STM32F0 Discovery I2C

[dflogeras]

Hi,

Chibi-3.0.3, and STM32F072 DISCO board. This is my first time dealing with the i2c-v2 driver, so I thought I'd give it a try.

I read the RM for my part and found the tables for I2C timings. Since my project has

Code: Select all

#define STM32_I2C1SW                        STM32_I2C1SW_HSI

The I2C freq should be 8MHz.

After looking that up in the table, and I wish to support 100kHz I2C mode, I ended up with:

Code: Select all

static const I2CConfig i2c_cfg_1 = {
  STM32_TIMINGR_PRESC(0x01U) |
  STM32_TIMINGR_SCLDEL(0x04U) | STM32_TIMINGR_SDADEL(0x02U) |
  STM32_TIMINGR_SCLH(0x0fU)  | STM32_TIMINGR_SCLL(0x13U),
  0,
  0
};

Sidenote, the example in testhal/STM32/STM32F37x/I2C that I have found using i2c-v2 driver is a bit confusing. The comment says one thing, then there is a I2C configuration which is commented out, and another replacing it, none of which seems to agree with the comment, or the manuals.

Anyway, I set up my two SDA/SCL pins as per the RM/Datasheet, and configure the device:

Code: Select all

  palSetPadMode(GPIOB, 8, PAL_MODE_ALTERNATE(1) | PAL_STM32_OTYPE_OPENDRAIN );
  palSetPadMode(GPIOB, 9, PAL_MODE_ALTERNATE(1) | PAL_STM32_OTYPE_OPENDRAIN );
  i2cInit();
  i2cStart( &I2CD1, &i2c_cfg_1 );

However, a subsequent call to i2cMasterTransmitTimeout() returns MSG_RESET. I'm not sure why, or how to proceed. I have all DBG defines enabled and nothing is causing an error/stack overflow that I can detect.

The I2C hardware works, as I began this project on a STM32L DISCOVERY, and am porting it to the F0. The only major differences are the I2C config/setup code. I can see on the scope that the SCL and SDA lines are attempting to communicate, but something must be different in the timing perhaps?

[Giovanni]

Hi,

Those configuration fields should be taken from the reference manual, the comment probably is copied by the original demo for the F373 and is not correct.

MSG_RESET is documented as:

Code: Select all

 * @retval MSG_RESET    if one or more I2C errors occurred, the errors can
 *                      be retrieved using @p i2cGetErrors().

You need to verify the nature of the error.

Giovanni

[dflogeras]

Hmm, between re-testing with the STM32L discovery, adding the i2cGetErrors() calls, and hooking up my logic analyzer, the problem went away. I'm guessing I may have had a loose connection to the devkit. My code is in version control and I don't see anything that I am doing differently/changed since my first post.

All seems to be communicating just as well as the i2cv1 driver, the UV index in my office is 0.02

Sorry for the noise, but sometimes if you don't post a question you spin your wheels.

[Giovanni]

np glad it is solved.

Giovanni

[dflogeras]

So my next step was to port to my homebrew F030F4 board. Worked like a charm, just thought you might like to know in case no one has yet used I2C on this device.