Allen-Bradley Analogue Input 1756-IF16 Scaling

[Rich0147]

Hi, I have a ControlLogix system with 1756-IF16 analogue inputs. I can't scale the inputs at the card as there is a requirement to facilitate re-scaling from the HMI.
I've configured the IF16 as follows:
Comm Format: Integer Data Single-Ended Mode
Input Range 0-20mA
What will be the Raw Values from the card at 4mA and 20mA please? This doesn't appear to be documented anywhere.

 

[joseph_e2]

I would probably use the module scaling to give a "raw" value of 0 to 2000 (or something like that). Let the module handle that first step of scaling. Then you can add your own scaling code that's edit-able from the HMI. Does that make sense?

 

[OkiePC]

Yes, the default scaling for that card will be % and the channel data raw numbers are floating point, so you can change the card scaling to 0.0 = 0mA, and 20.0 = 20mA and just display those values on the screen, and use the SCP AOI and its tag members can be edited from the HMI to allow runtime scaling.

 

[Rich0147]

Thanks for the responses. I've found that 4mA=-20030, 20mA=30921

 

[OkiePC]

Thanks for the responses. I've found that 4mA=-20030, 20mA=30921

Those are not default values, so you would either need to change them to something more user friendly, or leave them as-is and let the HMI display them as raw counts. In my opinion, they should be edited to make life easier on the end user not having to deal with a negative raw minimum. What version of firmware is in the controller? I think the version I glanced at yesterday was 31 and it had real numbers for the IF16 data values.

 

[chelton]

Scaling isn't available in integer mode The raw scaling is 1756-if16 is 0ma = -32768 21ma = 32767. User manual pg 31

 

[Contr_Conn]

Any explanation why you would be using Integer mode instead of Floating Point Data?
FP mode allows you to have scaling within the card in engineering units, so you don't have to deal with RA.
FP is also better even if you prefer to do scaling in the code, instead of using internal module scaling.

 

[robertmee]

Thanks for the responses. I've found that 4mA=-20030, 20mA=30921

All kinds of wrong with this method. Re-add the module using float, and scale the inputs to meaningful raw values like 4ma = 4000 and 20ma = 20000. Then you have raw values that are equivalent to the 4-20ma signal for troubleshooting, and you can use the HMI still to rescale the raw to engineering units.

 

[Rich0147]

All kinds of wrong with this method. Re-add the module using float, and scale the inputs to meaningful raw values like 4ma = 4000 and 20ma = 20000. Then you have raw values that are equivalent to the 4-20ma signal for troubleshooting, and you can use the HMI still to rescale the raw to engineering units.

Would you not have reduced signal resolution with this method?

 

[robertmee]

Would you not have reduced signal resolution with this method?

The resolution comes from the 16 bit A to D. A float has far more resolution to a significant digit than an integer. 2^16 is 65000 counts. So the integer range represents possible 65000 values. the 4000 to 20000 is still a real, IE 4000.0 to 20000.0, so at just 1 significant digit, that's 160,000 possible counts, were the module designed with a higher bit resolution . The AtoD is the limiting resolution for either method, not the scaling.

The key is configure the module as float.

 

[Rich0147]

Thanks very much for the responses. I've switched it to Floating Point as suggested.
Any ideas why some of the channels are displaying a small value and not indicating under-range when open-loop (disconnected)?

 

[robertmee]

Thanks very much for the responses. I've switched it to Floating Point as suggested.
Any ideas why some of the channels are displaying a small value and not indicating under-range when open-loop (disconnected)?

4-20 or 0-20?

 

[Rich0147]

4-20 or 0-20?

 

[robertmee]

Are your sensors configured as 4-20 or 0-20? Is your card set up as single ended or differential? If sensors are 4-20ma then you can use say a reading of 2000 to indicate an issue. If your sensors are 0-20ma, use the wire off feature of the model. It acts differently for how the module is configured. See the relevant section of the analog user manual:

 

[Rich0147]

Thanks again for your response. All instruments are 4-20mA. Card is configured single-ended.
I can apply the solution you suggest but just thought it strange that the in-built Underrange flag is not working on any of the even numbered channels, i.e. CH0, 2, 4, 6, 8, 10, 12, 14.