Basic Operation
873
SPNU563A – March 2018
Copyright © 2018, Texas Instruments Incorporated
Analog To Digital Converter (ADC) Module
At this point, the ADC can be configured for normal operation, and it corrects each digital result with the
error correction value loaded in ADCALR.
NOTE:
Prevent ADC Calibration Data From Being Overwritten
In calibration mode, the conversion result is written to ADCALR that overwrites any previous
calibration data; therefore, the ADCALR register must be read before a new conversion is
started.
For no correction, a value of 0x0000 must be written to ADCALR. In noncalibration mode, the ADCALR
register can be read and written. Any value written to ADCALR in normal mode (CAL_EN = 0) is added to
each digital result from the ADC core.
22.2.6.1.3 Mid-Point Calibration
Because of its connections to the ADC’s reference voltage (VrefHi, VrefLo), the precision of the calibration
reference is voltage independent. On the other hand, the accuracy of the switched bridge resistor (R1 &
R2) relies on the manufacturing process deviation. Consequently, the mid-point voltage’s accuracy can be
affected due to the imperfections in the two resistors (expected mismatch error is around 1.5%).
The switched reference voltage device has been specially designed to support a differential measurement
of its mid-point voltage. This ensures the accuracy of the mid-point reference, and hence the efficiency of
the calibration.
The differential mid-point calibration is software controlled; the algorithm (voltage source measurements
and associated calculation) is inserted within the calibration software module included in the application
program.
The basic differential mid-point calibration flow is illustrated here after:
1. The application program connects the voltage VrefHi to R1 and VrefLo to R2, (BRIDGE_EN = 0,
HILO = 0), launches a conversion of the input voltage V(cal1), and stores the digital result D(cal1) into
the memory.
2. Then the application program switches the voltage VrefHi to R2 and VrefLo to R1 (BRIDGE_EN = 0,
HILO = 1), converts this new input voltage V(cal2) and again stores the issued digital result D(cal2)
into the memory.
3. The actual value of the real middle point is obtained by computing the average of these two results.
[D(cal1)+D(cal2)] /2;
summarizes the mid-point calibration flow.