Announcement

Collapse
No announcement yet.

P2-550 RS485 Communication Port Failures

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • P2-550 RS485 Communication Port Failures

    I have two P2-550 PLCs connected RS485 using the network read/write blocks to get data from a the slave to the master. I been running for about 6 weeks and have had 4 slave port failures during this time. I have tried everything I can think of to restore the slave from powering down to removing the program and reloading but nothing works. I can install a new CPU and it will run for about 3 weeks before the replacement fails. I have a case going for the problem and we have sent the failed CPUs in for analysis but haven't heard anything. I have organized the reads and writes up between digital and analog points. My read and write analogs are about 30 points each. Could I be overloading the communication to the point where the slave shuts its port down? Hard to believe that I have four failed RS485 ports in this short amount of time. Any ideas?

  • #2
    have had 4 slave port failures during this time
    1) What are the TX/RX lights doing?
    2) What are instruction status bits doing? ( Have you put counters on them to see if they are going in progress, getting errors, etc?? Also this should be tested using manual polling.)
    3) Auto poll or manual? If manual are instructions interlocked?
    4) Have you monitored the RS485 Buffer Full Bit (95%) that is configurable in the hardware configuration?

    If lights are not responding and in progress status bits are counting then it sounds like you may be getting something on the network damaging the port.

    1) Have you checked for voltages on the network cabling?
    2) Are you using termination resistors on both ends of the network?

    Comment


    • #3
      What is the total cable distance?
      Does it run outdoors?
      If you have an urgent issue, please contact AutomationDirect's Technical Support team.

      AutomationDirect.com Technical Support: 1(800) 633-0405 or (770) 844-4200 Email Tech Support

      Comment


      • #4
        Originally posted by techme View Post

        1) What are the TX/RX lights doing?
        2) What are instruction status bits doing? ( Have you put counters on them to see if they are going in progress, getting errors, etc?? Also this should be tested using manual polling.)
        3) Auto poll or manual? If manual are instructions interlocked?
        4) Have you monitored the RS485 Buffer Full Bit (95%) that is configurable in the hardware configuration?

        If lights are not responding and in progress status bits are counting then it sounds like you may be getting something on the network damaging the port.

        1) Have you checked for voltages on the network cabling?
        2) Are you using termination resistors on both ends of the network?
        1) Master TX light is flashing trying to request from the slave
        2) I have status bits on a couple of the reads and they are show normal, no errors. All data is coming across fine until the ports fail
        3) Auto polling
        4) Buffer bit doesn't show full but I am not sure where it is configured to set the flag

        Haven't checked the voltage on the cable and we are not using terminators now. I had them on but removed them.
        The cable length is approx 2500 ft between CPUs and is armored direct bury cable.

        For now, here is my attempt at a solution since the failure yesterday was on both the Master and Slave

        We will install a MOXA TCC-120I signal repeater (with isolation) at each of the CPUs and let the repeater handle the long distance transmission.
        We will terminate at the CPU and the MOXA since that run is approx 18". We should have that installed early next week.
        I will update status on our progress.

        I do have a ticket on this issue 1654389.
        Last edited by LesYarbrough; 10-03-2019, 09:11 AM. Reason: Update since we installed the MOXA signal repeaters...no issues and we have had numerous weather events over the summer.

        Comment


        • #5
          2500' of cable is a mighty long antenna. Lightning wouldn't have to strike anywhere close and it could still pick up a lot of stray EMF.

          I think you are the right track with your modifications.
          If you have an urgent issue, please contact AutomationDirect's Technical Support team.

          AutomationDirect.com Technical Support: 1(800) 633-0405 or (770) 844-4200 Email Tech Support

          Comment


          • #6
            Originally posted by Do-more PE View Post
            2500' of cable is a mighty long antenna. Lightning wouldn't have to strike anywhere close and it could still pick up a lot of stray EMF.

            I think you are the right track with your modifications.
            I agree. 485 isolator/repeaters is a good choice first step to fix this. It will be interesting if AD support ever gets back with their findings to see if they can point out what happened to the other CPUs ports.

            Comment


            • #7
              woah. 2,500' on a 2-wire RS-485, AND you removed the termination resistors that help prevent signal reflections?
              Last edited by MikeN; 05-02-2019, 04:55 PM.

              Comment

              Working...
              X