Ethernet Traffic & PVP lockups

[69FIREBIRD 0]

Hi, I'm currently looking at an Ethernet network containing approx 10 Micrologix 1100 processors and 3 Panelview plus HMI's (all running the same application but generally viewing different screens and only one has active data logging). The network is physically laid out over a production line using 5 port CTRLink EISK5-100T switches (approx 10) to "daisy chain" the Ethernet network down the length of the production line. The PLC's and HMI's are patched into these switches at various points. With basic projects in the HMI's the screens tend to operate fine, but once trending, data logging and heavier use of the alarm feature are added the panelview’s tend to lock up - each screen has the current time displayed for reference and this freezes up periodically approx every 1-2 minute for a period of 1-2 minutes the buttons still function in respect that when a button is pressed while locked up when the screen unlocks the button action occurs. I have looked through Rockwell’s knowledgebase and "tuned up" trending and data logging. I have spoken to Rockwell tech support and they believe it will be a network issue, they have asked that I look at the PLC -> PLC messaging which I am in the process of doing to see if messaging between PLC’s can be better managed to improve the network traffic. I have noted that if a panelview is disconnected form the network then although the screen is not functioning by updating variables it does not lock up at all, it then looks to me like an issue with the data server when not connected it can’t update and therefore doesn’t lock up but when connected is able to update but is slow due to network traffic issues and this reflects in the screen locking up. Can anyone advise on possible fixes? I have also been looking into an Ethernet traffic analyzer “Ethereal” (a free web download app) but am a little short on knowledge on what to look for, can anyone advise on this and what to look for? Thanks in advance! Apologies for the length of the post.

[Charlie Maher 0]

The RPI settings make a big impact on EtherNet. Low RPI values can eat up the bandwidth fast. What are the update rates set to on the HMI's? Is the network a stand alone network or are you sharing bandwidth with the company printers? How much distance is the network covering? Sorry for all of the questions Charlie

[Ken Roach 194]

I doubt this is a network hardware or wiring issue. The MicroLogix 1100 is often overwhelmed by applications that would suit an SLC-5/05, just because of its ability to service Ethernet communications. Definitely examine your controller-to-controller messaging. If you have a complex set of peer-to-peer messages you may be using up controller CPU time as well as consuming more than the intended number of TCP connections. Use the RSLinx "station diagnostics" to learn a little about he ML1100's network port loading.

[69FIREBIRD 0]

Hi Charlie, I assume that by RPI you are referring to the screen refresh rate? If so the screens very in this respect from 0.1 second to 2 seconds this reflects screens that have variables that switch quickly and quick updates make for smooth consistent variable updates while others don’t have any information updating i.e. menu orientated screens. As already mentioned one application covers the whole production line (to help keep track of project updates / revisions) but most of the time the different PVP stations are on different screens and I therefore assume are only updating tags relevant to that screen being displayed rather than tripling up on network traffic. The network is standalone all IP addresses on the network are of the range 192.168.1.xxx. The network covers a distance of approx 70-80 meters. Thanks for your reply, feel free to comment on the above. Thanks Ken, will have a look tomorrow and read up on MLX1100 comms.