Finisar AN-2030 16a Unsigned fixed-point binary format for slopes, 16b Format for offsets

AN-2030: Digital Diagnostic Monitoring Interface for Optical Transceivers F i n i s a r

1The slope constants at addresses 76, 80,84, and 88, are unsigned fixed-point binary

2numbers. The slope will therefore always be positive. The binary point is in between

3the upper and lower bytes, i.e., between the eight and ninth most significant bits. The

4most significant byte is the integer portion in the range 0 to +255. The least significant

5byte represents the fractional portion in the range of 0.00391 (1/256) to 0.9961

6(255/256). The smallest real number that can be represented by this format is 0.00391

7(1/256); the largest real number that can be represented using this format is 255.9961

8(255 + 255/256). Slopes are defined, and conversion formulas found, in the “External

9Calibration” section. Examples of this format are illustrated below:

10	Table 3.16a: Unsigned fixed-point binary format for slopes
11
12
		Decimal	Binary Value		Hexadecimal Value
		Value	MSB	LSB	High Byte	Low Byte
		0.0000	00000000	00000000	00	00
		0.0039	00000000	00000001	00	01
		1.0000	00000001	00000000	01	00
		1.0313	00000001	00001000	01	08
		1.9961	00000001	11111111	01	FF
		2.0000	00000010	00000000	02	00
		255.9921	11111111	11111110	FF	FE
		255.9961	11111111	11111111	FF	FF

13
14	The calibration offsets are 16-bit signed twos complement binary numbers. The offsets
15
16	are defined by the formulas in the “External Calibration” section. The least significant bit
17	represents the same units as described above under “Internal Calibration” for the
18	corresponding analog parameter, e.g., 2μA for bias current, 0.1μW for optical power,
19	etc. The range of possible integer values is from +32767 to -32768. Examples of this
20	format are shown below.
21
22			Table 3.16b: Format for offsets
23
		Decimal	Binary Value		Hexadecimal Value
		Value	High Byte	Low Byte	High Byte	Low Byte
		+32767	01111111	1111111	7F	FF
		+3	00000000	00000011	00	03
		+2	00000000	00000010	00	02
		+1	00000000	00000001	00	01
		0	00000000	00000000	00	00
		-1	11111111	11111111	FF	FF
		-2	11111111	11111110	FF	FE
		-3	11111111	11111101	FF	FD
		-32768	10000000	00000000	80	00
24

25External calibration of received optical power makes use of single-precision floating-

26point numbers as defined by IEEE Standard for Binary Floating-Point Arithmetic, IEEE

27Std 754-1985. Briefly, this format utilizes four bytes (32 bits) to represent real

28numbers. The first and most significant bit is the sign bit; the next eight bits indicate an

29exponent in the range of +126 to –127; the remaining 23 bits represent the mantissa.

30The 32 bits are therefore arranged as in Table 3.16c below.

9/26/02 Revision D Page 22