187

2545M–AVR–09/07

ATmega48/88/168

Figure 19-5. Start Bit Sampling

When the clock recovery logic detects a high (idle) to low (start) transition on the RxDn line, the

start bit detection sequence is initiated. Let sample 1 denote the first zero-sample as shown in

the figure. The clock recovery logic then uses samples 8, 9, and 10 for Normal mode, and sam-

ples 4, 5, and 6 for Double Speed mode (indicated with sample numbers inside boxes on the

figure), to decide if a valid start bit is received. If two or more of these three samples have logical

high levels (the majority wins), the start bit is rejected as a noise spike and the Receiver starts

looking for the next high to low-transition. If however, a valid start bit is detected, the clock recov-

ery logic is synchronized and the data recovery can begin. The synchronization process is

repeated for each start bit.

19.8.2 Asynchronous Data Recovery

When the receiver clock is synchronized to the start bit, the data recovery can begin. The data

recovery unit uses a state machine that has 16 states for each bit in Normal mode and eight

states for each bit in Double Speed mode. Figure 19-6 shows the sampling of the data bits and

the parity bit. Each of the samples is given a number that is equal to the state of the recovery

unit.

Figure 19-6. Sampling of Data and Parity Bit

The decision of the logic level of the received bit is taken by doing a majority voting of the logic

value to the three samples in the center of the received bit. The center samples are emphasized

on the figure by having the sample number inside boxes. The majority voting process is done as

follows: If two or all three samples have high levels, the received bit is registered to be a logic 1.

If two or all three samples have low levels, the received bit is registered to be a logic 0. This

majority voting process acts as a low pass filter for the incoming signal on the RxDn pin. The

recovery process is then repeated until a complete frame is received. Including the first stop bit.

Note that the Receiver only uses the first stop bit of a frame.

Figure 19-7 shows the sampling of the stop bit and the earliest possible beginning of the start bit

of the next frame.

12345678 9 10 11 12 13 14 15 16 12

STARTIDLE

BIT 0

1234 5 678120

RxD

Sample

(U2X = 0)

Sample

(U2X = 1)

12345678 9 10 11 12 13 14 15 16 1

BIT n

1234 5 6781

RxD

Sample

(U2X = 0)

Sample

(U2X = 1)

188

2545M–AVR–09/07

ATmega48/88/168

Figure 19-7. Stop Bit Sampling and Next Start Bit Sampling

The same majority voting is done to the stop bit as done for the other bits in the frame. If the stop

bit is registered to have a logic 0 value, the Frame Error (FEn) Flag will be set.

A new high to low transition indicating the start bit of a new frame can come right after the last of

the bits used for majority voting. For Normal Speed mode, the first low level sample can be at

point marked (A) in Figure 19-7. For Double Speed mode the first low level must be delayed to

(B). (C) marks a stop bit of full length. The early start bit detection influences the operational

range of the Receiver.

19.8.3 Asynchronous Operational Range

The operational range of the Receiver is dependent on the mismatch between the received bit

rate and the internally generated baud rate. If the Transmitter is sending frames at too fast or too

slow bit rates, or the internally generated baud rate of the Receiver does not have a similar (see

Table 19-2) base frequency, the Receiver will not be able to synchronize the frames to the start

bit.

The following equations can be used to calculate the ratio of the incoming data rate and internal

receiver baud rate.

D Sum of character size and parity size (D = 5 to 10 bit)

S Samples per bit. S = 16 for Normal Speed mode and S = 8 for Double Speed

mode.

First sample number used for majority voting. S

= 8 for normal speed and S

= 4

for Double Speed mode.

Middle sample number used for majority voting. S

= 9 for normal speed and

= 5 for Double Speed mode.

slow

is the ratio of the slowest incoming data rate that can be accepted in relation to the

receiver baud rate. R

fast

is the ratio of the fastest incoming data rate that can be

accepted in relation to the receiver baud rate.

Table 19-2 and Table 19-3 list the maximum receiver baud rate error that can be tolerated. Note

that Normal Speed mode has higher toleration of baud rate variations.

12345678 9 10 0/1 0/1 0/1

STOP 1

1234 5 6 0/1

RxD

Sample

(U2X = 0)

Sample

(U2X = 1)

(A) (B) (C)

slow

+()S

S 1– DS⋅ S

---------------------------------------

--- -

= R

fast

+()S

D 1+()SS

--------------------------------

---

189

2545M–AVR–09/07

ATmega48/88/168

The recommendations of the maximum receiver baud rate error was made under the assump-

tion that the Receiver and Transmitter equally divides the maximum total error.

There are two possible sources for the receivers baud rate error. The Receiver’s system clock

(XTAL) will always have some minor instability over the supply voltage range and the tempera-

ture range. When using a crystal to generate the system clock, this is rarely a problem, but for a

resonator the system clock may differ more than 2% depending of the resonators tolerance. The

second source for the error is more controllable. The baud rate generator can not always do an

exact division of the system frequency to get the baud rate wanted. In this case an UBRRn value

that gives an acceptable low error can be used if possible.

19.9 Multi-processor Communication Mode

Setting the Multi-processor Communication mode (MPCMn) bit in UCSRnA enables a filtering

function of incoming frames received by the USART Receiver. Frames that do not contain

address information will be ignored and not put into the receive buffer. This effectively reduces

the number of incoming frames that has to be handled by the CPU, in a system with multiple

MCUs that communicate via the same serial bus. The Transmitter is unaffected by the MPCMn

setting, but has to be used differently when it is a part of a system utilizing the Multi-processor

Communication mode.

If the Receiver is set up to receive frames that contain 5 to 8 data bits, then the first stop bit indi-

cates if the frame contains data or address information. If the Receiver is set up for frames with

Table 19-2. Recommended Maximum Receiver Baud Rate Error for Normal Speed Mode

(U2Xn = 0)

# (Data+Parity Bit) R

slow

(%) R

fast

(%) Max Total Error (%)

Recommended Max

Receiver Error (%)

5 93.20 106.67 +6.67/-6.8 ± 3.0

6 94.12 105.79 +5.79/-5.88 ± 2.5

7 94.81 105.11 +5.11/-5.19 ± 2.0

8 95.36 104.58 +4.58/-4.54 ± 2.0

9 95.81 104.14 +4.14/-4.19 ± 1.5

10 96.17 103.78 +3.78/-3.83 ± 1.5

Table 19-3. Recommended Maximum Receiver Baud Rate Error for Double Speed Mode

(U2Xn = 1)

# (Data+Parity Bit) R

slow

(%) R

fast

(%) Max Total Error (%)

Recommended Max

Receiver Error (%)

5 94.12 105.66 +5.66/-5.88 ± 2.5

6 94.92 104.92 +4.92/-5.08 ± 2.0

7 95.52 104,35 +4.35/-4.48 ± 1.5

8 96.00 103.90 +3.90/-4.00 ± 1.5

9 96.39 103.53 +3.53/-3.61 ± 1.5

10 96.70 103.23 +3.23/-3.30 ± 1.0

Part #	ATMEGA48-20AU
Description	MCU 8BIT ATMEGA RISC 4KB FLASH 3.3V/5V 32TQFP - Trays
Category	IC
Availability	Out of Stock
Qty	0

ATMEGA48-20AU

Related Items

Technical Document