STM32 I2C Slave (SMBUS) mode software reference design

Everyone is familiar with I2C, which has two lines, CLK and DATA. I believe everyone is more familiar with stm32's I2C. It uses the writing controller method, and the controller directly completes the I2C timing operation. Users do not need to care about the specific logic generated. However, in most cases, the I2C Master mode is used, that is, the master device mode, and it is rarely used as the slave mode. This article talks about how to use stm32 I2C as a slave mode. More strictly speaking, this article talks about the smbus mode.

From the official website stm32 manual, we found the difference between smbus and I2C, you can understand it by yourself:

When actually using it, you can set and use smbus as I2C. From the code, except that the I2C Clock is set to 20K, there is no obvious difference. The following is the setting of I2C smbus mode:

void IIC_Init(void)  

{  

  GPIO_InitTypeDef GPIO_InitStructure;  

  I2C_InitTypeDef I2C_InitStructure;  

//  RCC_APB2PeriphClockCmd  (RCC_APB2Periph_GPIOB , ENABLE);  

   /* Configure I2C1 pins: SCL and SDA */   

   GPIO_InitStructure.GPIO_Pin =   GPIO_Pin_6|GPIO_Pin_7; //6:SCL 7:SDA  

   GPIO_InitStructure.GPIO_Speed = GPIO_Speed_50MHz;  

   GPIO_InitStructure.GPIO_Mode = GPIO_Mode_AF_OD;//Out_PP;  

   GPIO_Init(GPIOB, &GPIO_InitStructure);   

//   RCC_APB1PeriphClockCmd  (RCC_APB1Periph_I2C1, ENABLE);   

   // Initialize I2C1 clock

   // Reset I2C1 device clock in order to avoid non-cleared error flags

   //__I2C_RCC_RESET(CPAL_I2C_CLK [Device]);

   RCC_APB1PeriphResetCmd(RCC_APB1Periph_I2C1,ENABLE);

   RCC_APB1PeriphResetCmd(RCC_APB1Periph_I2C1,DISABLE);

   // Enable I2C1 device clock

   RCC_APB1PeriphClockCmd(RCC_APB1Periph_I2C1,ENABLE);

   I2C_DeInit(I2C1);  

   I2C_InitStructure.I2C_Mode = I2C_Mode_SMBusDevice;  

   I2C_InitStructure.I2C_DutyCycle = I2C_DutyCycle_2;  

   I2C_InitStructure.I2C_OwnAddress1 = 0x76;//slave addr  

   I2C_InitStructure.I2C_Ack = I2C_Ack_Enable;  

   I2C_InitStructure.I2C_AcknowledgedAddress = I2C_AcknowledgedAddress_7bit;  

   I2C_InitStructure.I2C_ClockSpeed ​​= 20000;//20K

   I2C_Init(I2C1, &I2C_InitStructure);  

   I2C_ITConfig(I2C1, I2C_IT_EVT|I2C_IT_BUF, ENABLE);  

   I2C_ITConfig(I2C1, I2C_IT_ERR, ENABLE);

   I2C_Cmd(I2C1,ENABLE);  

}  

void NVIC_Configuration(void)  

{  

  NVIC_InitTypeDef NVIC_InitStructure;  

     /* Configure and enable I2Cx event interrupt -------------------------------*/  

    NVIC_InitStructure.NVIC_IRQChannel = I2C1_EV_IRQn;  

    NVIC_InitStructure.NVIC_IRQChannelPreemptionPriority = 0;  

    NVIC_InitStructure.NVIC_IRQChannelSubPriority = 2;  

    NVIC_InitStructure.NVIC_IRQChannelCmd = ENABLE;  

    NVIC_Init(&NVIC_InitStructure);  

}  

From the above code, we can see that I2C initializes I2C IO mapping, working mode I2C_Mode_SMBusDevice, slave address 0x76 (customizable), I2C Clock, I2C interrupt vector and other parameters. After the settings are completed, whenever I2C generates data, an IRQ interrupt will be generated and the interrupt processing function I2C1_EV_IRQn will be entered, as shown below:

void I2C1_EV_IRQHandler(void)  

{  

uint32_t  I2CFlagStatus;  

static uint8_t IIC_Data=0;

I2CFlagStatus = I2C_GetLastEvent(I2C1);  // =>  (SR2<<16|SR1)  

// print_info("I2CFlagStatus=0x%x rn",I2CFlagStatus);

IIC1_EV_INT_Flag=1;

IIC_Get_Data_Flag=1;

if ((I2CFlagStatus & I2C_SR1_ADDR) != 0)//bit1:addr matched  

{

//print_info("4rn");

if(I2CFlagStatus & I2C_SR1_TXE) //bit7 Data register empty (transmitters)  

{//read          

   Rx_Idx=0;

   Tx_Idx = 0;    

   I2C_SendData(I2C1, I2C_Buffer_Tx[Tx_Idx]);

}

else

{   

}  

}

else if((I2CFlagStatus & I2C_SR1_RXNE) != 0)//bit6  RxNE    -Data register not empty (receivers))  

{

IIC_Data=I2C_ReceiveData(I2C1);

I2C_Buffer_Rx[Frame_Idx][Rx_Idx] = IIC_Data;

Rx_Idx++;

}

else if((I2CFlagStatus & I2C_SR1_STOPF) != 0)//bit4  STOPF -Stop detection (slave mode)  

{

//I2C_Buffer_Rx[0] = num-1;      

I2C1->CR1 |= 0x1000;//CR1_PEC_Set;  

Rx_Idx_t=Rx_Idx;

Rx_Idx=0;

Frame_Idx++;

I2C1->SR1=0;  

I2C1->SR2=0;  

}

else

{  

}  

}  

In the interrupt handling function, after an interrupt is generated, the flag bit IIC_Get_Data_Flag is set to 1, and the current I2C status is obtained according to the read I2CFlagStatus status bit, and the data is obtained. Then, in the main function, the data generated by I2C is processed according to the flag bit. After the processing is completed, the IIC_Get_Data_Flag flag bit is set to 0. This is a good way to process data. (PS: The interrupt handling function cannot handle too complex and time-consuming transactions, so as not to affect the real-time nature of the interrupt response, so the data processing process is placed in the main function)