Other Parts Discussed in Thread: BQ24296,
Hi
The customer feedback that :
unsigned int ret = 0;
ret = bq24296_config interface((unsigned char) (bq24296 CON1)
(unsigned char)
(val)
(unsigned char)
(CON1 SYS MIN MASK)
(unsigned char)
(CON1 SYS MIN
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.
Hi
The customer feedback that :
Hi Star,
Does the customer disable charge when battery terminal is not connected? VSYS is expected to be maintained even with no battery is connected if charge is disabled.
I am not aware of any change to the logic of the IC. I am not sure why the customer is observing different behavior.
When the battery terminal is not connected to the battery, after VSYS outputs 3.6V, both VSYS and the battery terminal will stop outputting.
It would be helpful to see this behavior in a waveform capture. Can you help to request this from the customer. With no battery connected and charge enabled VSYS oscillating is expected.
Best Regards,
Garrett
Hi Garrett
Thanks for your reply.
When the battery is not connected, the CE pin is pulled low, when start charge the VSYS waveform as below:
When the battery is not connected, the CE pin is pulled up, when start charge the VSYS waveform as below:
Please help check it.
Thanks
Star
Hi Star,
Thank you for providing the waveform capture. Please see my comments below.
I have confirmed the below waveform is the expected behavior of the BQ24297 when charge is enabled, but battery is not connected. Both SYS and BAT voltage should oscillate.
Since the customer waveform does not match this expectation I have a few additional questions.
1)Can you help to provide a more detailed explanation of the customer test procedure? What is triggering VSYS to initially rise in the waveform above? At what point in the waveform is the customer pulling /CE pin low? What is the expected current draw of the load attached to SYS?
2)Can the customer provide a waveform showing VBUS, VSYS, and VBAT instead of just VSYS?
3)Does the device behave as expected when a battery is connected?
Best Regards,
Garrett
Hi Garrett
Thanks for your reply.
1) When the motherboard is not connected to the battery, connect the charger: BQ24297 inputs 5V, then the CE PIN is pulled up to the RENG voltage.
VSYS will be powered on for 2-3 seconds and then drop, and VBAT will not be powered on. The expected current for the load attached to VSYS is 2A.
They replaced different batches of BQ24297, and the waveforms obtained were inconsistent. The waveform measured now is the most recently produced IC, and the VSYS of the older IC will continue to be output.
2) The following are the power-on waveforms of VBUS, VSYS, VBAT, and CE
3) When the battery is connected, the device can work normally and can be charged. When the battery is completely out of power, the battery cannot be charged, which is equivalent to the situation where the battery is not connected. The battery cannot be activated even when connected to a charger.
Waiting for your reply.
Thanks
Star
Hi Star,
Thank you for the explanation and additional waveforms. Given this new information I need to do some research and testing and will get back to you.
Best Regards,
Garrett
Hi Star,
Firstly regarding 3) can you help to get more information regarding the protection circuit used by the customer when their battery is completely out of power? How does the circuit disconnect the battery terminal from the BAT pin? What is the mechanism to reconnect the battery to BAT pin (for example does the protector require a certain voltage level)?
The expected current for the load attached to VSYS is 2A.
Please have customer double check expected current draw. Is current draw expected to be minimal at initial powerup? If current draw was 2A right at power on VSYS would never rise above 2V due to the soft-start feature described on page 21 of the datasheet. When I test on EVM even a light load such as 200mA prevents SYS from ramping up when no battery is attached.
Then focusing on the most recent waveform where /CE is always low there appears to be a key event which pulls both BAT and SYS low. I have a few questions for customer regarding this event.
-Is there a spike in current draw either at SYS or even possibly BAT?
-Can you capture SW node voltage along with VSYS with /CE pulled low? I would like to check if buck converter is potentially stopping.
-What is status of all registers following the drop in VSYS and VBAT?
-Following the drop do you ever observe VSYS or VBAT being powered again?
