• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 31 subscribers
  • Views 350 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM2432: Error occurs when using IgH EtherCAT Master ver. 1.6.0

Kanae
Mastermind 9825 points
Part Number: AM2432
Other Parts Discussed in Thread: TMDS243EVM

Hi Support Team,

Error occurs when using IgH EtherCAT Master ver. 1.6.0

When the customer's ECAT slave board with AM2432 is used,
depending on the type of ECAT master, an error may or may not occur.

  • Error occurs
    -IgH EtherCAT Master for Linux ver.1.6.0
  • Error does not occur
    -IgH EtherCAT Master for Linux ver. 1.5.2
    -TwinCAT
    -ECAT Conformance Test

Error appears when ECAT is started with the master and the red error LED of the ECAT is blinking.
The same error  also occurs on TI's TMDS243EVM.

In the customer examination, Beckhoff's ECAT controller and Renesas' CPU (with built-in ECAT controller stack)
were also confirmed to work without any problem, but only AM243x is not working properly.
(According to the checks by Beckhoff and Renesas, no special modifications have been made for IgH.)

Q. Has an error ever occurred depending on the version of IgH EtherCAT Master for Linux?
     Please let us know if similar problems have occurred in the past and what we can do to solve them.

Best Regards,
Kanae

  • 0
    TI__Mastermind 24030 points

    Unfortunately we do not test with various versions of IgH stack. So what has changed in IgH from 1.5.2 to 1.6.0 is not clear. Traditionally there is a mismatch in IgH configuration I copied below:

    IgH master by default does not work with TI ESC as it uses non-optimal PDO config which requires TI ESC to increase process path latency per device by 400ns.

    • activate overlapping pdos leads also to overlapped data within the domain memory area. so the complete domain must be read before writing to the domain.
    • ecrt_slave_config_reg_pdo_entry by hand is not neccessary.
    • ethercat cstruct creates non-working watchdog values in ec_sync_info_t. for me ONLY EC_WD_DEFAULT works.

     This a 6 year old thread in TI E2E has workaround suggestions on master side (and still relevant I believe)

    https://e2e.ti.com/support/processors-group/processors/f/processors-forum/666649/processor-sdk-am335x-issues-with-igh-ethercat-master/2451531

      Pekka

  • 0 in reply to Pekka Varis
    TI__Mastermind 24030 points

    Could you also add the error logs. From above I understand the only thing happening is red error LED is on?

    Also with IgH 1.6.0 are you referring to https://gitlab.com/etherlab.org/ethercat/-/tags/1.6-alpha ? 

  • 0 in reply to Pekka Varis
    Mastermind 9825 points

    Hi Pekka,

    Thank you for your support!

    Customer's response to your confirmation is as follows.

    [Details about the error]
    -Red error LED flashes.
    The blinking pattern is a double flash.
    The specification of the double flash is 200ms on, 200ms off, 200ms on, 1 second off, and it looks as if the LED is on twice and off once.

    -When you start ECAT by executing the following command, the error LED starts to double-flash after about 10 seconds.

    $ sudo /etc/init.d/ethercat restart

    -Execute the following command during the double flash, and you will see the error in SAFEOP.

    $ sudo ethercat slaves
    0 0:0 SAFEOP+ERROR E AM263x/AM243x/AM64x (PRU-ICSS)

    -Double flush means "process data/EtherCAT watchdog timeout" so I assume there is a problem with PDO.

    [Information about the version of IgH]

    Please check the update as follows.

    As you understand, the version we are using is devel-1.6.0.
    We will update the information.
    We stated that the error occurs with IgH 1.6.0, but it also occurs with IgH 1.5.2.

    Here are the two ver. that we are using
    devel-1.6.0
    stable-1.5.2

    Please let us know if there are any workarounds related to the possible issues
    from the above error information details.

    Best Regards,
    Kanae

  • 0 in reply to Kanae
    TI__Mastermind 24030 points

    1.5.2

    Pekka Varis said:

    IgH master by default does not work with TI ESC as it uses non-optimal PDO config which requires TI ESC to increase process path latency per device by 400ns.

    • activate overlapping pdos leads also to overlapped data within the domain memory area. so the complete domain must be read before writing to the domain.
    • ecrt_slave_config_reg_pdo_entry by hand is not neccessary.
    • ethercat cstruct creates non-working watchdog values in ec_sync_info_t. for me ONLY EC_WD_DEFAULT works.

     This a 6 year old thread in TI E2E has workaround suggestions on master side (and still relevant I believe)

    https://e2e.ti.com/support/processors-group/processors/f/processors-forum/666649/processor-sdk-am335x-issues-with-igh-ethercat-master/2451531

    Did you try this? 

    Kanae said:
    Error does not occur
    -IgH EtherCAT Master for Linux ver. 1.5.2
    Kanae said:
    We stated that the error occurs with IgH 1.6.0, but it also occurs with IgH 1.5.2.

    On 1.5.2 it works of or it does not work? Assuming 1.5.2 does not work try the above.

      Pekka

  • 0 in reply to Pekka Varis
    Mastermind 9825 points

    Hi Pekka,

    Thank you for your reply.


    Pekka Varis said:
    Did you try this?

    It is being tested at the customer.


    Pekka Varis said:
    On 1.5.2 it works of or it does not work? Assuming 1.5.2 does not work try the above.

    According to the updated information, the problem is also occurring in IgH 1.5.2,
    so the above measures will be tried with the customer first.

    I will share the results of the customer's testing here as soon as I have them.

    Best Regards,
    Kanae