Remote rack loss detection with CLX and Ethernet?

Paul351W · Jan 18, 2010

I am setting up a new system consisting of 3 racks, one Panelview Plus 1500, and two processors.

One processor will act as the master control for a test site, with the second processor on a remote rack running high speed controls. The other rack only contains a 1756-ENBT module and I/O modules to serve the master PLC.

I am passing produced and consumed tags between the two processor racks and I have set up a heartbeat signal between the two so I can detect a connection loss, but how do I detect a communication loss with the remote rack? I've been searching the user manuals and help files for a couple hours and haven't found anything yet.

I want to shut the whole test stand down if communication is lost between any of the devices on the Ethernet network.

lostcontrol · Jan 18, 2010

Perhaps the easiest way is to use the GSV, Module, Rack Name & Entry Status.

http://www.plctalk.net/qanda/showthread.php?t=47388

You could also do some searching, there may be better options, but this should work quite well....

Paul351W · Jan 18, 2010

I'm used to the PLC5 Remote I/O world of having rack fault bits, so there not being a built in bit to check ethernet communication status is blowing my mind a bit here.

I will take a look into what you posted

Eddie Willers · Jan 18, 2010

In v17 if you use a UDT with the first element being a DINT named "CONNECTION_STATUS" as the P/C Tag datatype, the connection status becomes part of the Consumed Tag.

Otherwise there's the GSV method, which is the standard way I monitor I/O and P/C Tag connections.

There are a lot of Status items that existed as pre-defined data table elements in the PLC-5 and SLC that are accessible only by GSV or SSV in the Logix. It saves on unnecessary diagnostic background labor by the operating system, but does add a little labor (and understanding) for the programmer.

Paul351W · Jan 18, 2010

I'm still running v16 here, so I will have to stick with the GSV method. I have the two processor racks running at my desk, and the Entrystatus values are 4 on both processors, yet the slot status bits are all 1's...

I guess that tells me not to trust the slot status bits all the time. I don't have all the proper cards in the racks on my desk so maybe that's why the values aren't matching up. I do know the two processors are communicating because of the heartbeats that I programmed though.

Thanks for pointing me in the right direction guys, i couldn't find any obvious clues anywhere else.

Eddie Willers · Jan 18, 2010

For your "remote rack" connection to the chassis that contains the remote controllers, you only need to configure a "Rack Optimization" connection if you are also going to use Discrete I/O modules in that chassis. If the connection is done solely to set up a controller target for Produced/Consumed Tags, set the Connection Type to "none".

Slot Status bits are only for the Discrete I/O modules that are configured as "Rack Optimized" connections in a Rack Optimized chassis. The Produced/Consumed Tag connection to your remote controller does not get a Slot Status bit.

A quick way to check the overall connection status to a Rack Optimized chassis is to check to see if the DINT value of the Slot Status Dword = -1, which indicates a total connection failure.

MasterBlaster · Jan 18, 2010

Paul351W said:
I guess that tells me not to trust the slot status bits all the time. I don't have all the proper cards in the racks on my desk so maybe that's why the values aren't matching up. I do know the two processors are communicating because of the heartbeats that I programmed though.

You cannot accurately read the bits directly. The values are as follows:

4 = Running
7 = Waiting (cannot see the target)
5 = Shutting Down (Closing the connection)
1 = Faulted (No connection present)
3 = Connecting

Hope this helps.

Bill

rdrast · Jan 19, 2010

Paul351W said:
I'm used to the PLC5 Remote I/O world of having rack fault bits, so there not being a built in bit to check ethernet communication status is blowing my mind a bit here.

I will take a look into what you posted

Well, one reason that doesn't work in Logix, is that you don't know which processor in a system might own which I/O. Or even care if something is missing. Different polling times for modules will also affect just a 'Bit Field'.

I usually poll (sequentially and continuously) each and every I/O module in a system for it's status, which can include a lot more then "There/Not There". Electronic Fused output modules report back fuse status, diagnostic modules can report back wire-breaks, shorts, and the like.

