MSP432P4111: I2C slower than expected

Matthias Wissen

Part Number: MSP432P4111
Other Parts Discussed in Thread: TCA8418

Hello,
I have encountered some issues with using I2C in out application.
I would like to connect a TCA8418 to a MSP432P4111 via I2C in fast mode (400 kHZ). According to the datasheet, the TCA8418 requires 0.6 us for I2C clock high time (t_sch) and 1.3 us for I2C clock low time (t_scl).
I measured the timing with an oscilloscope. In our application t_sch is 1.15 us and t_scl is 1.37 us, so our t_scl is very close the required time of 1.3 us and t_sch is way above. Is that normal?

Also I noticed that the measured bus clock rate is not 400 kHz but 342 kHz. Our system clock is 48 MHz and the prescaler value in UCB3BRW register is 120. So why don't we get the full 400 kHz? If I want to get 400 kHz I have to change the software code to 480 kHz.

I changed the SDA and SCL pull up resistors from 4k7 to 1k2. Then the bus clock rate increased from 342 kHz to 384 kHz. So it seems, that the clock rate depends on the pull up resistor (and the capacitance?), but why? Is there a proper way to get 400 kHz?

Thanks for your help in advance!

Matthias

over 4 years ago

0 Chester Gillon over 4 years ago

Guru 92251 points

Matthias Wissen said:
So it seems, that the clock rate depends on the pull up resistor (and the capacitance?), but why?

Does TM4C1294KCPDT: I2C SCL frequency help?

It's a different device, but think the same principle could apply.

0 Matthias Wissen over 4 years ago in reply to Chester Gillon

Prodigy 10 points

Hi,
thank you for the quick response. I already thought about clock stretching, but wasn't sure.
In the request from your link t_scl and t_sch are within the limits. When I replaced the pull ups by stronger pull ups, I realized that t_scl falls below 1.3 us. I've measured 1.23 us. Why is this possible? Is there something I can do to extend the low time a little bit?

0 Chester Gillon over 4 years ago in reply to Matthias Wissen

Guru 92251 points

in your initial post you mentioned changing the software code to 480 kHz to get 400 kHz on the I2C bus.

With the change to stronger pull-ups what frequency was the software trying to use?

0 Matthias Wissen over 4 years ago in reply to Chester Gillon

Prodigy 10 points

Code is still on 400 kHz. 480 kHz was just for a quick test, but removed this value from my code.

0 Chester Gillon over 4 years ago in reply to Matthias Wissen

Guru 92251 points

Matthias Wissen said:
When I replaced the pull ups by stronger pull ups, I realized that t_scl falls below 1.3 us. I've measured 1.23 us.

Looking in MSP432P411x, MSP432P401x SimpleLink Mixed-Signal Microcontrollers datasheet (Rev. B) t_LOW appears on the figure but there is no value in the table:

I.e. not sure what is the minimum t_LOW the MSP432P4111 is supposed to generate as a I2C master at 400 kHz.

Perhaps someone from TI can clarify.

0 Matthias Wissen over 4 years ago in reply to Chester Gillon

Prodigy 10 points

I found the timing values in the TCA8418 datasheet:

Arm-based microcontrollers

Arm-based microcontrollers forum

MSP432P4111: I2C slower than expected