Post by thread:[PIC] sn75176 problem

hi! i am making a serial connection rs485 between pic18f452 and pic16f628. The procedure is one send byte message the other get it and replies as an ok msg. This happens three times. Then the receiver becomes the sender and the same procedure occurs. Now, the program works if i connect RX and TX pins directly. But when i use sn75176, one for each pic, it doesn't work. I set the DeRe pin high before the message sending and clear it after sending, this is the nearly only difference at the code. No data transmission. If i put a big delay(150ms) after sending msg, two or three transmissions seems to happen. Then it is locked again. What can the problem be? Do you have a guess. I will send the codes when i prepare them for sending. Thanks! Note: I use no interrupts. Note2: I have put pullup res to the rx pins. -- View this message in context: www.nabble.com/sn75176-problem-tp24172473p24172473.html Sent from the PIC - [PIC] mailing list archive at Nabble.com.

chaconne escreveu: > hi! > i am making a serial connection rs485 between pic18f452 and pic16f628. > The procedure is one send byte message the other get it and replies as an ok > msg. This happens three times. Then the receiver becomes the sender and the > same procedure occurs. > Now, the program works if i connect RX and TX pins directly. But when i use > sn75176, one for each pic, it doesn't work. I set the DeRe pin high before > the message sending and clear it after sending, this is the nearly only > difference at the code. No data transmission. If i put a big delay(150ms) > after sending msg, two or three transmissions seems to happen. Then it is > locked again. What can the problem be? Do you have a guess. I will send the > codes when i prepare them for sending. > > Thanks! > > Note: I use no interrupts. > Note2: I have put pullup res to the rx pins. > Are you waiting for the last byte transmission to complete before disabling the transmitter? If you disable the transmitter right after writing the last byte to TXREG, then it will not be transmitted (at least not complete). You should wait some time also before the other device enable its transmitter. Regards, Isaac __________________________________________________ Faça ligações para outros computadores com o novo Yahoo! Messenger http://br.beta.messenger.yahoo.com/

Codes are like below. I have put a 150ms delay after TXREG = byte, one or two more transmission are done. The process as i observe is, picA sends one byte, picB gets it, picB send OK msg but picA doesnt get it, also picB gets one more message although there are no sent msg from picA. Note: I only send one byte in one transmission. Thanks. Code: // serLib.h #ifndef SERLIB_H_ #define SERLIB_H_ #endif /*SERLIB_H_*/ #define BAUD 9600 #define FOSC 4000000L #define NINE 0 /* Use 9bit communication? FALSE=8bit */ #define DIVIDER ((int)(FOSC/(16UL * BAUD) -1)) #define HIGH_SPEED 1 #if NINE == 1 #define NINE_BITS 0x40 #else #define NINE_BITS 0 #endif #if HIGH_SPEED == 1 #define SPEED 0x4 #else #define SPEED 0 #endif #define RX_PIN TRISB1 #define TX_PIN TRISB2 //void SendChar(unsigned char); unsigned char ReceiveChar(void); unsigned char getche(void); void init_comms(void); void init_comms() { TRISB1 = 1; TRISB2 = 0; SPBRG = DIVIDER; RCSTA = (NINE_BITS|0x90); TXSTA = (SPEED|NINE_BITS|0x20); //RCIE = 1; //GIE = 1; } void SendChar(unsigned char byte) { serialEnable = SEND_ENABLE; while(!TXIF) continue; TXREG = byte; serialEnable = RECEIVE_ENABLE; } unsigned char ReceiveChar() { return RCREG; } Code: // SerialProcesses.h void SerialProcesses() { char temp; if(serialMode == WAIT_MODE) { if(RCIF) { if(serialIndexTX == 3) { arrayDataOut[serialIndexRC] = ReceiveChar(); serialIndexRC++; if(serialIndexRC > 2) { serialIndexTX = 0; serialMode = SEND_MODE; } else { serialMode = SEND_MODE; b_sendOk = 1; } } else { if(ReceiveChar() == 55) { serialMode = SEND_MODE; } } RCIF = 0; } } else { if(b_sendOk == 1) { SendChar(55); serialMode = WAIT_MODE; b_sendOk = 0; } else { SendChar(arrayDataIn[serialIndexTX]); if(++serialIndexTX > 2) { serialIndexRC = 0; serialMode = WAIT_MODE; } else { serialMode = WAIT_MODE; } } } return; } Code: // StartUp.h void StartUp() { TRISA = 0; TRISB = 0b10000010; PORTA = 0; PORTB = 0; GIE = 0; init_comms(); return; } Code: // picA Main void main(void) { StartUp(); serialMode = SEND_MODE; serialIndexRC = 3; serialEnable = SEND_ENABLE; serialEnableReg = 1; while(1) { DataIn(); SerialProcesses(); DataOut(); } return; } Code: // picB Main void main(void) { StartUp(); serialMode = WAIT_MODE; serialIndexRC = 3; serialEnable = RECEIVE_ENABLE; serialEnableReg = 1; while(1) { DataIn(); SerialProcesses(); DataOut(); } return; } -- View this message in context: www.nabble.com/sn75176-problem-tp24172473p24174430.html Sent from the PIC - [PIC] mailing list archive at Nabble.com.

