7.3.11.2 Temperature and Nonlinearity Compensation
The PGA300 implements a third-order TC and NL compensation of the bridge offset, bridge span, and bridge nonlinearity. Equation 4 shows the digital temperature and nonlinearity compensation algorithm implemented in the PGA300:
Equation 4. DAC = (h
0 + h
1T + h
2T
2 + h
3T
3) + (g
0 + g
1T + g
2T
2 + g
3T
3) × P + (n
0 + n
1T + n
2T
2 + n
3T
3) × P
2 + (m
0 + m
1T + m
2T
2 + m
3T
3) × P
3
where
- DAC = Digitally compensated value at the input of the DAC
- hx, gx, nx and mx are the TC and NL compensation coefficients programmed in EEPROM
- P is the offset and gain compensated pressure value
- T is the offset and gain compensated temperature value
Equation 4 has 16 coefficients, and therefore requires at least 16 different measurement points to compute a unique set of 16 coefficients.
The 16 different P ADC and T ADC measurements can be made, for example, at four temperatures and at four different pressures:
- The P Gain and T Gain values must be set to a fixed value for all measurements.
- At each measurement point, the P ADC value and the T ADC value must be recorded in order to compute the 16 coefficients.
- Measuring P ADC and T ADC at different temperatures and pressures may sometimes be expensive. In this case, there are three approaches:
- Use a model of the bridge to estimate the P ADC and T ADC measurements instead of actually measuring them.
- Use batch modeling, in which a family of sense elements is characterized across temperature, and the TC coefficients of the compensation equation are determined prior to calibration. On a production line, measurements are made at a limited number of temperature and pressure set points, and coefficients are adjusted accordingly. Discuss details with TI application engineers on the E2E community.
- Reduce the number of coefficients by reducing the order of TC compensation. Discuss the procedure to use fewer coefficients with TI application engineers on the E2E community.