Updating PLC time from time server

[littlereiter99]

Hi everyone,

Looking for some assistance with using the SNTP AOI from AB. I have tried my laptop set up as a time server and a time server we have on the network but when I run the instruction I get General error 16#00ff on the SNTP MSG READ msg. Running V32 on a L73 hopping through a EN2T/D on Version 11.003.

Thanks for any help,

Derek

 

[PreLC]

Is there an extended error code on the MSG?

I was able to find some documentation in the EIP Socket Interface manual:

https://literature.rockwellautomation.com/idc/groups/literature/documents/at/enet-at002_-en-p.pdf

 

[littlereiter99]

Is there an extended error code on the MSG?

I was able to find some documentation in the EIP Socket Interface manual:

https://literature.rockwellautomation.com/idc/groups/literature/documents/at/enet-at002_-en-p.pdf

It is working now. I would love to give a good explanation on why, but when i logged in today after all night running the msg every five minutes it has updated the clock and the msg is dn with no errors. Nothing changed from when I left, the msg was still failing every 5 minutes. thanks for the reply and wish i could post a final solution to this...

 

[thingstodo]

It is working now. I would love to give a good explanation on why, but when i logged in today after all night running the msg every five minutes it has updated the clock and the msg is dn with no errors. Nothing changed from when I left, the msg was still failing every 5 minutes. thanks for the reply and wish i could post a final solution to this...

Your 'the SNTP AOI from AB' was imported from Rung_SNTP_AOIv2_02.L5X, downloaded from the rockwell Sample code, right?

I'm trying to get that running, am getting an error bit on the AOI, but I don't see where the error code is that you read. The AOI structure is large, so I may just be missing it!

I'm running an L73 on version 33.013 firmware, through a ENBT/A rev 6.001 (it's in the test rack). I'm trying to figure out whether the firewall will let the time request through, or maybe I have the time server address wrong, etc etc

EDIT UPDATE:
Yup - I just missed the ERROR CODES (there are more than 1)

I found an error code, assuming that .ERRC is an error code:
5 in SKT_Create.ERRC,
0 in SKT_Open.ERRC,
ff in SKT_Write.ERRC,
0 in SKT_Read.ERRC,
0 in SKT_Delete.ERRC.
It's possible that the ENBT/A is too old to support this function.

I also have an EN2T/D version 10.010 that I know is not routed to the subnet that has the NTP server.
When I run the AOI to that EN2T, I get an error code of 0 in SKT_Create.ERRC but the other error codes stay the same as above

I'm not on site so I can't switch out the ENBT/A with a more modern module to check if that fixes the problem.

Update 2:
Rockwell document https://literature.rockwellautomation.com/idc/groups/literature/documents/at/enet-at002_-en-p.pdf
Page 51. Error code 5 = Ethernet module does not have firmware that supports Logix Sockets (for the ENBT/A)

The EN2T/D 10.010 does not give me that error for SKT_Create.ERRC, but does give me ff on SKT_Write,ERRC. There are several causes, and I need to use the extended error code to look that up. SKT_Write.EXERR appears to be the extended error code .. and it is 33 = Open Sockets (TCP Client) is unable to connect to a third-party device (Slave). That make sense for the EN2T/D .. which can't route to the NTP server.

For the ENBT/A the extended error code SKT_Write.EXERR = 33. Perhaps that has something to do with the ENBT/A being too old to support sockets.

 

[Ken Roach]

You are correct that the 1756-ENBT is too old to support sockets. All the documentation tells you this, so you can discard any information or testing attempting to do so.

So instead, troubleshoot the 1756-EN2T installation, which appears to be telling you plainly and correctly that it can't connect to the NTP server.

The next troubleshooting steps I would take are more basic: I would verify that the -EN2T is configured for and connected to a physical and logical Ethernet network that can connect to the NTP server. And I would install another computer or device that is known to have a functional NTP client (like your Windows PC running its own time client, or something like the NIST time client) as close to the -EN2T as possible, including literally plugging into the same Ethernet port with the same cable.

 

[harryting]

I'm reading this thread with interest. We have used 1756-TIME before that's been discontinued (or I can't find it for sale). EN2T will sync to PTP source automatically but PTP server are not cheap. The AOI seems like a reasonable option.

 

[thingstodo]

..All the documentation tells you this ...

I know .. RTFM ..

in my defence, locating the FM sometimes takes a bit of effort

So instead, troubleshoot the 1756-EN2T installation, which appears to be telling you plainly and correctly that it can't connect to the NTP server.

Yup, that's why I tried it.

I think I mentioned, I'm located remote. I can't change out the ethernet card that is connected to the Control Domain until I drive to site. The EN2T/D is connected to IO, and I could steal it for testing .. if I drove out there. The plant does not have many places to actually use an ENBT/A, so it is likely to stay where it is longer term (except for testing)

The VM running Studio is on the Control Domain and can ping the NTP server and the ENBT. The only thing holding me back there *COULD* be the firewall, as the VM is on a different subnet than the ENBT. The layer 3 switch or the firewall *COULD* prevent the Control Domain VLAN from communicating. I doubt it, but it could.

And I would install another computer or device that is known to have a functional NTP client (like your Windows PC running its own time client, or something like the NIST time client) as close to the -EN2T as possible, including literally plugging into the same Ethernet port with the same cable.

Good ideas, but I have some limitations. There is only the one port on one switch in that test area that is part of the Control Domain VLAN. The IT guys are quite paranoid, so I'm not sure if I can even change out the ENBT/A to an EN2T without their help. But I'll try that first when I get to site.

Some days it might just be faster and easier to test things out on the production PLCs (but I won't)

EDIT: I guess I never mentioned that we had an EWEB card as the time grand master, communicating to a NTP server, and that it had a dedicated L61 to operate as the local time master for the rack, which then bridged to 2 different subnets using 2 different EN2Ts. All of the Logix PLCs were on those subnets.

IT's Network design has dictated that former time server rack is only connected to 1 VLAN. And now there are 4 VLANS in the Control Domain, one per Cisco Switch stack. When the time master moved to 1 VLAN, we lost time sync to half the Logix. As we moved Logix to the other 2 VLANs, it got worse.

So this is not an IMMEDIATE-PANIC type of problem. It's more of a nagging, it-should-get-done type of problem.

It will likely not progress until I get to site sometime in mid January. I was hoping to get something done this last week before Xmas .. but it didn't happen.

 

[joseph_e2]

I'm reading this thread with interest. We have used 1756-TIME before that's been discontinued (or I can't find it for sale). EN2T will sync to PTP source automatically but PTP server are not cheap. The AOI seems like a reasonable option.

The 1756-TIME does indeed show discontinued, but, interestingly, its discontinuation date is January 31, 2023:
AB Lifecycle status


They do refer to a product from Aparian:
https://www.aparian.com/products/timesync
Migration from the -TIME module to that would definitely require some engineering effort.