BQ24295: Really slow charging

Chris Schramm

Part Number: BQ24295

Dear All,

I've just received of our prototype PCB with a host MCU connected to BQ24295. I2C, Boost and SYS Power worked right from the start, awesome!

But I have a strange observation: it seems that charging is really really really slow, between 0 and 15mA @ VBat=3.8V (standard protected 3.7V LiPo with 2.000mAh)

Here's the schematic for reference:

TS1 is 10k

I have by far not yet completed implementing all I2C registers, but I've started with those I thought might be important for my current (haha) problem:

REG8:6 says we're connected to USB (that's alright, because the PMIC's D+/D- are connected to MCU's D+/D-)
REG8:4 says we're in "Fast Charge" state (hahahahahaha)

REG1 says OTG and Charging are enabled

REG2:2+ gives 1024 (just as a side note: datasheet says that after a reset it will be 512mA although the default is 1024)
REG0 says the input current limit is 500 mA

Any help is greatly appreciated

over 7 years ago

0 Chris Schramm over 7 years ago

Intellectual 590 points

Meanwhile I read the fault register which is empty. R106 is 330 ohm instead of 317.

0 Chris Schramm over 7 years ago in reply to Chris Schramm

Intellectual 590 points

and another update: as soon as I remove VBAT, VBUS drops to about 2.4V - what is this?!

0 Fernando Lopez over 7 years ago

TI__Expert 8230 points

Hello Chris,

Looking at your schematic, the bootstrap capacitor, C105, is connected incorrectly on the schematic. It should be tied to the SW and not the output of the converter (SYS). Without the bootstrap cap correctly connected, the internal high side FET won't be able to turn on and the converter won't be able to operate as expected, which seems to be what you are observing.

0 Chris Schramm over 7 years ago in reply to Fernando Lopez

Intellectual 590 points

uuuuh what a shame! Lucky me, I was able to re-solder C105 and to test the change - and yes, you're right, now everything is working perfectly. Thank you so much!
One last question: does CE need a pull up/down RESISTOR or is it fine enough to hook it to MCU'S GPIO? Obviously, it will only pulled when MCU is powered by BQ24295

0 Fernando Lopez over 7 years ago in reply to Chris Schramm

TI__Expert 8230 points

Hi Chris,

CE/ can be pulled down with a resistor if you want to enable charge by default. It's fine to let the MCU control it with a push-pull GPIO as well, but the risk here is if the MCU hangs and it isn't able to enable charge, you could run into a scenario where you can't charge a dead system.

Power management

Power management forum

BQ24295: Really slow charging