Processor Fault Indication

Kroger2013

Member
Join Date
Dec 2013
Location
USA
Posts
32
Hello all, I am in need of some advice here. We had an issue with our waster water treatment plant here a week or so ago with one of our 1769-L32E processors faulting out which controls a skimmer. Another 1769-L32E runs our pumps and there is no hardwired I/O between the two. I have been tasked with shutting the pumps off if the processor (1769-L32E) faults out, shut the pumps off which are controlled by the other processor.

By the way, I did correct the issue with the processor faulting out, but the one's in charge of our waste water plant want this failsafe installed. Both processors are networked, but no I/O configuation is set up between the two currently.

Thanks for the help!
 
Simplest way is to set up a Produced Tag in each controller that the other consumes.

It will need to be a tag of a UDT. It need contain no data, can if you want, but include as the first element, a member called "Connection_Status" (the name is irrelevant), but make it of data-type CONNECTION_STATUS (It is already defined).

The Connection_Status member contains two BOOL flags "RunMode" and "ConnectionFaulted". These flags are maintained by the system, and in conjunction tell you if the remote processor is running, and actually producing valid data.

It really is all you need to know....
 
Another way to do this, which works on any platform, is to have one master PLC read a changing data register from the other slave PLC. I usually read seconds from the real time clock. If the read data is unchanged on several successive reads then trip an alarm. This is often referred to as a heartbeat check.
 
Another way to do this, which works on any platform, is to have one master PLC read a changing data register from the other slave PLC. I usually read seconds from the real time clock. If the read data is unchanged on several successive reads then trip an alarm. This is often referred to as a heartbeat check.

Of course that would work "on any platform", but to have to wait one or two seconds before indication that any data transfer is "invalid" could be one or two seconds too late. And it requires a fair chunk of code in each PLC, complete with message instructions in each PLC. And you don't want to be firing off messages too frequently either.

Kroger2013 is using the CompactLogix platform, so using the Produced/Consumed tag method I suggested is by far the simplest, fastest, most effective, and least expensive method.

Furthermore, it has no "connection" overhead if used on an existing data-transfer. The only drawback is that it subtracts 4 bytes of possible data transfer from the 500 available for a produced/consumed tag (The 2 BOOL flags occupy one 32-bit word, or 4 bytes, in the data-type)

I would be happy to write up a "walk-through" how to set this up, if requested. Can even provide a video of it in operation, if requested, as I just set this up between 2 ControlLogix controllers in about 5 minutes.
 
Hello all, I am in need of some advice here. We had an issue with our waster water treatment plant here a week or so ago with one of our 1769-L32E processors faulting out which controls a skimmer. Another 1769-L32E runs our pumps and there is no hardwired I/O between the two. I have been tasked with shutting the pumps off if the processor (1769-L32E) faults out, shut the pumps off which are controlled by the other processor.

By the way, I did correct the issue with the processor faulting out, but the one's in charge of our waste water plant want this failsafe installed. Both processors are networked, but no I/O configuation is set up between the two currently.

Thanks for the help!

If you want to go really simple and old-school, you could wire an output from the skimmer plc to an interposing relay and then wire that relay contact to an input on the pump PLC.

Of course, this assumes that installing the wire is feasible and that your outputs on the skimmer system failsafe to an off state when the processor is faulted.
 
But suppose comms are good but one plc failts...will the connection status tell you that?

Well, in that case as a part of the UDT data he can copy over the processor status register and read the bit. Then if the coms are up, but the proc faults he can see that, and if coms faults, he can see that.
 
But suppose comms are good but one plc failts...will the connection status tell you that?

Yes it will, because the remote processor won't be in RunMode

I can do a screenshot if you want it... and here it is. I faulted one processor and the other processor sees it as "not in Run Mode"

2017-10-07_025342.jpg
 
Last edited:
If you want to go really simple and old-school, you could wire an output from the skimmer plc to an interposing relay and then wire that relay contact to an input on the pump PLC.

Of course, this assumes that installing the wire is feasible and that your outputs on the skimmer system failsafe to an off state when the processor is faulted.

No wires or relays needed....
 
I thought about the old school way, but since they are both networked, I wanted to go the route of using communications between the two processors.

Thanks for the replies guys, this forum has been a lifesaver in the middle of the night because of the knowledge that you all freely share!

What daba recommended is what I was leaning toward wanting to do.
 
I thought about the old school way, but since they are both networked, I wanted to go the route of using communications between the two processors.

Thanks for the replies guys, this forum has been a lifesaver in the middle of the night because of the knowledge that you all freely share!

What daba recommended is what I was leaning toward wanting to do.

I wholly recommend it for ease of use, speed of response, and accuracy. If you got existing Produced/Consumed data, just rejig it so that one of them includes the Connection Status as the first part of a UDT, follow it with the original data structure you will be replacing.
 

Similar Topics

Hi Guys Today we went through the worst situation that PLC went to major fault Mode due to Trap by one of the Timer with -ve Value Moved into...
Replies
4
Views
1,013
Good Morning , I have some new Powerflex 525 Drives - This morning I got a F 114 uC Fault. Sounds like it is an issue with the microprocessor...
Replies
11
Views
13,527
Any idea why a MSG instruction (see attached screen capture) would cause an L81E Processor to crash hard with a T01:C61 Fault? After the fault the...
Replies
11
Views
13,886
I have a SLC 5/05 processor with a Major Fault 1F39h. The processor is running fine with the fault, but I would like to know what is causing it? I...
Replies
13
Views
3,764
Hey Guys, So I am lucky enough to be neck deep with an old piece of machinery that is controlled by Siemens S5 gear. It is here that I should...
Replies
2
Views
2,107
Back
Top Bottom