Allen Bradley 1769-IF4 Analog Input (Calibration/Accuracy Question))

Hi, I'm using an Allen Bradley 1769-IF4 Analog input crd with a Compact Logix L33ER. Its 14 bit (0-16384), and I'm taking raw readings. I am applying and measuring (with a multi-meter) a precise voltage of 1.50 volts across the analog input terminals. I'm configured for a single ended input signal, 0 to 10V.

The input value I am getting is 2350 which equals 10* 2350/16384 = 1.43 Volts.

The card specs say that it is internally calibrated yet this error is approximately 4.7%, way outside the specs for the card (0.2%). I am not observing any steady state offset or noise on the signal.

I know I can apply a fudge calibration factor in PLC code, but am curious if anyone else has used this card and achieved better accuracy, and what procedure they were using.

Thank you

.... just thinking out loud ...

My assumption is that 0-16384 corresponds to [0 to 10 V] (nominal operating range), not to [-0.5 to 10.5] (full scale range)?

agrajag42 said:

.... just thinking out loud ...

My assumption is that 0-16384 corresponds to [0 to 10 V] (nominal operating range), not to [-0.5 to 10.5] (full scale range)?

Click to expand...

That is what the book says (0 to 16383 = 0.0 to 10.0 vdc).

We use these cards with 4-20mA inputs and I often see a little bit of deviation from the manual... maybe five to ten counts, not 150.

In any case, for our purposes, we do a two point scaling of the raw into engineering units. I don't call it a calibration, just a scaling to real world conditions.

Apply full scale voltage to the input and see what you get.

The other thing that I am curious about... What is your precise 1.5 volts source? Are you measuring it at the same time? If so, the 2nd meter (counting your PLC input point as the first) perhaps has low enough impedance to affect the 1st.

Something I remembered from my circuits class was:

"Every meter impacts the circuit it is measuring to some extent, just as any tire-pressure gauge changes the measured tire pressure slightly as some air is let out to operate the gauge. While some impact is inevitable, it can be minimized through good meter design.

Since voltmeters are always connected in parallel with the component or components under test, any current through the voltmeter will contribute to the overall current in the tested circuit, potentially affecting the voltage being measured. A perfect voltmeter has infinite resistance, so that it draws no current from the circuit under test. However, perfect voltmeters only exist in the pages of textbooks, not in real life!"

I pulled this from this website for reference:
http://www.allaboutcircuits.com/textbook/direct-current/chpt-8/voltmeter-impact-measured-circuit/

Just playing devils advocate...

Could your multi-meter be out of calibration?