PLC to PLC MSG Instruction Affecting Panelview Performance

[GTUnit]

I have a weird problem.
HMI response to operator input gets extremely slow under a certain condition:

Vendors pallet wrapping machine consists of a Micrologix 1400 and Panelview Plus 7 700. Their Panelview application is bloated but the panel response and screen reaction time was adequate.

We needed to interlock the wrapper with a another system running a L82E.

The following are observations of HMI sluggishness based on what changes were made.

Connected wrapper to L82E switch. No noticeable HMI sluggishness.
Enabled Read MSG. L82E reading 1 INT from Micrologix1400. Once a second. Minor HMI sluggishness.
Enabled Write MSG. L82E writing 1 INT to Micrologix1400. Once a second. Major HMI sluggishness.

Micrologix 172.16.11.2 / 255.255.255.0
Panelview. 172.16.11.3 / 255.255.255.0
L82E 172.16.1.2 / 255.255.0.0.
All other controls hardware on network 172.16.3.x / 255.255.255.0

Network traffic may seem like a logical conclusion but the Panelview works fine connected to the L82E network. The ONLY time it seems to take a performance hit is when I enable the write MSG instruction.

What is going on?

 

[Ken Roach]

Wow, that's bizarre. I just can't imagine one or two messages per second pushing the MicroLogix over a performance cliff.

The first thing I would check is the triggering logic on the 1/sec MSG instruction. Make sure you didn't inadvertently use the /DN bit so that it re-triggers as fast as possible. I trust that you know what you're doing with one-scan triggers, but I've seen that sort of error (and made it myself !).

Does the MSG instruction complete without an error in the ControlLogix ? If you can, add an RTO timer to evaluate how long between /EN and /DN.

Can you mirror the MicroLogix port to Wireshark or otherwise capture the data hitting that port ?

Is there a router or VLAN separating the Class B from the Class C network devices ? What's the Default Gateway setting on the MicroLogix ?

I wonder if something's causing the switch to dump its ARP table, or causing the MicroLogix to close all of its connections to the PanelView.

 

[JTCat]

Hi,
Have you tried using two reads, one on each processor, as opposed to a read and write?

 

[JeremyM]

What’s your messaging interval? If you’re unsure, it’s likely that these messages are being triggered as quickly as possible, as Ken suggested.