Best Regards,
Garrett
Hi Garrett
Thanks for your reply.
can you help to get more information regarding the protection circuit when the battery is completely out of power? How does the circuit disconnect the battery terminal from the BAT pin? What is the mechanism to reconnect the battery to BAT pin (for example does the protector require a certain voltage level)?
A: After the battery discharge voltage is lower than 2.8V, the discharge will be stopped; after battery protection, the battery needs to be charged to recover. After the battery is activated, it will continue to output voltage to the outside.
Please have double check expected current draw. Is current draw expected to be minimal at initial powerup? If current draw was 2A right at power on VSYS would never rise above 2V due to the soft-start feature described on page 21 of the datasheet. When I test on EVM even a light load such as 200mA prevents SYS from ramping up when no battery is attached.
A: VSYS will supply power to the main control. During the boot process, the current is dynamic, ranging from tens of mA to more than 1A.
Is there a spike in current draw either at SYS or even possibly BAT?
A:It is to measure the status of VSYS and BAT when the battery is out of power in the shutdown state. When charging is performed in the shutdown state, there will normally be no excessive current peak.
Can you capture SW node voltage along with VSYS with /CE pulled low? I would like to check if buck converter is potentially stopping.
A: The following are the waveforms of VSYS and SW when CE is pulled low with battery .
The following are the waveforms of VSYS and SW when CE is pulled low and without battery
Following the drop do you ever observe VSYS or VBAT being powered again?
A: After powering down, VSYS and VBAT were not observed to be powered back on.
Register value in charging state as below:
Register value after booting:
Please help check it.
Thanks
Star
Hi Star,
We are reviewing the new information and will get back to you later this week.
Best Regards,
Garrett
Hi Star,
Thank you for the additional details, please see my comments below.
after battery protection, the battery needs to be charged to recover. After the battery is activated, it will continue to output voltage to the outside.
What is the trigger to activate the battery? Is it voltage based or current based?
For the waveforms showing SW node without battery can you please provide a waveform with the original timescale of 400 msec along with a zoom in to a smaller time scale at the moment when VSYS and VBAT drop. In the waveforms provided VSYS is normal. I need to review SW node behavior for the event when VSYS drops.
I have also reviewed the register statuses and the one that stands out to me is REG08 indicates PG_STAT = 0 (not power good) after booting. Earlier in the thread you mentioned CE is pulled up to REGN and one of the waveforms shows CE dropping to GND at same time VSYS drops. This makes me suspect the issue may be relating to the customers input supply not being stable or for some reason REGN LDO is turning off.
Best Regards,
Garrett
Hi Star,
I wanted to share a theory which may explain the behavior.
Comparing the register status during charge verse after booting I noticed after booting that the input current limit was reduced from 2A setting to 500mA setting. You mentioned the customers load is expected to be ~2A. The SYS load increasing after initial boot up above the 500mA IINLIM threshold may be the cause of VSYS crashing. Then the soft start feature does not allow SYS to ramp back up since no battery is present and load is > 100mA.
To debug can you request the customer to force D+/D- detection to set IINLIM to to a value of 2A or higher and see if the VSYS crash behavior is still observed.
Best Regards,
Garrett
RILIM's current resistor is 0R, the hardware does not limit the current, which is controlled by the software. The current mentioned in the reply is reduced by 500mA from 2A, which is a value set by the software according to the type of charger when the host is running normally.Now without the battery, connect the charger, VSYS will drop, then the software is not running, then the current is not limited.
Hello,
The register dump provided by Star earlier in the thread indicates that after booting up the current limit is initially set to 500mA.
My best theory for the behavior is that VSYS crashes due to this lower current limit before the software can update the current limit to 2A. When battery is connected there is no issue with SYS startup and software can increase the current limit, but when battery is not connected or completely discharged VSYS is not able to ramp up properly and the input current limit is never able to be set to the higher value.
Best Regards,
Garrett