RM48L952: Help required urgent for RM48L952 i2c driver is not working generated by halcogen.

Hello Murali,

Have you applied a scope to see what is happening on the bus? i.e., are you getting any signals to the pin?

Have you tried loopback mode?

Please insure the appropriate settings without conflicts in the PINMUX tab and also select the correct baud rate according to the required baud rate of the bus on which the device is to be placed. Have you observed your SCL line and are you seeing the appropriate clock rate?

What is the state of the other devices on the network? Note that in I2C the SDA line is driven low and floats high (thus the need for the pullup.) If you have another device on the bus that is driving low, the Hercules device will be unable to communicate on the line. This is a feature of I2C to allow slaves to extend the timing to accommodate slower devices on the network.

Additional information would be beneficial regarding your application as well. Is the Hercules device the master of slave on the bus?

Have you also worked through the examples provided in Halcogen --> Help --> Examples. I believe there are I2C examples included for both master and slave operation.

We are interfacing TMP112 I2C based temperature sensor with hercules RM48L952
1)I tried with scope we are not getting signal on both SDA and SCL lines.
2)I tried with loopback it is working.
3)I have tried with lower as well as higher baudrate.
4) only one device on I2C network.
5) We are operating on hercules MCU is master and TMP112 sensor is slave.
6) PINMUX I2C peripheral checked and unchecked mode is also tried. 10K Pullup resistor is connected to SDA and SCL lines.
7) With i2c interrupt and polling method is also tried still getting hang in below mentioned part of code.

It is getting struck in below function which is generated by Halcogen
void i2cSendByte(i2cBASE_t *i2c, uint8 byte)
while ((i2c->STR & (uint32)I2C_TX_INT) == 0U) { }

Code for referring as below


i2cInit();

i2cSetOwnAdd(i2cREG1,Master_Address);

i2cSetSlaveAdd(i2cREG1, Slave_Addr_write_tempSensor); /* Configure address of Slave to talk to */

i2cSetDirection(i2cREG1, I2C_TRANSMITTER); /* Set direction to Transmitter */ /* Note: Optional - It is done in Init */

i2cSetMode(i2cREG1, I2C_MASTER); /* Set mode as Master */

i2cSetStart(i2cREG1); /* Initiate Start condition for Transmission */

i2cSendByte(i2cREG1,Slave_Addr_write_tempSensor);
// while(i2cIsBusBusy(i2cREG1) == true); /* Wait until Bus Busy is cleared */
i2cSendByte(i2cREG1,CONFIG_BYTE1);
i2cSendByte(i2cREG1,CONFIG_BYTE2);
// i2cSetStop(i2cREG1);

i2cClearSCD(i2cREG1); /* Clear Stop Condition detect flag */

i2cSetDirection(i2cREG1, I2C_RECEIVER);
i2cSetMode(i2cREG1, I2C_MASTER); /* Set mode as Master */
// for(delay=0;delay<1000000;delay++); /* Simple Delay before starting Next Block */ /* Depends on how quick the Slave gets ready */

while(1)
{
i2cSetStart(i2cREG1);
i2cSendByte(i2cREG1,Slave_Addr_read_tempSensor);

// while(i2cIsBusBusy(i2cREG1) == true); /* Wait until Bus Busy is cleared */
i2cSendByte(i2cREG1,TEMPRATURE_REGISTER);
i2c_rcv_data=i2cReceiveByte(i2cREG1);
i2c_rcv_data=i2cReceiveByte(i2cREG1);
// i2cSetStop(i2cREG1);

i2cClearSCD(i2cREG1); /* Clear Stop Condition detect flag */
}


Thanks and Regards,
Murali Prasad

Hello Murali,

Murali Prasad said:

1)I tried with scope we are not getting signal on both SDA and SCL lines.
2)I tried with loopback it is working.

What type of loopback did you use? digital or analog? In either case, the fact that the code functions in loopback points to an issue with the Hardware. i.e., something is driving the signals other than the MCU. If Hercules is the Master, it should have control of SCL at a minimum so this signal should be clocking. If these signals are driven low always, there is something driving them that way (try disconnecting the external device and see what happens to SCL).

the other points have varying levels of signifigance to the issue you are seeing.

Murali Prasad said:

3)I have tried with lower as well as higher baudrate.

The external device should define which baud rate to use. However, given you aren't getting any SCL output at all, baud rate selection is not relevant to your problem at the moment.

Murali Prasad said:

4) only one device on I2C network.
5) We are operating on hercules MCU is master and TMP112 sensor is slave.

You need to understand what impact the slave device, TMP112, is having on your signal lines. Is it driving them low continuously? Is there a problem with the hookup of the TMP112? What are it's pull requirements?

Murali Prasad said:

6) PINMUX I2C peripheral checked and unchecked mode is also tried. 10K Pullup resistor is connected to SDA and SCL lines.

If you are using the I2C peripheral, you need to make sure the appropriate function is brought to the pin and there are no conflicts. Checking the I2C peripheral in the list at the top of the configuration page is the main step but you should confirm that the I2C SCL and SDA lines are checked as well. Also, if you are using a Launchpad, please double check that the specific pins you are using (as noted in the pinmux table) are the same ones that apply where you have connected the temp sensor on the Launchpad.

Murali Prasad said:

7) With i2c interrupt and polling method is also tried still getting hang in below mentioned part of code.

Since you have stated the SW is working in loopback mode, I would be doubtful that there are specific issues in your base I2C code and given that you do not see any output on the SDA or SCL lines, also leads me to believe this is either a pin mux configuration issue or a HW connection issue.

Hi Chuck,

Thanks for the response.

1) We have used digital loopback.
2) There is no conflict in Pinmux selection.
3) TMP112(Ti make) supports 100Khz to 2.85Mhz speed so we have tried I2C standard baudrate of 100Khz and 400Khz.
4) SDA and SCL lines will be in high state since we have pulled-up using 10K resisitors.
5) i2c read and write is function were written as per TMP112 waveform/Communication frame format.
6) If there is any deviation in the code kindly let me know.
7) We have checked TMP112 power supply inputs were ok.

If we use polling method and interrupt also firmware is hanging in below code.

void i2cSendByte(i2cBASE_t *i2c, uint8 byte)
{

/*SAFETYMCUSW 28 D MR:NA <APPROVED> "Potentially infinite loop found - Hardware Status check for execution sequence" */
while ((i2c->STR & (uint32)I2C_TX_INT) == 0U) <<<<<--------------------------It gets struck here,
{
} /* Wait */

i2c->DXR = (uint32)byte;

}

Above comment @ while loop says it is a potentially infinite loop that is why MCU is getting hangs here.

Kindly help us to resolve the issue as early as possible we have project deadlines.

Regards,
Murali Prasad

Murali,

If the software is working in loopback mode, why would you suspect any issues with the software? For further confirmation of the software, you can also try the analog loopback mode which will exercise also the analog buffers of the device.

Please double check the pins on your board that you are using. The I2C functions are on pins B2 and C3 (pins 71 and 72 on J11 expansion connector.)

To cross check the pinmux selection, you could select the I2C SDA and I2C SCL lines in non-functional mode (GPIO mode) and toggle them with software or through direct register manipulation in CCS. If you see the signals change on the C3 and B2 pins, then the PINMUX is set correctly.