Writing data to UART in Python and reading them from C

Question

I am writing a byte to serial port using Python.

import serial

ser = serial.Serial ("/dev/ttyACM0")    
ser.baudrate = 115200                   
ser.write('\x57')
ser.close()

When I connect TX to RX I have no problem to read that byte (sent from Python code), using GtkTerm. But when I am trying to read this data on micro controller using C, I always read 240. But when I use GtkTerm to send hexadecimal data directly (View -> Send Hexadecimal data), I read (on microcontroller) appropriate value. What could be wrong?

C code:

char byte = getc_();
printf_("1 byte received: i: %i \n",byte);

get_c() function:

char getc_()
{
    #ifdef LIB_MUTEX
    mutex_lock(&mutex_getc_);
    #endif

    char res = uart_read();

    #ifdef LIB_MUTEX
    mutex_unlock(&mutex_getc_);
    #endif

    return res;
}

Parity, baudrate, startbits, stopbits, flow control. Are you reading the byte using the same UART when reading it with GtkTerm, btw? It would explain why it works, because the UART should always be configured correctly to decode what it encoded. — Ulrich Eckhardt
– Ulrich Eckhardt, Commented Dec 1, 2015 at 17:43
Check the error flags for the microcontroller's UART. Hopefully there's one for overflow and one for framing. If the framing one is set, you either have a bad baudrate or a signal problem. If the overflow is set, you're sending too much data. If you have access to an oscilloscope, probe the TX line from the PC and see if its any different in the two cases - python vs gtkterm. — Brian McFarland
– Brian McFarland, Commented Dec 1, 2015 at 18:51
You may see some junk from the UART when you first start using it. Protocols that work over serial ports need to put some effort into synchronizing the two ends and can't just assume that the first byte in on one side will be the first byte out on the other. — David Schwartz
– David Schwartz, Commented Dec 1, 2015 at 19:45

chux · Accepted Answer · 2015-12-01 18:42:49Z

1

Likely wrong communication rate. Change receiver/transmitting baud to maybe 1/4 or 1/6 -- Suggest 19200. (or speed up transmitting/reciver)

240 is 0xF0. With RS- 232, data is sent

Start bit - LS bit - next bit - ... -next bit - MS bit - Stop bit
// or 
0 - LS bit - next bit - ... -next bit - MS bit - 1
// or 0xF0
0 - 0 - 0 - 0 - 0 - 1 - 1 - 1 - 1 - 1

If the receiving end is seeing too many 0 bits (look at least significant bit first), and data received is in bit groups of 0's and 1's, it usually means data is sent at a slower baud than what the receiver is using.

Another clue to deciphering mis-match baud is what is the ratio of the count of bytes sent versus received. Given various bit patterns this is not a hard rule, but the side with more data is likely the one at the higher baud.

edited Dec 1, 2015 at 18:42

answered Dec 1, 2015 at 18:37

chux

158k17 gold badges160 silver badges311 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

user3053231 · Accepted Answer · 2015-12-02 16:44:15Z

0

After some time I find this solution. First byte I always receive 240, but with additional small sleeps I get the correct value which I sent.

import serial
import time
ser = serial.Serial (port = "/dev/ttyACM0", bytesize = 8, stopbits = 1)
ser.baudrate = 115200 
sleep_time = 0.05  
ser.write('\x41') #240
time.sleep(sleep_time)
ser.write('\x41') #right value A
time.sleep(sleep_time)
ser.write('\x41') #right value A
...
ser.close()

answered Dec 2, 2015 at 16:44

user3053231

Collectives™ on Stack Overflow

Writing data to UART in Python and reading them from C

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related