• Not Answered

EIOC communication with Powerflex drives

Hi experts,

We have a few PowerFlex drives controlled through EIOC over EthernetIP. Every time we write a new setpoint, we get an "Output transfer error" that recovers on its own within a second.

Note that it's not a full communication failure to the drive. The control module that sends discrete start/stop commands to the drive doesn't generate any errors.

We've already tried adjusting the RPI, but it didn't help.

Any ideas what could be causing this issue would be highly appreciated.

alex

• 20 May 2025 3:16 PM
• Cancel

6 Replies

• Andre Dicaire

  

  • 20 May 2025 3:41 PM

  What is current RPI and Module scan time?
  
  How are you configured to write values to the LDT signals?
  
  Ie. are you using external reference parameter or a DO. DC or EDC IO reference? Is the external reference an Output type and is it wired or set by an expression?
  
  Module writes to the LDT and LDT is sent asynchronous to device. I’m thinking RPI needs to be faster than module scan rate.
  
  Assuming your LDT are not reporting RPI error which can happen if you have too many LDTs on a Physical device such that the desired RPI cannot be met.
  
  I’d start there.

  Andre Dicaire

  • Cancel
• alex

  

  • 20 May 2025 6:09 PM

  In reply to Andre Dicaire:

  Thanks for reply. currently RPI set to 1000 and module scan rate is 1 second. But we did try setting RPI to 400 it did not help. But i will try again with RPI lower then scan rate
  
  Im using standard Analog output block to reference LDT.

  • Cancel
• Andre Dicaire

  

  • 21 May 2025 7:10 PM

  In reply to alex:

  • Output Transfer error is described as follows in BOL:

    A wired link to an external reference parameter cannot transfer its value to the external reference because it is rejected at the destination. See the example for the Input Transfer Error. from a transfer function block.

    You are using an AO block, and I assume you have this wired via the CAS_IN of the block.  

    Emerson recommends using an Output Parameter as an External reference for Bus and Ethernet IO Signals.  However, this is mainly to allow the module to manipulate the value of the signal even when status is bad.  AO block will stop writing if status is bad and that prevents the module from resetting the signal value if needed before comms are restored ( go to fail safe/passive value).  If this is not a concern, the AO should work and write to the signal when status is good.

    The Output Transfer is indicative of a wired connection in the module diagram unable to confirm successful write, which is often a wired connection to an external ref output parameter because the destination can't accept the value as sent (datatype mismatch or bad status).  

    An IO Output Error would typically indicate an issue with the IO connection of a block.  
    

    I would try some test configurations starting with just an external reference to the signal and confirm this can be manipulated manually.  And try with an AO block in MAN.  if the issue happens only when the reference is wired (which writes every scan of the module) make sure the LDT signal type matches what you are wiring to it. This may be more to do with the module or the AO out scale versus the signal data type.  Having a test module to work with and explore timing issues and such would help to troubleshoot.

  Andre Dicaire

  • Cancel
• alex

  

  • 22 May 2025 12:54 PM

  In reply to Andre Dicaire:

  Thank you Andre, ill start testing.

  • Cancel
• Casey Houchens

  

  • 6 Jun 2025 10:16 AM

  In reply to Andre Dicaire:

  We ran into a problem yesterday when doing a redundancy switchover on an EIOC that is communicating with four PowerFlex VFDs via 20-750-ENET cards. The communication is fine before the switchover, but the drives all fault on communication errors immediately after the switchover. Diagnostics shows everything is Good before doing the switchover. I’m wondering if this is normal behavior due to switching which EIOC has control of the Ethernet port, or is there some setting to prevent this?
  Thanks in advance for any feedback.

  • Cancel
• Andre Dicaire

  

  • 6 Jun 2025 5:28 PM

  In reply to Casey Houchens:

  When I first ran some tests on the EIOC with E300 relays, we saw the same thing. If I recall, the relays would reset when the master (EIOC) would change. The EIOC uses two different IPC addresses for the Primary and the Secondary, with the secondary card being assigned the next IP address configured on the Primary. You don't really see this second set of IP addresses (v15 PK controllers actually show this now when you configure networks on the node for use in the Ethernet port.)
  
  So if you do a switchover and the master node changes its IP address, the drives have to drop the old master and accept the new master. If we used wireshark to capture the traffic between the EIOC and the drives we would likely get some insight as we see the new active EIOC initiate the creation of the connection of the EIOC and the defined PDT's. Do the drives have communication parameters that can change how it handles a changed IP address of the master? Typically Rockwell uses IP swapping to maintain the same IP address for the active PLC or master device. Their drives are likely be default expecting the IP address to not change. But is there an option that allows the drive to hold while it accepts a new connection?
  
  I don't know any details on the drive in this respect, but the IP address change on switch over seems to me the issue, and that is not changeable on the EIOC.

  Andre Dicaire

  • Cancel