A Modbus Developers' Community

A Control.com syndicated community

Link to Control.com main site
Control.com main forum

Link to the Modbus.org site
Modbus.org main site
Users' Forum

To view additional discussions, click here.


Return to Modbus Developers' Community Home Page

Mark V DCS Modbus problem
May 3, 2008 3:45 am, by HANYWEPCO
Subject : Applications
from the Automation List dept.
Text :
We have Mark V Control system connected to ABB Bailey DCS System through Modbus serial port. Some data are transferred as Packed (16 bits) then de-multiplexed in DCS to get 16 alarm digits. We have a problem that one alarm appeared in MarkV but did not come to DCS alarms although it is configured on both sides. This subject alarm came at the same second as another alarm but this other alarm came to DCS. Please can anyone explain why this can happen? Can the scan rate in MarkV cause this problem, i.e. if two alarms come at the same time in MarkV, can one of them be lost in DCS serial link? Why do we use Packed data for signals and when to use them and when to use logic (i.e. what is the difference between PACKED, LOGIC)?
Reply


  • Re: Mark V DCS Modbus problem
    May 4, 2008 7:29 pm, by CSA
    MODBUS data is "gathered" from Mark V turbine control panels by the operator interface (<I> or GE Mark V HMI). There is no direct connection between a DCS and the Mark V; MODBUS connections are made "through" the operator interface.

    Have you tried forcing the logic signal for the suspect alarm on the Mark V (usually it needs to be forced to a logic "1") and observing it on the DCS? Does the DCS display the alarm when it's associated logic signal is forced on the Mark V? I have seen several instances where the wrong signal is placed in the packed register for a particular alarm, and when that alarm is annunciated by the Mark V operator interface it is not annunciated on the DCS because the wrong logic pointname was used in MODBUS.DAT.

    Also, how fast is the DCS asking for data via MODBUS? Once per second? Faster? Slower?

    If two alarms are annunciated with the same time that means they were detected during the same scan of the CSP. Scan rate should not affect MODBUS information. I have also seen some DCS programmers "multiplex" the MODBUS signals, particularly packed registers, and if the alarm logic was only picked up for one or two scans of the CSP (which is usually one-eighth of a second per scan) the DCS might not detect a change of state occurring that fast. How long did the alarm condition exist? (You should be able to tell us that from the alarm printout information.)

    MODBUS won't "broadcast" an alarm the instant it occurs; it's a periodic transfer of information. I believe the MODBUS application on the operator interface polls the Mark V for the data when it gets a request, sends the data to the MODBUS Master, then waits for another request for information before polling the Mark V for the data and sending the data to the Master.

    MODBUS is fairly well described in the Mark V Application Manual, GEH-5980, for <I>s. There is a separate manual for GE Mark V HMI implementations of MODBUS, but it's virtually identical from a set-up perspective. More recent GE Mark V HMIs use CIMPLICITY to handle MODBUS communications instead of TCI. Packing is just a more efficient use of registers for transferring data which, usually, doesn't change very often or very fast.
    Reply


  • The man who sets out to carry a cat by its tail learns something that will always be useful and which never will grow dim or doubtful. - Mark Twain
    Your use of this site is subject to the terms and conditions set forth under Legal Notices and the Privacy Policy. Please read those terms and conditions carefully. Subject to the rights expressly reserved to others under Legal Notices, the content of this site and the compilation thereof is ©1999-2006 Control Technology Corporation. Powered by Python, Zope, and Squishdot, inspiration by Slashdot. (Vers.ZSQ1.0)