MLX90320_12 Datasheet, PDF (15/33 Pages) Melexis Microelectronic Systems – Automotive sensor interface

MLX90320
Automotive sensor interface
T₁ is the output of the temperature chain corresponding to the room temperature. The ADC of the
temperature chain outputs 10bits but 12 bits are stored (address 0 to 2 of the EEPROM). The MSB
must always be 0 and the other 11 bits are obtained from an average of the previous temperature
readings. This gives more accuracy to the output of the temperature chain. The T1 value used in
equation 2 is in the [0..1] range with an 11 bit resolution.
G₀ is the zero order fine gain coefficient (independent from the temperature) used to adjust accurately
the output span at room temperature. 12 bits are stored (address 12 to 14 of the EEPROM) but only
the 10 first are used. The two MSB must be 0. The G₀ value used in equation 2 is in the [0..1] range
with a 10 bit resolution.
G₁ is the first order fine gain coefficient used to compensate the sensor sensitivity drift with
temperature. 12 bits are stored (address 8 to 10 of the EEPROM). The MSB is the sign bit (two’s
complement): If G1[11] = 1 then G1 is negative, if G1[11] = 0 then G1 is positive. The G1 value used in
equation 2 is in the [-2..2] range with an 11 bit resolution.
G₂ is the second order fine gain coefficient used to compensate the sensor sensitivity drift with
temperature. 12 bits are stored (address 4 to 6 of the EEPROM). The MSB is the sign bit (two’s
complement): If G2[11] = 1 then G2 is negative, if G2[11] = 0 then G2 is positive. The G2 value used in
equation 2 is in the [-2..2] range with an 11 bit resolution.
The ALU computes equation 2 with 12 bits but the result is truncated to 10 bits because the Gain DAC is a 10 bit
DAC. When the MLX90320 is not able to compensate for the sensor sensitivity drift with temperature, the fine
gain calibration parameters stored in EEPROM will lead to a FNGain_real value out of the [0..1] range. In this case
the MLX90320 will indicate an overflow in the digital calculations by putting the output voltage in a fault band.
When this occurs, a reset of the chip is required to go back into the normal mode of operation. Typical total gains
with the corresponding sensor offset ranges that can be compensated can be found in table 3.
6.4.2 The Offset calibration of the sensor signal chain.
The purpose of the 7-bit offset DAC is to be able to adjust the MLX90320 output offset anywhere in the 0.5V to
4.5V range. The voltage at the output of the coarse offset DAC can be calculated by the formula:
CSOff _{ana log}
=
 ^{0.905 ×VDD}
−
CSOff _digital
127
× (0.905 ×V_DD
−
0.06
×
V
DD
)_{}
×
V_DD
5
Equation 3
Explanation of parameters used in equation 1:
CSOff_analog is the voltage at the output of the coarse offset DAC.
CSOff_digital is the digital decimal value of the coarse offset (7 bits stored in address 11 and 15 of the
EEPROM).
The voltage span at the output of the coarse offset DAC is large enough to allow the user to calibrate a
decreasing output characteristic with for example 4.5V as output offset and 0.5V as output full scale. This output
characteristic is only possible by inverting the inputs (setting the IINV bit).
Besides the programmable coarse offset, there is also a 10-bits programmable fine offset stage which allows
adjusting the MLX90320 output offset with a high resolution (at least a resolution of 0.1% of the supply voltage).
The voltage at the output of the fine offset DAC can be calculated by the formula:
3901090320
Rev 007
Page 15 of 33
Data Sheet
Feb/12