Co Processor on PLC5

[BeepBob]

Hello all,

I have been experiencing dropping network connections on the Co Pro quite a bit, 3 times a month. The remedy that we do to bring the network connection back is by bouncing the port on the switch that the co pro is connected to.

I read through the co pro manual and I do not see a recommendation for duplex mode and speed.

I see references to 10 Mbps ...so that is half-duplex?

I called Tech Connect and they have no clue...it seems this thing is just too old.


Anyone here experience this network behaviour on the co pro?

Maybe the NIC on the co pro is dying?

 

[chelton]

Check the diagnostics info on the builtin web server on the PLC5. Do you have a lot of FCS errors?
If your switch set to Auto neg the duplex/rate?
Cabling / incorrect configuration won't help

 

[Ken Roach]

If it's an A-B branded "Co-Processor" then it's almost certainly 10 Mb/s and half-duplex Ethernet. There was a late model "Ethernet Sidecar" that had 100 Mb and Auto-Negotiate but that would be clearly labeled.

Does your switch have any diagnostics available ? When the connection drops is there a Link status LED ?

It could be the CoPro's Ethernet port, or the switch's Ethernet port, or the cable, or some other failure or crash inside the module. The fact that the link is re-established by "bouncing" the switch side suggests a problem limited to the physical hardware or low-level elements of the link, rather than a stack crash.

 

[BeepBob]

Check the diagnostics info on the builtin web server on the PLC5. Do you have a lot of FCS errors?
If your switch set to Auto neg the duplex/rate?
Cabling / incorrect configuration won't help

Hmm FCS errors is at 0.

 

[BeepBob]

The PLC5 web server serves up neat information regarding the Ethernet connection.

Does it capture the statistics on the Co pro as well?

 

[Ken Roach]

I don't know what kind of statistics are available from the coprocessor. I think it's an OS/9 machine and I don't know if its TCP/IP stack included an HTTP server, or SNMP features.

You're not getting a CoPro crash or fault, right ?

When you "bounce" the switch, are you just disconnecting and reconnecting the cable, or disabling and re-enabling, or power-cycling ?

The 1771-DMC4 had an AUI port, not a built-in RJ45 jack as I recall. Have you swapped out the AUI/10BaseT transceiver ?

 

[BeepBob]

I don't know what kind of statistics are available from the coprocessor. I think it's an OS/9 machine and I don't know if its TCP/IP stack included an HTTP server, or SNMP features.

You're not getting a CoPro crash or fault, right ?

When you "bounce" the switch, are you just disconnecting and reconnecting the cable, or disabling and re-enabling, or power-cycling ?

The 1771-DMC4 had an AUI port, not a built-in RJ45 jack as I recall. Have you swapped out the AUI/10BaseT transceiver ?

Wealth of information as always Ken, thank you.

I do not get a Co Pro crash or fault.

When I bounce the port on the switch, I just " change state to administratively down" then back up.

I also tried disconnecting and connecting the cable, both solve the issue.

Now swapping out the AUI/10BaseT transceiver, do I go on ebay and find that part an try to swap it in?

 

[chelton]

Wealth of information as always Ken, thank you.

I do not get a Co Pro crash or fault.

When I bounce the port on the switch, I just " change state to administratively down" then back up.

I also tried disconnecting and connecting the cable, both solve the issue.

Now swapping out the AUI/10BaseT transceiver, do I go on ebay and find that part an try to swap it in?

What are the symptoms before you "bounce the port"?
Can you ping the IP of the device?
Is it a MSG command that stop's responding?
Are you able to connect via logix?
Have you tried changing ports on the switch?

 

[JohnCalderwood]

When I read the "PLC5 co-processor." I thought " we had issues with ours", but then it looks like you have a modern one ! We had the 1771-DMC, and had issues with ours as that last Rockwell Engineer that knew about them in the UK retired a few years ago.

 

[BeepBob]

What are the symptoms before you "bounce the port"?
Can you ping the IP of the device?
Is it a MSG command that stop's responding?
Are you able to connect via logix?
Have you tried changing ports on the switch?

1. Cannot ping the Co-Pro

2. The Co Pro runs a C program that sends a number of WORDS to another system

3. Logix will not see the Co Pro

4. Changing the port on the switch will work, as that is similar to bouncing the port

I think what this amount to is the NIC interface on the Co Pro failing. Rockwell Tech Support is stump. This will light a bigger fire on my management to provide funding ASAP to throw these things out and replace em with current Gen PLC.

 

[joseph_e2]

If you move it to another switch port and leave it there, will it eventually fail again? I might try replacing the network cable too if it's not too much trouble.

 

[PreLC]

What sort of code is running on the Co-processor?
An IPC with OPCUA+C Executible should be able to very easily replace the CoPro without needing to rewire the PLC5, with really low cost.