How to program for a failed meter.

Saturn_Europa said:

RS Logix 5000 v20.04 Contol Logix L-7 Processor

I have been tasked with recording the highest daily turbidity value on a reclaim water system. I am going to take a 10 second moving average and compare it to the previous 10 second moving average and keep the highest one. (Still trying to figure out how to reset it after 24 hrs). Suggestions?

There are two turbidity meters on this system. One for primary and one for backup. Is there a slick way to select between which turbidity meter the program will use for calculation? I would like the program to de-select the meter if it starts giving bad readings. Obviously anything under 4 mA would indicate a bad reading. Has anyone done this before?

Click to expand...

Bering C Sparky said:

Could you not just MOVE 0 to destination after 24 hours, then your next 10 second sample would be your high for the beginning of that day.

One way you could select between the 2 meters would be designate them as meter #1 and meter #2.
Use a compare to say that if meter # 1 < raw value of 4ma to latch a bit. (Use this bit to trigger an alarm also so you know meter #1 is faulty)
Then use this bit as XIC = meter 1 XIO = meter 2. Use these instructions to move the proper meter value into your math calculation for 10 second average.

You can put a button to unlatch the bit after you check/repair meter 1 and put it back in service.

(Of course meter 1 would be the primary reading and meter 2 would be your secondary should #1 fail, if you wanted to have either as primary you would just have to add logic to do that)

I think that is what you are asking, anyway. Figured I would take a shot at it, don't get to program as much as I use to and its all stating to slowly fade away.

EDIT: I had some jiberish in this post before which I deleted and added ladder drawing.

Click to expand...

Bit_Bucket_07 said:

Yes. I've done this sort of thing with redundant thermocouples. You say that a signal below 4 mA is obviously "bad", but I think that you need to better define what a "bad reading" is, prior to planning a means for avoiding using it within your control logic.

With a redundant thermocouple, I have typically chosen the higher of the two temperatures to employ as the process temperature, unless one of them has approached full scale. Open thermocouple channels are typically setup for fail safe high, and I've seen damaged thermocouples report temperatures that are higher than are feasible for the process being observed. How will you know whether a reading should be considered bad?

Click to expand...