatible...
with testing, building test-environment, collecting test-equipment, this will get an full week job.

Hi db1981,

Thanks for your reply,

As much as I could understand, currently linuxCNC Ethercat master supports only one domain(which basically does both read and write) but for my drives(or similar drives) we require separate domain for both read and write.
Is this what you want to tell?

if yes, we need to make changes in lcec.main and lcec_conf  to incorporate two domain for both Rx and Tx respectively.

As I want to start the operation with Ethercat, if you could guide me regarding how to change and what to incorporate,  I am happy to help in making changes for the same.

To start kindly provide me with some material where I can get some insight on how to make changes.

Thanks in advance

regards
Abhishek
 

Hi db1981,

As per your suggestion, I started going through lcec_conf.c and lcec_main.c, I discovered a great many things and according to that I made some changes to lcec_main.c,
Here are my current changes.

1) I have added one more domain structure entry in lcec_master.

The original was:

// create domain
    if (!(master->r_domain = ecrt_master_create_domain(master->master))) {
   rtapi_print_msg (RTAPI_MSG_ERR, LCEC_MSG_PFX "master %s domain creation failed\n", master->name);
   goto fail2;
}


I have modified it to:

// create domain
    if (!(master->r_domain = ecrt_master_create_domain(master->master))) {
      rtapi_print_msg (RTAPI_MSG_ERR, LCEC_MSG_PFX "master %s domain creation failed\n", master->name);
      goto fail2;
    }
    if (!(master->w_domain = ecrt_master_create_domain(master->master))) {
      rtapi_print_msg (RTAPI_MSG_ERR, LCEC_MSG_PFX "master %s domain creation failed\n", master->name);
      goto fail2;
    }


My next step are
1. I will be updating the read and write specific call backs in lcec_main.c 
2. Add the domain specific structures, I guess that it will be inside lcec.h.

Awaiting your response,
I hope I am moving in the right direction.

Regards,
Abhishek