BQ25622: BQ25622 Termination Current Issue in High-Load Condition

Hello, 

Thank you for reaching out via E2E. Please see my comments below. 

I have attempted to replicate your test conditions, and when I test with a BQ25622 IC with VREG set to 4.1V, 800mA SYS load, ICHG = 1.2A, and ITERM = 500mA I observe termination occur near 500mA as expected. The exact termination current on the specific unit I tested was 494mA and a waveform capture of my result is below. 

With my test results in mind I have a couple questions. 

1)Can you help to provide a waveform showing SYS load when charge current is approaching 500mA? 

2)Can you please share a register dump from the BQ25622 during condition 0 when charge current is below 500mA, but termination has not yet occurred so I can review all your settings as well as charger IC status outputs at the time. 

Best Regards,

Garrett 

Hi, Garrett 

This time, when using the logger to record, I increased the SYS voltage (if SYS current recording is needed, it may require modifying the circuit and re-measuring, which can be provided). Additionally, I recorded related information from the charger ADC. I also dumped the register when the charge current was 988mA and 424mA (below the set value of 500mA), confirming that the cutoff current 0x12 = 0x190 remained in the same state. For other register values, please refer to the charger log (you can search for "0x02 Data" in the document to find them). The test results with the increased SYS voltage can be found in the Charger Data.
Please let me know if you need further clarification or additional details. Thank you!

Charger log

Charger Data

Hello, 

Thank you for your response. We are reviewing the provided attachments and will get back to you with comment. 

Best Regards,

Garrett

Hi KaiPo, 

I have reviewed the provided data and register log and have the following comments. 

The register log when IBAT = 424mA shows no abnormal conditions besides charge current below termination threshold. Firstly, have you tested this on multiple units and is the behavior consistent across multiple units? 

Secondly, I understand it may require board modification, but ideally we do need to review a waveform capture showing either IBAT or ISYS over the few hundred msec before termination. Considering termination occurs at expected falling threshold (near 500mA) when SYS load is low my lead theory is SYS load has transient behavior which is impacting charge current. A waveform capture will a small enough time division will help to confirm this.

Charge current needs to remain stably below termination threshold for approx. 50 msec for termination to occur. 

Best Regards,

Garrett 

Hi Garrett,

Thank you for your response.

1. We have observed the charge current dropping below the termination threshold on other units as well.

2. Currently, the logging interval of our recorder is approximately 0.5 seconds, which is the limit of the current setup.
   I have conducted a test using this setting, and the results can be found in the link below.
   I will try alternative methods in the coming days to shorten the logging interval.

   Link

To summarize, maybe the charge is ending later than we expected because the current becomes unstable as it gets close to the termination threshold?

Best Regards,

KaiPo

Hi KaiPo, 

Thank you for your answers. 

0.5 second intervals will not be small enough to see system load transients impact on charge current, but understand there are limitations on your current hardware setup. Allow me to further explain what is possibly occurring in your testing when main screen is on. When there are fast load transients, specifically short drops in SYS current draw the current intended for SYS regulation will briefly flow to BAT output instead. These brief increases in IBAT when charge current is below termination threshold cause IBAT to not remain below the termination threshold for required 50 msec. 

You end up observing termination at a lower IBAT measurement in your logging because at lower IBAT even the brief increases due to load transients remains below the termination threshold. 

KaiPo Chang said:

To summarize, maybe the charge is ending later than we expected because the current becomes unstable as it gets close to the termination threshold?

Correct, termination is most likely ending later because charge current does remain below threshold due to instability caused by sys load transients. As charge current further reduces even the peaks of the instability remain below the ITERM threshold and charge termination can occur. 

If this longer charging time with your main screen on is a concern a couple workaround ideas I can think of are to either increase ITERM threshold above 500mA, only when you are in this operation state or utilize built in ADC to stop charge independent to automatic termination by disabling charge when ADC outputs IBAT < 500mA Iterm. 

Best Regards,

Garrett 

Hi Garrett,

I have shortened the current sampling time to around 5ms on average.
The results are shown in the figure below.

For complete data, please refer to the provided link.

Additionally, I would like to ask whether the interaction between ISYS and IBAT is a characteristic of this IC.
Is it possible to improve this issue through circuit design?

Best Regards,

KaiPo

Hi Garrett,

Thank you for your response.

I’d like to ask another question.
Regarding the charge termination condition, besides the mentioned "remain below threshold for full 50msec deglitch time," is there any corresponding condition for oscillations in IBAT?
If it’s simply "remain below threshold for full 50msec deglitch time," then it shouldn’t wait until 350mA to terminate.
It should terminate after the current remains below 500mA for a continuous 50ms.

Best Regards,

KaiPo