Strange problem

jaden · Oct 15, 2019

I had a customer call about a machine counter not working. We checked hardware, determined the input was correctly functioning. In tracing through the AB 5/03 program, I found one branch with the Input XIC was missing. I added that and its working great. Customer says that I was the last one connected to the PLC (six weeks ago) and it worked perfectly until 3 days ago.
The battery light was not on. How likely is it to lose one piece of a program without the rest being scrambled?

_Dock_ · Oct 15, 2019

About as likely as a dead president coming back to life and running for office this year. Send the bill and advise them to leave the machine alone next time.

jaden · Oct 15, 2019

They have no problem paying, just happy its running. Its just bugging me how that could occur.

Timbert · Oct 15, 2019

_Dock_ said:
About as likely as a dead president coming back to life and running for office this year. Send the bill and advise them to leave the machine alone next time.

Yeah. They can pay your fee from winning the lottery, because that is probably more likely than just one instruction "vanishing."

_Dock_ · Oct 15, 2019

There is no way im aware of that a program would delete a rung UNLESS that program was being booted from external memory that wasn’t updated. If a power failure occurred and a program booted from a memory that wasn’t updated from the edit that was made then it is possible.

Ron Beaufort · Oct 15, 2019

just guessing ...

there is a memory module installed - which does NOT have the "problem code" included ... and the bits are set to load the memory from that module ... then a power cycle (etc.) caused the processor to reload the code from the memory module ...

bet: slightly more than pocket change ...

I'd start by checking the settings of the status bits shown below ... in most cases, all of these are left at the default setting of 0 ...
.

_Dock_ · Oct 15, 2019

Ron,

We posted at the same time, as always tour teacher ways is better then my brief synopsis. See Ron’s post.

Kev77 · Oct 15, 2019

I second the memory module theory. I had a customer last week that said there program changed after 10 years. Come to find out they had some sort of issue and it booted from the memory module that wasn't updated after startup.

Aabeck · Oct 16, 2019

As far as a program changing - one of the first things the CPU does going to Run is check the program compared to its Checksum (or whatever it's called now) value and it they don't concur it will fault.

This is the same Checksum RSLogix looks at to see if the offline file matches the online, or you have to do an upload.

If somehow one bit got changed/read wrong that might be able to change XIC to XIO or similar, but losing BST .... NXB XIC B#/x.x BND would be impossible in the code.

James Mcquade · Oct 16, 2019

you forgot one important person, Bubba.
he likes to have things done his way, makes changes, and tells no one.
guess he has started working there. since a maintenance guy left here, things work much better and the calls in the night have been reduced 60%.
james

Strange problem

jaden

Member

_Dock_

Member

jaden

Member

Timbert

Member

_Dock_

Member

Ron Beaufort

Lifetime Supporting Member

_Dock_

Member

Kev77

Member

Aabeck

Member

James Mcquade

Member

Similar Topics