Omega CN1166 manual Error and Exception Responses, CRC16 Calculation

G.Error and Exception Responses

If the controller receives a message which contains a corrupted character (parity check fail, framing error etc.), or if the CRC16 check fails, the controller ignores the message. If the message is otherwise syntactically flawed (e.g. the byte count or word count is incorrect) the controller will also not reply.

However, if the controller receives a syntactically correct message which nonetheless contains an illegal value, it will send an exception response, consisting of five bytes as follows:

Addr. Func. Exception No.	CRC16
	HI	LO

The Function Number field consists of the function number contained in the message which caused the error, with its top bit set (i.e. function 3 becomes 0x83), and the Exception Number is one of the codes contained in the following table:

Code	Name	Cause
1	ILLEGAL FUNCTION	Function Number out of range
2	ILLEGAL DATA ADDRESS	Parameter ID out of range or not supported
3	ILLEGAL DATA VALUE	Attempt to write invalid data/action not
		carried out
4	DEVICE FAILURE	N/A
5	ACKNOWLEDGE	N/A
6	BUSY	N/A
7	NEGATIVE	N/A
	ACKNOWLEDGE

H.CRC16 Calculation

This is a 16-bit cyclic redundancy checksum. It is calculated in accordance with a formula which involves recursive division of the data by a polynomial, with the input to each division being the remainder of the results of the previous one.

The formula specifies that input is treated as a continuous bit-stream binary number, with the most significant bit being transmitted first. However, the transmitting device sends the least significant bit first.

According to the formula, the dividing polynomial is 216 + 215 + 22 + 1 (Hex 18005), but this is modified in two ways:

i.Because the bit-order is reversed, the binary pattern is reversed also, making the MSB the rightmost bit, and

O077-4 Volume II 4-5