The parity bit, unlike the start and stop bits, is an optional parameter, used in serial communications to determine if the data character being transmitted is correctly received by the remote device.
The parity bit can have one of the following five specifications:
|none||Specifies that the local system must not create a parity bit for data characters being transmitted. It also indicates that the local system does not check for a parity bit in data received from a remote host.|
|even|| Specifies that the total number of binary 1s, in a single
character, adds up to an even number. If they do not, the parity bit
must be a 1 to ensure that the total number of binary 1s is even.
For example, if the letter a (binary 1100001) is transmitted under even parity, the sending system adds the number of binary 1s, which in this case is three, and makes the parity bit a 1 to maintain an even number of binary 1s. If the letter A (binary 1000001) is transmitted under the same circumstances, the parity bit would be a 0, thus keeping the total number of binary 1s an even number.
|odd||Operates under the same guidelines as even parity except that the total number of binary 1s must be an odd number.|
|space||Specifies that the parity bit will always be a binary zero. Another term used for space parity is bit filling, which is derived from its use as a filler for seven-bit data being transmitted to a device which can only accept eight bit data. Such devices see the space parity bit as an additional data bit for the transmitted character.|
|mark||Operates under the same guidelines as space parity except that the parity bit is always a binary 1. The mark parity bit acts only as a filler.|