Hi,

I'm using Modbus TCP communication with ABB M102 Relays which is RTU protocal. So i have converter and taking as TCP Protocal. The time delay here is more than 2s.

Daisy chain format is used and per loop 10 slaves are connected with 2 analogue signals each and 10 digital signals.

Controller is AC800M controller and CI867 cards are used.

So if any body can lend me a helping hand, it would be great.

 

Hello,

1. verify you are connected to port 1 of CI867 (100mbit/s) not 2 (10 mbit/s)

2. operate the MBTCPREAD REQ parameter from a faster pulse generator so reads are more frequent.

3. Set polltimeout=0 to disable the diagnostic request.

4.check the block for error codes. code number will tell you if you have a problem and what it is.

5.replace your ethernet switch. (test with a standard basic home switch.) sometimes the problem is the switch. same applies to the ethernet cables.use CAT 5 or higher.

I was the manager of ABB instrumentation and control department so I know these controllers very well. (still supporting customers)
Hope I helped.

Good luck,
Mark

 

> I'm using Modbus TCP communication with ABB M102 Relays which is RTU protocal.

So really you're using RTU serial protocol ?

Yes, you will get delays. The delays will increase significantly when you have slaves not responding. The delay for each dead slave is the comms timeout * number of retries. You need to check where the timeout and number of retries are configured - its probably in your TCP to RS232 converter interface, but it may be limited by your M102 relays as well.

If you have a retry interval of 500ms and 3 retries then you have a 1.5 second delay for each message to each dead slave. If you have 10 dead slaves on the same multidrop interface (eg some plant is off for maintenance) and 2 messages per slave you have a 30 second delay.

Rob
www[.]lymac.co.nz