AB Logic 500 question: SCL command: offsets(?)

The input card (NI-4) has a range of 3277 to 16384 I believe. I found that someone had sub 3277 from the source value and had 0 for the offset. The scaled range of the device is 0-300, and the offset should be -75 I believe. The rate is 229.


In other places in the logic, they also sub 3277 from the source value, (which is the same word used in the above question) but left an offset value in the scl command. However the offset value in the command isn't what I thought it should be. Here's what they have:
Attached are pics from the logic. The logic is scaling the outlet temperature, using a rosemount device (4-20ma, -50 to 730F range)
I believe the rate should be 595, but they have 601. I believe the offset should be -245, however the offset in the logic is -50.

What am I missing?


Thanks in advance..


Newbie

Greetings New Controls Guy ...



and welcome to the forum ... I wish I had time to look through your program right now - but it will be later tomorrow before I get a chance to do that ...



in the meantime, the two-part article on Scaling that I wrote some time ago should help answer your questions ... you might have to read the whole thing to “tie it all together" - but (if I understand what you’re asking) the part that applies directly to your question starts with the last paragraph on page 11 in Part 2 ...



Part 1

Part 2



hope this helps ...

Ron Beaufort said:

Greetings New Controls Guy ...



and welcome to the forum ... I wish I had time to look through your program right now - but it will be later tomorrow before I get a chance to do that ...



in the meantime, the two-part article on Scaling that I wrote some time ago should help answer your questions ... you might have to read the whole thing to “tie it all together" - but (if I understand what you’re asking) the part that applies directly to your question starts with the last paragraph on page 11 in Part 2 ...



Part 1

Part 2



hope this helps ...

Click to expand...

Thanks. That does clarify a lot of things. There is logic in place to protect against values >16384 and <3277, with accompanying mov statements.
With the adjustment of the Input value by 3277, I can see now that the offset would be whatever the minimum value of the scaled range (0-300 would be 0, and for the -50 to 730F range it would be -50).
Makes sense, complete sense.

I'm still stumped about the rate though.. a range from -50F to 730F should have a rate of 595, yet they have 601. That still seems wrong.

I'm still stumped about the rate though.. a range from -50F to 730F should have a rate of 595, yet they have 601. That still seems wrong.

Click to expand...

Since the rate value is pretty close to what you calculate it should be, then it could be that the sensor came with a calibration card and the actual output best matches a straight line with the slightly different slope that you see in the program. It could also be that your plant calibrates in-house (we do this) and the last cal for that sensor resulted in a straight-line fit that best matches 601, not 595.

Paul T said:

Since the rate value is pretty close to what you calculate it should be, then it could be that the sensor came with a calibration card and the actual output best matches a straight line with the slightly different slope that you see in the program. It could also be that your plant calibrates in-house (we do this) and the last cal for that sensor resulted in a straight-line fit that best matches 601, not 595.

Click to expand...

Unfortunately it hasn't been calibrated in over 10 years. When the unit was installed in 96 the scls were set. Nobody has bothered with them since then. I've been fixing issues on the unit and also have revised/corrected things in the logic.
I'm used to using the scp argument, but this specific processor doesn't allow that command.
There are a number of rtds on the unit and they all use the scl command. I've corrected a fan bearing rtd scl issue, and realized that some of the other scl values didn't look right.
So I figured I'd better get some help from programmers more experienced with the scl command.

Well, sometimes the existing logic is just wrong <g>. This is why comments in a program are always helpful, even if they later show that the logic is wrong. 601 corresponds to an engineering range of about 788, which is a strange number in any case.

If the xducer hasn't been calibrated in ten years, though, the up to 1% error being introduced by the scale factor is probably no worse than the rest of the system. Also, I just re-read your first post. I haven't used Rosemount in... geez... about ten years! but as I recall they were a medium to high-end xducer. I'll still bet that it's related to a cal card that came with the unit when new. Usually, when you write enough logic to mess with the slope of the curve, you also have to use slightly different endpoints for the transducer low/high values. I haven't looked at the logic yet, but if the code makes provisions for a low value that is not -50, and a high value that is not 730, then that's certainly why the scale factor isn't 595. Upshot- comments are always good!