I reset the BQ24272 watchdog timer every 10 seconds via I2C. Some times it seems like the BQ24272 pull down the SDA before the ninth bit and so it crashes the communication with the maser.
Correct transaction
Problematic transaction
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
I reset the BQ24272 watchdog timer every 10 seconds via I2C. Some times it seems like the BQ24272 pull down the SDA before the ninth bit and so it crashes the communication with the maser.
Correct transaction
Problematic transaction
Hello,
I tested a BQ24272 EVM during charging and monitored the SCL and SDA lines with the scope. For my testing I shortened the watchdog reset to once a second to try to catch any issues that might be evident in the I2C communication.
Below is a sample of the waveforms that I observed consistently during charging. The top is a 50us and the bottom is 20us.
I did not observe any issues with the communication or any observed interruptions in the charging.
I do see that the problematic transaction you have captured bears a strong resemblance to the second half of the sampled communication. However there was no fault in the communication observed in my testing. However the baud rate of the EVM & USB interface is lower to 100kbaud.
Have you tested at a lower baud rate? Do you still see communications issues?
Thank you for your time Ryan!
I change the capacitor of BYPASS PIN (pin 23) to 1uF from 0.1uF that a had in the first place. The datasheet recommend 0.1uF minimum capacitance. The things are better now. I still testing it. I will try lower baud rate, too. I will keep you informed.
In fact, in my case, the larger capacitor improve the i2c communication. Now the issues appears more seldom. Do you think the reason may be poor grounding? In my prototype I have not solder the thermal pad to ground because I have soldered the IC by hand.
Another issue a discovered is that some times the charge is not terminated. The TE is enabled and the termination current is set to 400mA. The consumption of the rest components is 40mA.
In low charge mode is the "charge current termination" disabled?
Hello Anastasios,
The larger bypass capacitor should attenuate any higher frequency noise better which makes the improvement in the I2C understandable.
The disconnected thermal pad is something that needs to be fixed with the datasheet explicitly stating it should be connected directly to a ground potential. Are you able to solder the pad?
When the device does not terminate does the current simply continue to taper lower and lower?
Are you only seeing this failure to terminate when the LOW_CHG bit is enabled?
Hello,
I order a stencil and I will solder the BQ24272 properly. I believe the issue with the I2C is due to unsoldering thermal pad. Don't you believe? The lower baud rate didn't help the I2C.
When the device does not terminate, the current continues to lower near zero mA. When the safe timer expires the charging terminates and 'Battery OVP' fault return. This happen when LOW_CHG bit is enabled. Until now I haven't noticed it when the LOW_CHG bit is disabled.
Thank you
Hello Anastasios,
When the LOW_CHG bit is enabled the safety is slowed and the termination is disabled. This is why you are not seeing seeing termination in when that bit is set.