In the datasheet of the AD7134 I cannot find the exact representation of the gain calibration value. Is this 20 bit value also a signed twos complement value (similar to the offset calibration value)?
Thanks. So I conclude that GAIN_CHx is unsigned and the resulting gain equals to 0.5+(GAIN_CHx)*(0.95/1e6), with GAIN_CHx the decimal representation of the 20bits unsigned value. The 0.95ppm value is likely just 1/(2^20) = 0.953674ppm. So that would make the formula then to be 0.5+(GAIN_CHx)/(2^20) if GAIN_CHx is represented as a unsigned decimal number or 0.5+ GAIN_CHx when assuming GAIN_CHx to be a fixed point number with Wordlength = 20 and Fraction length = 20).