The digital output from each temperature measurement is stored in the read-only temperature register. The temperature register of the TMP102 device is configured as a 12-bit, read-only register (configuration register EM bit = 0, see the Extended Mode (EM) section), or as a 13-bit, read-only register (configuration register EM bit = 1) that stores the output of the most recent conversion. Two bytes must be read to obtain data and are listed in Table 6-8 and Table 6-9. Byte 1 is the most significant byte (MSB), followed by byte 2, the least significant byte (LSB). The first 12 bits (13 bits in extended mode) are used to indicate temperature. The least significant byte does not have to be read if that information is not needed. The data format for temperature is summarized in Table 6-2 and Table 6-3. One LSB equals 0.0625°C. Negative numbers are represented in binary twos-complement format. Following power-up or reset, the temperature register reads 0°C until the first conversion is complete. Bit D0 of byte 2 indicates normal mode (EM bit = 0) or extended mode (EM bit = 1) , and can be used to distinguish between the two temperature register data formats. The unused bits in the temperature register always read 0.

Table 6-2 does not list all temperatures. Use the following rules to obtain the digital data format for a given temperature or the temperature for a given digital data format.

To convert positive temperatures to a digital data format:

1. Divide the temperature by the resolution
2. Convert the result to binary code with a 12-bit, left-justified format, and MSB = 0 to denote a positive sign.

   Example: (50°C) / (0.0625°C / LSB) = 800 = 320h = 0011 0010 0000

To convert a positive digital data format to temperature:

1. Convert the 12-bit, left-justified binary temperature result, with the MSB = 0 to denote a positive sign, to a decimal number.
2. Multiply the decimal number by the resolution to obtain the positive temperature.

   Example: 0011 0010 0000 = 320h = 800 × (0.0625°C / LSB) = 50°C

To convert negative temperatures to a digital data format:

1. Divide the absolute value of the temperature by the resolution, and convert the result to binary code with a 12-bit, left-justified format.
2. Generate the twos complement of the result by complementing the binary number and adding one. Denote a negative number with MSB = 1.

   Example: (|–25°C|) / (0.0625°C / LSB) = 400 = 190h = 0001 1001 0000

   Two's complement format: 1110 0110 1111 + 1 = 1110 0111 0000

To convert a negative digital data format to temperature:

1. Generate the twos compliment of the 12-bit, left-justified binary number of the temperature result (with MSB = 1, denoting negative temperature result) by complementing the binary number and adding one. This represents the binary number of the absolute value of the temperature.
2. Convert to decimal number and multiply by the resolution to get the absolute temperature, then multiply by –1 for the negative sign.

   Example: 1110 0111 0000 has twos compliment of 0001 1001 0000 = 0001 1000 1111 + 1

   Convert to temperature: 0001 1001 0000 = 190h = 400; 400 × (0.0625°C / LSB) = 25°C = (|–25°C|); (|–25°C|) × (–1) = –25°C