COIC5130A Datasheet, PDF (8/20 Pages) List of Unclassifed Manufacturers – Programmable Reed-Solomon Error Correction Encoder and Decoder

COic5130A Specifications
8
Co^~Optic
Message Decoding
During message decoding the device clocks in the encoded message block from the DIN7 - DIN0 input bus on the rising edge of CLK. The
device then computes polynomials for this data using a high performance Reed-Solomon coding algorithm, and then proceeds to use the
calculated polynomials to correct any errors, if possible. These polynomials are a series of complex equations that will indicate not only the
position of incorrect bytes but will also produce the necessary information to correct them. The timing of the decoding operation is defined
Decoder Message Output Timing
DOUT7 - DOUT0
K Data Bytes
R Parity Bytes
DATARDY
CORR
The action taken by the decoder with respect to a given message block is determined in all cases by the quantity of errors received. The
device reports its action via the CORR output pin. If P = R, a HIGH output on the CORR pin indicates that the message block being output
is correct.A LOW output indicates that a correction was not performed because there were too many errors or no errors. In situations where
P has been chosen to be not equal to R, the meaning of the UNCORR pin is that more than P/2 errors have occurred.
Decoder Optional Status Bytes
The optional status bytes, that may be output at the end of the message block, indicate the number of errors encountered and may be used
to determine the exact meaning of the CORR output when P ≠ R. The first status byte has all of its bits set LOW except bit 7, which is the
CA (correction attempted) flag. The CA flag goes LOW when E ≤ R/2 (the message block was successfully decoded). Therefore, anytime CA
is HIGH, the message block should be retransmitted since there were probably too many errors to reliably correct.Also, if P = R then CA will
always be LOW when the CORR output pin is HIGH and HIGH when CORR is LOW.
In the second status byte, bits 4 through O are a binary count of the number of errors encountered. Bit 4 is most significant. The remaining
bits are always LOW.