Texas Instruments MSP430x1xx - Manual (Page 241)

Baud Rate Considerations
12-25
USART Peripheral Interface, UART Mode
12.6.3 Support of Multiprocessor Modes for Reduced Use of MSP430 Resources
Communication systems can use multiprocessor modes with multiple-
character idle-line or address-bit protocols. The first character can be a target
address, a message identifier, or can have another definition. This character
is interpreted by the software and, if it is of any significance to the application,
the succeeding characters are collected and further activities are defined. An
insignificant first character would stop activity for the processing device. This
application is supported by the wake-up interrupt feature in the receive
operation, and sends wake-up conditions along with a transmission. Avoiding
activity on insignificant characters reduces consumption of MSP430
resources and the system can remain in the most efficient power-conserving
mode.
In addition to the multiprocessor modes, rejecting erroneous characters saves
MSP430 resources. This practice prevents interrupt handling of the erroneous
characters. The processor waits in the most efficient power-conserving mode
until a character is processed.
12.7 Baud Rate Considerations
The MSP430 baud-rate generator uses a divider and a modulator. A given
crystal frequency and a required baud rate determines the required division
factor N:
N =
BRCLK
baud rate
The required division factor N usually has an integer part and a fraction. The
divider in the baud rate generator realizes the integer portion of the division
factor N, and the modulator meets the fractional part as closely as possible.
The factor N is defined as:
N
+
UBR
) 1
n
S
n–1
i + 0
mi
Where:
N: Target division factor
UBR: 16-bit representation of registers UBR1 and UBR0
i: Actual bit in the frame
n: Number of bits in the frame
m
i
: Data of the actual modulation bit
Baud rate +
BRCLK
N
+
BRCLK
UBR ) 1
n
ȍ n * 1
i + 0
m
i
12.7.1 Bit Timing in Transmit Operation
The timing for each individual bit in one frame or character is the sum of the
actual bit timings as shown in Figure 12–27. The baud-rate generation error
shown in Figure 12–28 in relation to the required ideal timing, is calculated for
each individual bit. The relevant error information is the error relative to the
actual bit, not the overall relative error.