1746-NI8 Config settings lost?

[Conor 4]

Hi guys, Yesterday in work we has a power "blip". A lot of the plant went down. One of these was a SLC. This SLC has Digital Input and Output card. It also has an Analogue Input card, 1746-NI8. When this piece of plant came back online, none of the Analogues were working. I checked out everything and found the power led was on but the channel led's were blank. I then went online and checked, no Analogues working. So I went off-line and checked the Channel configuration. I found that all the channels were set totally differently than what was originally programmed into them, the enable check box was even unchecked. I changed them all back to the correct settings and downloaded the code. Machine back online everything was ok. Has anyone seen this before or know of any issues that may have caused this to happen. The program was at the correct stage of the process where it had stopped at when the power blipped, so I don't think it was the battery in the processor. Conor

[rguimond 1]

My guess is that the original program was stored to an EEPROM, and that the analog card settings weren't set up properly when it was saved. After the power blip, the program stored on EEPROM was copied to RAM.

[Conor 4]

But this machine had ran fine for at least 2 years. I had it powered off before and it came back fine each time?

[IO_Rack 181]

I can share a similar story but can not confirm what had actually happened. It's a fairly remote site with generator back up. They had mentioned power issues. Customer had a ML1200 with an IF2OF2 analog module. When I arrived, the processor was faulted with a Major Error 71h. (cannot communicate with expansion I/O.) I reset the fault from Logix but the input data was way off. I checked the configuration and found it to be 0 to 10V (default). The application was definitely 4 to 20mA and the drawings confirmed that. The program was present and OK and there was no EEPROM. Also, the customer did not appear to have any programming tools. They did have electricians replace the sensors before calling us. I reseated the module and changed the configuration to 4 to 20mA. Edited 13 Apr 2013 by IO_Rack

[malc cope 1]

I suppose it depends on what form the blip took, but I have issues in the past where , one where a loose wire grounded briefly and another where energising a contactor caused a short circuit, both on different sites, but both caused apparent loss of the program within a SLC meaning that the whole code had to be downloaded again, may be a similar issue?

[kaiser_will 65]

I remember an issue years ago of a similar event that had a root-cause failure of a failing rack power supply. I believe the main rack power supply blipped and corrupted PART of the SLC memory. Again, the problem never reared its ugly head until the plant suffered a brown out and this machine just would not start up. I went online and found the program would not go into run mode as the analog input configuration words were completely whacked. Luckily I backup and archive the snot out of everything.

[Conor 4]

Hi guys, I was just checking this out again today and found what the issue was. I downloaded the NI8 manual as well, and found the Ladder Logic Configuration Examples on page 56. When I set up the card originally there was a line of code added to the Main routine SOR XIC S:1/15 COP #N9:0 #O:3.0 12 EOR S:1/15 is the first pass bit. Then I put in the correct values into N9 and checked Configuration of the card in the I/O cnfiguration and all seemed ok I had just started this reply and went back in to check the I/O Config and found that it was all over the place again. I checked the N9 integers and the had messed up figures in them again. I checked in the code and I can't seem to find anywhere that something is writing values into N9 I was wondering if anyone had any idea's? Thanks, Conor

[Ken Roach 194]

Most of the time when you have a power cycle, the SLC retains the current program in battery-backed RAM. But if there is a voltage spike or some other severe power disturbance that corrupts the RAM, the SLC has to load from EEPROM or have a program downloaded to it. That could explain why the machine worked fine during power cycles but not when there was a plant-wide sudden disruption. This new information is confusing. Is there an HMI on the system that could be writing to N9 ? Check for the presence of an EEPROM; at least knowing if one is present will help narrow down the possibilities.

[Conor 4]

Hi Ken, There is no HMI connected, but there is a Scada. I will check in the morning to see if there is anything writing to the PLC, but I don't think there is. I also don't think there is an EEPROM, again I will check in the morning Conor

[Conor 4]

Hi Ken, Without physically pullig out the processor, is there a way of checking to see if there is an EEPROM? I looked up the SLC manual and from there I was able to check system bit S:66 (Flash EEPROM size). This has a size of 512 in it. I also checked the N9 integers and they had the strange values in them again. I checked on my Scada and couldn't find anything that would be writing to this Integer either. Conor