Abov A96G166 - 0% Error Baud Rate; Figure 80. 0% Error Baud Rate Block Diagram

A96G166/A96A166/A96S166 User’s manual 15. USART 0/1
161
15.14 0% error baud rate
USART system of A96G166/A96A166/A96S166 supports floating point counter logic for 0% error of
baud rate. By using 8-bit floating point counter logic, cumulative error to below the decimal point can
be removed.
Floating point counter value is defined by baud rate error. In the baud rate formula, BAUD is
presented in the integer count value. For example, If you want to use the 57600 baud rate (fXIN =
16MHz), integer count value must be 16.36 value (BAUD+1 = 16000000/(16×57600) = 17.36). Here,
the accurate BAUD value is 16.36. To achieve the 0% error of baud rate, floating point counter value
must be 164 ((17-16.36) x 256 ≒ 164) and BAUD value must be 17. Namely you have to write the 164
(decimal number) in USART_FPCR and 17 (decimal number) in USART_BAUD.
Figure 80. 0% Error Baud Rate Block Diagram
Integer count value
Integer count value - 1
8bit Max floating point
count value
8bit floating point
counter
TXD clock
Generator
0
1
0% Error
Baud rate