nPM1300 resets with VSYSLOW when enabling external boost converters at VBAT ≈ 3.1–3.2 V

[Oleksii Shcherbyna]

Hi! We have a device that includes boost converters powered from the battery through the nPM1300.

When these converters are enabled, a short but relatively high current surge occurs, causing a brief drop in both VBAT and VSYS voltages. As a result, the internal POF comparator is triggered, causing the nPM1300 to reset.

We read the RSTCAUSE register and determined that the reset was caused by a VSYSLOW event.

Our setup:

• Battery: BZ503440, 850 mAh (we can attach our battery model for fuel gauging if needed).
• Discharge current limit: 1 A (configured in DTS file).
• VSYS capacitance: 42 uF. VOUT2 capacitance 42 uF.
• Load: two HV9150K6-G boost DC-DC converters on VSYS.

To address this issue, we implemented several changes:

• Added a delay before turning on the external switch that enables the boost converters (soft start).
• Added additional power supply capacitors.
• Reduced the POF comparator threshold by changing the POFVSYSTHRESHSEL field in the POFCONFIG register to 2.6 V. This change had the most noticeable effect.

However, we are still experiencing resets when the input voltage is around 3.1 .. 3.2 V.

We have the following questions:

• Is there any way to disable the POF comparator or prevent it from triggering in response to short VSYS voltage drops?
• Is it possible to configure the nPM1300 to ignore short-duration voltage dips or transient voltage drops on VSYS?
• Is it possible to set the POFVSYSTHRESHSEL threshold below 2.6 V? Or maybe some other ways to solve that.

Thanks,

[Amanda Hsieh]

Hi, 

No. In the nPM1300, the VSYS comparator (VSYSPOF) is an always-active hardware protection feature. It cannot be disabled nor controlled to a lower value than its lowest 2.6V value because of the PMIC architecture. 

Just thinking about your application loads and LiPo battery limits, for a standard LiPo cell an unloaded/lightly-loaded voltage of 3.1–3.2 V is already past the discharge knee and corresponds to roughly 0 to 1% remaining State of Charge (SOC). At this stage, there is almost no usable energy left, and the cell’s internal impedance increases sharply, meaning any high-current surge from enabling boost converters will anyway naturally cause a steep voltage collapse.
Even if you manage to extract the last bit of juice, it puts a lot more stress on the battery (aging) when you try to operate around that region with those high currents. 

Regards,
Amanda H.

[Oleksii Shcherbyna]

Hi Amanda,

Thank you. As VSYSPOF can't be disabled, so we'll handle this in firmware with fuel gauge to limit load at low state of charge: we won't enable boost converters or reduce load, once SoC or VBAT drops below some threshold.

Thanks again for your help!

Best regards,
Oleksii