Extremely slow response from Contrologix

[harryting]

I'm stumped, yet again. Call from operator saying they are seeing alarms, turns out to be all HMI com alarm to the PLC. Connected up to the PLC (1756-L73) and everything is in extremely slow motion.

- PLC is on a DLR, device level ring.
- ping and network stat (web page) seems good.

I finally was able to open the main task property page in Logix5000 and it shows 24 ms, which I guess is not bad.

So, in summary, network seems good. The problem seems to be between the network card and the PLC somewhere. Beside changing out the PLC, any other tips?

ETA: oh, btw, I put the PLC into program mode via key switch but no improvement.

 

[Operaghost]

Have you looked at the web server on your Ethernet/IP adapter to see it's utilization? Is there an issue with the DLR? Is it one PLC and some I/O and an HMI or are there multiple PLCs or HMIs?

The scan time alone doesn't really tell us anything without a reference. Is that longer than usual? 24ms is somewhat long for the L73, but it is by no means excessively long.

What is slow? Is it the HMI updating slowly or is your process running slowly?

OG

 

[Paullys50]

task monitor tool

 

[daba]

I have observed exactly this scenario if one or more of the I/O modules is set for a very fast RPI...

 

[chavak]

Maximum Tag update rate of HMI screens.......

 

[harryting]

More info,

- network utilization shows to be at most 2% on the network card
- system ran OK for over a year before this started one early morning when no one was around
- no other plcs on the same ring shows the same issue
- slow to both the HMI and rslogix

I'll check the RPI and look up task monitoring tool tomorrow even though I feel it's unlikely the cause.

Oh, its so slow as to make rslogix unusable especially when the tag table is open. A digital input takes over. 30 seconds to show up.

Yup, this is a weird one. I'll also see if we got a spare cpu..

 

[brendan.buchan]

Right click the controller at the top of the Controller Organiser, go to Properties > Advanced and check what the System Overhead Time Slice is set to.

I've noticed similar behaviour when this is set to a low value <20%.

Still doesn't explain why it's only showing up now, but worth checking.

 

[Operaghost]

What Brendan said!!

OG

 

[harryting]

From the offline file I have, the Overhead timeslice is set to 20%, which is probably the default. I also check the RPI from the offline files and here are what I got

-DI/DO - 20ms
-AI - 100ms
-AO - 12ms
-Prosoft modbus card - 5 ms

I suspect they are all in factory default settings.

 

[dbh6]

How is the PLC connected to the network? are you using a managed or unmanaged switch? Or is this a point to point connections to the HMI from the PLC ethernet card?

 

[harryting]

Update:

We pulled the DLR from one side of the network card and notice the problem went away. So at this point we just replaced the card with a spare and reconnected all connections and the problem is gone. Both the existing card and the replacement is of the same vantage with 4.xxx firmware.

At this point we decided to put the old card back in and see what happens. Sure enough, put the old card back and everything still works just fine so now we just sit back and keep an eye on it.

Since we will have a scheduled downtime coming up soon we will leave it alone for now and update all the ETAP and 1756-EN2TR firmwares then.