Strange data in MI

[LeonMötter]

So i have a weird problem where i have a source writing to the same MI (MI661 in a V560) that i am using to store the value of a temperature sensor.  The problem is that i don't know where it is coming from and the more i started looking the stranger it seems to get. I found the problem when i was looking at data from the temperature sensor a noticed that where a lot of points where the value fell below expected values. (normal is between 3400 and 350 and sometimes is noticed a value somewhere between 248 and 272). 

To find out in which part of my program i screwed up i disabled the sensor and i used the debugger to stop the PLC program. And then i noticed the first weird thing. Normally when the PLC is in stop mode you can write a Mi to zero and it should stay that way until the program is started again. In this case that doesn't happen. when i write a 0 to the MI it will immediately change back to the strange value it was on before, regardless of the state of the PLC.

Because of this, i thought maybe the source is not in the PLC itself but in one of the connected network devices. This PLC is a Modbus slave to 2 masters. The first is another Unitronics PLC (SM35). The second is a SCADA package (ignition). I disconnect them both and still something is writing to this MI.

So to summarize i have something writing to a MI even when the PLC is in stop mode and there is no network connection or anything else. 

Anybody any idea what could be going on here? do i have a ghost in my machine? 

[Fernando Castro]

1 hour ago, LeonMötter said:

So i have a weird problem where i have a source writing to the same MI (MI661 in a V560) that i am using to store the value of a temperature sensor.  The problem is that i don't know where it is coming from and the more i started looking the stranger it seems to get. I found the problem when i was looking at data from the temperature sensor a noticed that where a lot of points where the value fell below expected values. (normal is between 3400 and 350 and sometimes is noticed a value somewhere between 248 and 272). 

To find out in which part of my program i screwed up i disabled the sensor and i used the debugger to stop the PLC program. And then i noticed the first weird thing. Normally when the PLC is in stop mode you can write a Mi to zero and it should stay that way until the program is started again. In this case that doesn't happen. when i write a 0 to the MI it will immediately change back to the strange value it was on before, regardless of the state of the PLC.

Because of this, i thought maybe the source is not in the PLC itself but in one of the connected network devices. This PLC is a Modbus slave to 2 masters. The first is another Unitronics PLC (SM35). The second is a SCADA package (ignition). I disconnect them both and still something is writing to this MI.

So to summarize i have something writing to a MI even when the PLC is in stop mode and there is no network connection or anything else. 

Anybody any idea what could be going on here? do i have a ghost in my machine? 

I have found issues like this. have you in a previous program used an alarm and then deleted it?.

I found that if you delete the alarm group without reassigning the MIs/MBs first those registers get stuck with the previous assigned values, even using the search function won't find that operand being used somewhere else.

 

 

[Flex727]

Also remember that any operand in the middle of a vector will not be found using the search function.

[Ausman]

And......since your last download have you done a full Init & Reset?

Something else to try as well if all the above fails.  In the ladder, disable all those network comms in your program before doing the stop, and then set them to a test number/0 etc.  Maybe those details get hung onto vigorously  by the plc, even during stop.

cheers, Aus

[LeonMötter]

Thanks for all the responses. turns out the problem was more hardware related (maybe). i opened the PLC up en took the jumpers for the analog inputs off and back on again. After that the problem was gone. 

Could it have that one of jumpers had a lose connection and the value i saw in the debugger was jumping between 4-20 mA and either a thermo koppel or PTC input?

[Flex727]

Glad your problem is solved. Thanks for reporting back.