SLAS748G March 2011 – January 2024 DAC3482
PRODUCTION DATA
The block parity method uses the FRAME signal to determine the boundaries of the data block to compute parity. This mode is enabled by setting the frame_parity_ena bit in register config1.
A low-to-high transition of FRAME captured with the DATACLK rising edge determines the end point of the parity block and the beginning of the next one. In this method the parity bit of the completed block corresponds to the FRAME value captured on the DATACLK falling edge right after the STOP/START point.
The input parity value is defined to be the total number of logic 1s in the data block. A logic HIGH captured on the falling edge of DATACLK indicates odd parity or odd number of logic 1s, while a logic LOW indicates even parity or even number of logic 1s. If the expected parity does not match the number of logic 1s in the received data, then alarm_frame_parity in register config5 will be set to 1b. The main advantage of the block parity mode is that there is no need for an additional parity LVDS input.
Since the FRAME signal is used for parity testing in addition to FIFO syncing and frame boundary assignment, it is mandatory to take some extra steps to avoid device malfunction. If FRAME is used to reset the FIFO pointers continuously, the block size must be a multiple of 8 samples (each sample corresponding to 16-bits I and 16-bits Q).
In addition, the use of block parity in byte-wide input data mode requires the following steps: