I2C with restart - bq27510 fuel gauge IC - RM46

More info of my exercise is available at

I'm writing a blog on integrating the ti fuel gauge chips with Hercules microcontrollers.

The code didn't have the expected result. I'll have to do some more digging into the TRM.

One thing I was uncertain of when reading the example above, is that you've put the i2cSetStop after the i2cSetStart. In most examples I've seen, it is coming before i2cSetStart, just after i2cSetCount.

Note to self: It can't be too hard. I just want to replicate this exact scenario from the TRM with HALCoGen API:

I reviewed the TRM and Halcogen source code for one more time. Can you try the following?

i2cInit();
i2cSetMode(i2cREG1, I2C_MASTER);
i2cSetSlaveAdd(i2cREG1, Slave_Address);
i2cSetDirection(i2cREG1, I2C_TRANSMITTER);
i2cSetCount(i2cREG1 ,1);
i2cSetStart(i2cREG1); // generate the start phase, slave address, and R/W bit
i2cSend(i2cREG1, 1, TX_Data_Master); //send the cmd
i2cSetDirection(i2cREG1, I2C_RECEIVER);
i2cSetCount(i2cREG1, 2);
i2cSetStop(i2cREG1);   //set count and stop bit so that stop can be generated after receiving two bytes.
i2cSetStart(i2cREG1); // generate the start phase, slave address, and R/W bit
i2cReceive(i2cREG1, 2, RX_Data_Master); //read two bytes
while(i2cIsBusBusy(i2cREG1) == true); // wait for the reception to complete

RM needs to be 0. You can use the scope on the bus to see which portion of the waveform is not generated.

Thansk and regards,

Zhaohong

I have tried this, with RM=0.

What happens when I run that code,  is that indeed there is no stop condition, but the send part sends different data than what's expected.
We should send write address, and then 1 single value: 0x2e. Then we should get an ACK and a new start.
What happens is that we send a wrong value 0xff, then get an ACK, we send a wrong value 0xff again,  don't get an ACK, and get a new Start.

I was then expecting it would send the read address, but it sends the write address again after the restart.

Zhaohong, my scope isn't capable of capturing this info.

The capture with my logic analyze that I posted in the previous answer is te best I can give.

My end goal is to get the 100% completely same communication as in TRM for RM46, section 30.2.5.3:

I have been able to achieve this with a non-ti controller with I2C module, and got this capture on my analyzer:

This is exactly as what 30.2.5.3 documents.

As how to do this with Hercules and its I2C module, I'm currently at a loss.

Hi,

The first part of this post is a capture my almost working example - just that one stop condition too much. The second part is the capture of the latest code you have posted in the thread.

1: working, but with a STOP/START in stead of a restart:

Code:

#define Slave_Address  0x55

uint8_t TX_Data_Master[1] = { 0x2E};
uint8_t RX_Data_Master[2] = { 0 };



void bq27510Init() {

  int delay =0;

  ///////////////////////////////////////////////////////////////
  //        Master Transfer Functionality                      //
  ///////////////////////////////////////////////////////////////
  i2cInit();

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

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

  /* Configure Data count */
  i2cSetCount(i2cREG1, 1);

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

  /* Set Stop after programmed Count */
  i2cSetStop(i2cREG1);

  /* Transmit Start Condition */
  i2cSetStart(i2cREG1);

  /* Tranmit DATA_COUNT number of data in Polling mode */
  i2cSend(i2cREG1, 1, TX_Data_Master);

  /* Wait until Bus Busy is cleared */
  while(i2cIsBusBusy(i2cREG1) == true);

  /* Wait until Stop is detected */
  while(i2cIsStopDetected(i2cREG1) == 0);

  /* Clear the Stop condition */
  i2cClearSCD(i2cREG1);

  /* Simple Dealya before starting Next Block */
  /* Depends on how quick the Slave gets ready */
//  for(delay=0;delay<1000;delay++);


  ///////////////////////////////////////////////////////////////
  //        Master Receive Functionality                      //
  ///////////////////////////////////////////////////////////////


  /*****************************************/
  //// Start receving the data From Slave
  /*****************************************/

  /* wait until MST bit gets cleared, this takes
   * few cycles after Bus Busy is cleared */
  while(i2cIsMasterReady(i2cREG1) != true);

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

  /* Set direction to receiver */
  i2cSetDirection(i2cREG1, I2C_RECEIVER);

  /* Configure Data count */
  /* Note: Optional - It is done in Init, unless user want to change */
  i2cSetCount(i2cREG1, 2);

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

  /* Set Stop after programmed Count */
  i2cSetStop(i2cREG1);

  /* Transmit Start Condition */
  i2cSetStart(i2cREG1);

  /* Tranmit DATA_COUNT number of data in Polling mode */
  i2cReceive(i2cREG1, 2, RX_Data_Master);

  /* Wait until Bus Busy is cleared */
  while(i2cIsBusBusy(i2cREG1) == true);

  /* Wait until Stop is detected */
  while(i2cIsStopDetected(i2cREG1) == 0);

  /* Clear the Stop condition */
  i2cClearSCD(i2cREG1);


  asm(" nop");
  asm(" nop");
  asm(" nop");


}

Capture

Summary: there is a stop at step 5 that I want to not generate. The TI bq27510 IC expects a restart without a stop.

2: capture of the proposed test:

Code:

#define Slave_Address  0x55

uint8_t TX_Data_Master[1] = { 0x2E};
uint8_t RX_Data_Master[2] = { 0 };



void bq27510Init() {

	i2cInit();
	i2cSetMode(i2cREG1, I2C_MASTER);
	i2cSetSlaveAdd(i2cREG1, Slave_Address);
	i2cSetDirection(i2cREG1, I2C_TRANSMITTER);
	i2cSetCount(i2cREG1 ,1);
	i2cSetStart(i2cREG1); // generate the start phase, slave address, and R/W bit
	i2cSend(i2cREG1, 1, TX_Data_Master); //send the cmd 
	i2cSetDirection(i2cREG1, I2C_RECEIVER);
	i2cSetCount(i2cREG1, 2);
	i2cSetStop(i2cREG1);   //set count and stop bit so that stop can be generated after receiving two bytes.
	i2cSetStart(i2cREG1); // generate the start phase, slave address, and R/W bit
	i2cReceive(i2cREG1, 2, RX_Data_Master); //read two bytes
	while(i2cIsBusBusy(i2cREG1) == true); // wait for the reception to complete

}

Capture

Summary: this example generates a wrong communication. The first address is wrong (line 2: should be 0xaa), It's transmitting two datas (3 - 4 - 5 and 6) before the restart - these two data fields (3 and 5) are not what's in the transfer buffer, and a NACK is detected (6) after the second (unwanted) data. No final end condition generated (after 13).

I've also attached a CSV export of both captures.

2388.i2c.zip

Hi, the value RM is 'not set', in both cases.

(extract from i2c.h: (uint32)((uint32)0U << 7U) )

#define I2C_MDR_CONFIGVALUE       ((uint32)0x00000000U \
                                 | (uint32)((uint32)1U <<11U) \
                                 | (uint32)((uint32)1U <<10U) \
                                 | (uint32)((uint32)I2C_TRANSMITTER) \
                                 | (uint32)((uint32)I2C_7BIT_AMODE) \
                                 | (uint32)((uint32)0U << 7U) \
                                 | (uint32)((uint32)0U) \
                                 | (uint32)((uint32)I2C_8_BIT) \
								 | (uint32)I2C_RESET_OUT)

5684.I2C37_RAM_Poll.cI found a I2C test for accessing an I2C RAM. It is a validation test. Although it does not use Halcogen, I attach it here so that you can see what is required because this test code directly accesses the registers.  The I2C_Read function provides the functionality you are looking for.  In the mean time, I still need to figure out what is missing in the code sequence using Halcogen functions.

Thank you for the pointer to the memory example, Zhaohong Zhang. I will try to translate that to the register definitions of the RM46, and for the bq27510 - and run it to see if it works...

Hi Jan,

Have you progressed further with this issue?

In my case I am using TMS570LC43x Launchpad

In my case, I am interfacing with IRThermo sensor MLX90614; However, I need to find a way to stop the stop bit from being sent!

I have checked the TRM which seems to support that but when I try it, it doesn't really work.

Regards

Fouad

Fouad Hilly , I haven't been able to get this working.
It's not the end of the world in my case, because the sensor - even though the spec tells that it needs a RESTART - works with the STOP/START sequence.
I'd feel more comfortable though if I would be able to create that RESTART sequence.

Hi Jan,

Thank you for your reply,

I have dug further but in my case for the TMS570LC43, where I found few issues in the code generated by the HALCoGen, here is a link to a question with feedback to TI, which I have started: e2e.ti.com/.../1892569

I have posted some code as well, but as you are working on different MC, check the HL_i2c. and HL_i2c.h generated by the HALCoGen

Tips:

When you call i2cInit(), you are actually setting the stop bit! so you can either skip calling it, or you need to write it through registers.

The main thing I have noticed to make it to work, you need to set nIRS to "1", while the rest "0" then you start setting the rest of the bits, like master, start bit condition and so on.

The "count" is an important factor as well in the process, so you need to set it up with how many bytes to send for transmitting before you change the i2c to receiving.

When you start receiving, follow the same steps, but this time with "stop bit conditions set"

I hope this help

Regards

Fouad