Network slowly when using 1756-DHRIO and 1756-ENBT to bridge between RSView32 and PLC

[bekrab]

We have DH+ network on 6 PLC5.
We have RSView32 as a HMI.
Previously we link HMI with DH+ network by 1784-KT card but we have a problem on some node drop from time to time without any error.

So, we try to use Gateway Rack on controllogix platform by combination between 1756-DHRIO and 1756-ENBT.

This config is work. We can link RSView32 through gateway rack to DH+ network.

The one problem is response time. When we change setpoint in HMI, It is very slow (about 30 seconds) before we can see feedback in HMI.

Please advise.

 

[OkiePC]

What version of RSView32?
RSLinx?
Operating System?

Do you have switches and routers and other devices on the ethernet side?

Even with the gateway, you still have 7 nodes on DH+, so depending on how efficient your communications are structured, delays of a second or two are normal.

Delays of 30 seconds would make me look at wiring first, then packet rates on both networks (DH+ and Ethernet).

 

[bekrab]

OkiePC,

We use WindowXP SP2.
RSLinx 2.51
RSView32 7.1
There is no any device or PC except one RSView32 station connect on Ethernet network side.

We test by connect DH+ by using PLC5. Response time when we enter setpoint is normal.

So, We try by connect DH+ by 2 PLC5 Response time when we enter setpoint slower than one node. It is about 3 seconds.

When we connect 3 PLC5 into DH+ network, now we got response time about 8 seconds.

all PLC5 model is 5/30.
Number of tags per PLC5 node is about 1000 tags which is mixed both analog ,digital and alarm tags.

Is there any suggestion? Please advise.

Thank in advance.

 

[bekrab]

Any suggestions......

 

[Ken Roach]

If this were my system I would be analyzing it with protocol analyzers to figure out how much traffic there is on the DH+ system and where the delays are coming from on Ethernet.

How many nodes are there on the DH+ network ?

Are any of the controllers sending peer-to-peer messages to one another ? This kind of messaging is very often done incorrectly and generates more traffic than is necessary.

Check the Statistics on the 1756-DHRIO channel (using RSLinx Classic) and see if there are any errors or retries on the DH+ side.

 

[bekrab]

Thank you Ken for your recommendation.

Maximum we have 6 nodes but we test on 3 nodes and got un-acceptable delay.

We will try on your recommend and will let you know soonest.