Rosemount 644 Configuration Question

[tpatterson]

I have a Rosemount 644 on FF configured using DeltaV 7.4 The problem I am having is that when the temperature goes above (or below) the OUTPUT range, the output does not go Bad or Uncertain. I want the output to go Bad to indicate out of range.

What parameters should be configured in DeltaV to genrate this condition?

[rezabejd]

It must be rough being a supplier sometimes. I've been complaining to them to remove this feature that you want - as it manifests with later revision 3051 pressure and DP cells. I've been unable to find a way to turn it "off", so I doubt you'll have much success finding a way to turn it "on".

One of the development engineers for DeltaV thought it was a conformance issue - like it's written into ITK4 for outputs to behave as you suggest.

It's a nuisance for us, especially for flows, for which it is not uncommon to go above full scale, or especially to go below "0" slightly. With the "feature" you want, every time this happens, the associated PID controller sheds to manual. So for example, when someone swings from a dirty to a clean filter, or a spare pump is started up, an associated flow may briefly go upscale. I don't want the controller shedding to manual in such insatnces. Try operating like this for a while and see how well you like it.

In my mind, if a measurement is valid, why flag it as "bad" or "uncertain"? Isn't that one of the preeminent reasons we like digital integration, that we have much-increased data validation? I'm not familiar with the 644, but most fieldbus transmitters will indeed flag the signal as "bad" if, for example, the sensor fails or goes open-circuit.

I think what is needed, is a bit setting under "status opts" (for example) in the AI block, that makes "set status to uncertain / bad if reading exceeds out_scale" an option. Then you can have it your way and I can have it my way.

I'd be interested to understand better why you think you want to invoke this behavior. Signal statuses are great and open the door for engineers to do many clever things. We've been a ways down the road of cleverness and have backed off a fair piece.

[tpatterson]

The reason for wanting to do this is so that we know when something is abnormal (IE a thermocouple burn-out)

FF devices are supposed to define the PV status as either Good, Bad or Uncertain. I just want to know how to configure the device to give a PV BAD indication when the reading is out of my operating range.

[rezabejd]

T:

You will get the "bad" indication on sensor failure - burnout, short circuit, or otherwise - without any configuration. It's inherent in the way Rosemount builds the device.

Maybe you're thinking like HART, where you had to choose which way to go (upscale or down) on a sensor problem. No need to worry about this with FF, it's built-in. DeltaV does an good job reading in and propogating status in the "canned" function blocks, but you should take pains to see to it that your systems integrator is vigilant about reading & propogating status in CALC/LOGIC blocks and other blocks for which the user determines the outputs and actions (any block where you write an "expression").

In the early going, when we were setting up our system, we were checking the "target (mode) to MAN on bad PV" option on all our PID blocks. Soon after start-up, we were going back and un-checking them all. It was more of a nuisance for our operators to seek out the loops and reset the mode every time there was a little blip on the segment. Turns out FF handles mode shedding out-of-the-box and does so quite well.

I guess my point is, the SP-50 team and others who designed the function blocks and other accoutrements of FF, put a lot of thought into it. It does most of the things you want, without a lot of extra logic or configuration.