ADC calibration can be performed using the ADC calibration commands or by performing manual calibration. The calibration procedure is as follows:

1. Select the PGA or buffer operation, input channel, and PGA gain condition for calibration.
2. Preset the OFFSET register = 000000h and the GAIN register = 400000h.
3. Disable the high-pass filter for offset calibration. Short the inputs to the system, or use the input MUX to provide the input short. A system-level input short can yield more accurate calibration. After the input settles, either send the OFSCAL command or perform a manual calibration.
   1. OFSCAL command. After the command is sent, DRDY is driven low 81 conversion periods later to indicate calibration is complete. The OFFSET register is updated with the new calibration value. As shown in Figure 8-21, the first data output uses the new OFFSET value.
   2. Manual calibration. Wait at least 64 conversions for the digital filter to settle then average a number of data points to improve calibration accuracy. Write the value to the 24-bit OFFSET register.
4. Apply a gain calibration voltage. After the input settles, either send the GANCAL command or perform a manual calibration.
   1. GANCAL command. Apply a positive dc full-scale calibration voltage. After the command is sent, DRDY is driven low 81 conversion periods later to indicate calibration is complete. The ADC calculates GAIN such that the full-scale code is equal to the applied calibration signal. As shown in Figure 8-21, the first data output uses the new GAIN value.
   2. Manual calibration. Apply an ac signal coherent to the sample rate or dc calibration signal that are slightly below full-scale (for example, 2.4 V for gain = 1). Using a calibration signal less than full-scale range prevents clipped output codes that otherwise lead to incorrect calibration. Wait 64 conversions for the digital filter to settle then average a number of data points to improve calibration accuracy. For ac-signal calibration, use a number of coherent signal periods to compute the RMS value.

      Equation 7 computes the value of GAIN for manual calibration.

      Equation 7.

Figure 8-21 Calibration Command