Oakley · Jan 19, 2010

rdrast said:
I usually poll (sequentially and continuously) each and every I/O module in a system for it's status, which can include a lot more then "There/Not There". Electronic Fused output modules report back fuse status, diagnostic modules can report back wire-breaks, shorts, and the like.

I used to do this, but then decided that I didn't need to take up resources that often to do so. I poll once every minute. But then again, the routine for the polling includes 12 remote chassis.

Paul351W · Jan 19, 2010

I won't be able to independently set the communication formats different for the remote I/O rack and the remote controller rack because they will be tied to the same ENBT module in the test controller chassis, correct?

I figured I would just watch the communication status to the remote rack and monitor the status of cards on an individual basis depending on what would happen if they failed. We are understaffed and always pressured for time, so I don't get to take the time to monitor absolutely everything on the test stands unfortunately.

Eddie Willers · Jan 19, 2010

I won't be able to independently set the communication formats different for the remote I/O rack and the remote controller rack because they will be tied to the same ENBT module in the test controller chassis, correct?

Opposite. You can and should configure the EtherNet/IP connection formats to the remote I/O rack and to the remote controller rack separately.

The "root" 1756-ENBT acts as a "bridge", not as a "scanner". Multiple controllers could actually share a 1756-ENBT's bridging resources to different devices on the network.

I actually don't write "watchdogs", but rather depend on the GSV instructions to get me the connection status and controller mode information from the controllers I'm sharing Produced/Consumed Tags with.

Paul351W · Jan 19, 2010

I am adding a jpg of what my setup looks like roughly. The "main" PLC owns the I/O in the remote rack containing only an ENBT and I/O modules. The "fast" PLC and the main PLC communicate with produced and consumed tags.

So you are saying that I should set the ENBT in the main plc rack up as "none" for communication format, then set the ENBT in the fast plc rack up for "none", and set the remote I/O rack to Rack optimized?

Eddie Willers · Jan 19, 2010

So you are saying that I should set the ENBT in the main plc rack up as "none" for communication format, then set the ENBT in the fast plc rack up for "none", and set the remote I/O rack to Rack optimized?

Exactly correct.

If the "remote I/O rack" includes analog or diagnostic digital modules, they will have individual Module I/O connections, but if they are ordinary discrete modules a single Rack Optimized connection will cover them all.

mellis · Jan 19, 2010

So you are saying that I should set the ENBT in the main plc rack up as "none" for communication format, then set the ENBT in the fast plc rack up for "none", and set the remote I/O rack to Rack optimized?

You are making this harder than it is...

You are in the program for the main PLC.
You have an ENBT card in this rack in the I/O tree.
There is no choice for "none" or "rack optomization" for this card.
Under the ethernet you have 2 more ENBT cards, one for FAST and one for REMOTE.
On the one for FAST, you pick "none". (unless you want to look at the I/O in this rack)
On the one for REMOTE, you pick "rack optomization".

Paul351W · Jan 19, 2010

mellis said:
You are making this harder than it is...

You are in the program for the main PLC.
You have an ENBT card in this rack in the I/O tree.
There is no choice for "none" or "rack optomization" for this card.
Under the ethernet you have 2 more ENBT cards, one for FAST and one for REMOTE.
On the one for FAST, you pick "none". (unless you want to look at the I/O in this rack)
On the one for REMOTE, you pick "rack optomization".

When I started looking into the settings I realized what you just said. I am still pretty green to CL at this point. So I will set things the way you mentioned, and then in the fast plc set the ENBT for the main plc to none also?

I do have a mix of digital and analog cards in the remote I/O rack, I was planning on monitoring the analog card faults individually.

Remote rack loss detection with CLX and Ethernet?

Paul351W

Member

lostcontrol

Lifetime Supporting Member

Paul351W

Member

Eddie Willers

Member

Paul351W

Member

Eddie Willers

Member

MasterBlaster

Lifetime Supporting Member

rdrast

Lifetime Supporting Member

Oakley

Member

Paul351W

Member

Eddie Willers

Member

Paul351W

Member

Eddie Willers

Member

mellis

Member

Paul351W

Member

Similar Topics