• State Resolved
  • Replies 7 replies
  • Subscribers 35 subscribers
  • Views 117 views
  • Users 0 members are here
Support feedback
Options
Options
Related

DP83822H: Internal PHY Processing as an ESC

Satoshi Katayama
Prodigy 40 points
Part Number: DP83822H
Other Parts Discussed in Thread: AM2434, AM6442

Hello,

When performing EtherCAT communication with a configuration in which Port 0 is connected to the EtherCAT MDevice side and Port 1 is connected to an SDevice, I understand that data received from the MDevice (Master) passes through Port 0, goes through the EtherCAT Processing Unit (EPU) inside the PHY, and then reaches the SDevice (Slave) via Port 1.

On the other hand, for transmissions from the SDevice to the MDevice, I understand that the data flows from Port 1 to Port 0 without passing through the EPU.

Is it correct to assume that the DP83822H operates according to this same specification?

  • 0
    TI__Mastermind 21506 points

    Hi Katayama-san,

    Because our device does not have any switch functions and just passes the incoming signal to the MAC or to Ethernet, this will depend on how you implement our PHY in your system.

    Best,

    J

  • 0 in reply to J
    Prodigy 40 points

    Hello,

    Thank you very much for your reply. I realized that there was a mistake in my original question.
    Since the EPU is located inside the ESC, it should be considered a processor-side function.

    We are currently designing a circuit using a combination of the AM2434 and the DP83822.
    In the AM2434 ESC internal architecture shown in the figure below（AM2434 without Port2/3）, is it correct to understand that communication to the SDevice（Port0→Port1 of AM2434） goes through the EPU, while communication to the MDevice（Port1→Port0 of AM2434） does not go through the EPU?





    https://ipc2u.com/articles/knowledge-base/overview-of-the-ethercat-protocol-and-devices-based-on-it/

  • 0 in reply to Satoshi Katayama
    TI__Mastermind 21506 points

    Hi Katayama-san,

    I see. Let me reassign this post to Sitara team so they can better support with the question on the processor. From the PHY side, there should be no issue.

    Best,

    J

  • 0 in reply to J
    TI__Expert 5700 points

    Hi Katayama-san,

    As far as I undersatnd an EtherCAT SDevice controller (ESC) based on the AM2434 PRU-ICSSG0 or PRU-ICSSG1 can have maximum two accessible Ethernet MII ports (out of 4 maximum). Let's call them Port 0 and Port 1. I understand that your ESC accepts the downstream transfer from the MDevice on Port 0, and there is one SDevice attached on Port 1. My understanding from above EtherCAT protocol description:  https://ipc2u.com/articles/knowledge-base/overview-of-the-ethercat-protocol-and-devices-based-on-it/ is that the EPU captures the original command frame that enters the ECS through the Port 0. But after additional research I understand the EPU shall be responsible to manage NOT only the downstream but also the upstream to the MDevice. I may need to confirm this with the hardware topic owner.

    What is the purpose behind your question ?

    Are you trying to understand the difference between time delays in downstream and upstream direction ?

    Thanks for your clarification !

    Kind Regards,

    Anastas Yordanov 

  • 0 in reply to Anastas Yordanov
    Prodigy 40 points

    Hello,

    At present, EtherCAT communication is operating normally, so there are no issues.
    In this inquiry, we would like to ask about the specifications.

    We understand that, on the AM2434, the EPU is also involved in upstream communication.
    Could you please provide your responses to the following questions?

    • On the AM2434, is it possible to configure the system so that upstream EtherCAT communication does not pass through the EPU, in order to comply with the upstream EtherCAT communication specifications?
    • If any configuration is required, how should the user perform this configuration?
    • Is there any documentation that states that, on the AM2434, the EPU is also involved in upstream communication?
  • +1 in reply to Satoshi Katayama
    TI__Expert 5700 points

    Hello Satoshi Katayama-san,

    I would like to apologize I may have the wrong understanding regarding any EPU involvement in the ESC to MDevice upstream communication path.

    As I further researched on ESC EPU I came upon the following resources

    Select the Right Industrial Standard: EtherCAT

    and in particular the diagram and explanations in this document from Beckhoff Automation: 

    Beckhoff on EtherCAT Master Redundancy Mode

    Section,TwinCAT 3/TE1000 XAE/I/O/EtherCAT/General

    Section,TwinCAT 3/TE1000 XAE/I/O/EtherCAT/EtherCAT master - Advanced Settings/Redundancy Mode

    In terms of AM6442 I have found a hint in the below Sitara Ethernet FAQ:

    AM6442 Ethernet FAQ

    that EPU is NOT involved in the upstream path from ESC Port1 through Port 0 to the MDevice, as follows:

    Please let me know if I shall reach out to our EtherCAT software expert regarding any ESC EPU configuration questions / doubts ?

    Thanks for your understanding

    Best Regards,

    Anastas Yordanov

  • 0 in reply to Anastas Yordanov
    Prodigy 40 points
    Thank you very much for your response. I understand that in AM2434, the EPU is not involved in the upstream bus. This issue has now been resolved. Thank you for your support.