I am doing simulation at proteus and i didn't connect pullups to the A and B lines.I have connected pullup only to the RX pins. Is þt necessary at simulation? 2009/6/24 Dario Greggio <.....adpm.toKILLspam@spam@inwind.it> > Yusuf Özkay ha scritto: > > You are right, the problem is about frame error. But what is the reason > of > > it, i couldn't find it yet. > > > > > Well, it can be anything, from glitches at power suppplly, to glithes in > the state of 485 A-B lines when switching from TX to RX - btw, do you > have pull-up and pull-down at A and B? >

I put pullup and pulldown with 4k7 and I put 1k between A and B lines. It didn't work. I will try it in real simulation now. Hope it works. Thanks. 24 Haziran 2009 13:09 tarihinde Dario Greggio <adpm.toKILLspaminwind.it> yazdý: > Yusuf Özkay ha scritto: > > I am doing simulation at proteus and i didn't connect pullups to the A > and B > > lines.I have connected pullup only to the RX pins. Is þt necessary at > > simulation? > > > Oh well, i don't know: but, without them, A and B will float and the > receiver will get "spurious bytes" thus causing the RX errors... >

I tried at real simulation same thing happens.One byte is sent from A, that is received by B. B sends one byte, and that is received by A. A sends one byte but this byte is not received by B. No more communication happens. 24 Haziran 2009 13:20 tarihinde Yusuf Özkay <.....yusufozkayKILLspam.....gmail.com> yazdý: {Quote hidden}> I put pullup and pulldown with 4k7 and I put 1k between A and B lines. It > didn't work. I will try it in real simulation now. Hope it works. > Thanks. > > 24 Haziran 2009 13:09 tarihinde Dario Greggio <EraseMEadpm.tospam_OUTTakeThisOuTinwind.it> yazdý: > > Yusuf Özkay ha scritto: >> > I am doing simulation at proteus and i didn't connect pullups to the A >> and B >> > lines.I have connected pullup only to the RX pins. Is þt necessary at >> > simulation? >> >> >> Oh well, i don't know: but, without them, A and B will float and the >> receiver will get "spurious bytes" thus causing the RX errors... >> -

Yusuf, 2009/6/24 Yusuf Özkay <yusufozkayspam_OUTgmail.com>: > I tried at real simulation same thing happens.One byte is sent from A, that > is received by B. B sends one byte, and that is received by A. A sends one > byte but this byte is not received by B. No more communication happens. > Are you sure that the RX/TX switch is operating correctly? If the timing is too long between the end of the transmit packet and the transmitter switching off then you _could_ get the the effect above, It will depend on the amount of processing / packet verification going on as well. For example, in the above: if the second response from 'A' is just an acknowlegement, it could be sending it before B is ready to receive. It didn't happen the first time because 'A' had to process the data, calculate CRCs etc. Even if you have the RX permanently enabled, it's the TX switching on and off that is going to create the issue. RP