Incorrect output from nPM1300 when powered through VBUS

[ryeanne]

Hi there,

I've designed a custom board implementing the nPM1300 and am currently in the process of examining the board without any firmware. Throughout testing, I have tried to power on the board using VBUS and VBAT.

The issue I am facing is:

- The output of VSYS and VOUT1/VOUT2 are correct when powering directly through VBAT using a power supply set at 3.7V. The current draw of the system sits at around 30 mA.

- The output of VSYS is incorrect (approx. 2V) when powered directly through VBUS using a power supply set at 5V. The current draw of the system is 0-0.5 mA.

- This causes the output of VOUT1/VOUT2 and an external buck regulator that I have connected to VSYS to be incorrect.

- The only additional connection to VBUS apart from the nPM1300 VBUS input is an ESD protection diode.

- I have also tried powering through VBUS using a USB cable connected to a power bank and laptop. When using the power bank, it delivers 5.15V into VBUS for around 20 seconds before dropping to around 2V.

Is this a hardware or firmware issue? I understand that there are some functions that can only be programmed through firmware, however I was under the assumption that the buck regulators and basic functionality (including being powered through VBUS) were possible without any firmware commands.

Thank you!

[Tharaka]

Hi

With USB (VBUS), nPM1300 limits current to 100 mA by default. Without proper CC1/CC2 connections or firmware to raise the limit, nPM1300 SYSREG block can shut down under load, potentially even short inrush current spikes at startup can trigger this causing VSYS to drop to ~2 V.

Regards 

Tharaka

[ryeanne]

Hi Tharaka,

I tried to limit my PSU to < 100 mA when powering directly through VBUS to see if transient current spikes were an issue but the issue remained. I was also under the impression that the nPM1300 could be powered directly through VBUS without CC negotiation, however it is connected to the USB-C port and seemed to work as intended with my laptop USB.

I did notice as I was probing voltages however that when I accidentally shorted VBUS and VSYS that the all of the outputs and bucks would turn on with the correct voltages, and current draw went up to 30mA as expected. I know that this is probably not recommended, but what could be the reason why everything works properly after that?

[Tharaka]

Hi

So based on your new information, 

VBUS is collapsing, not VSYS:

This indicates the power supply itself is shutting down/limiting, not the nPM1300's internal regulation.
VBUS is an input for nPM1300. The only way it can make it collapse is pulling too much current on it compared to what it can deliver. However, when VBUS is at 2.00 V, then nPM1300 will not pull any current on it,  so it it stays at 2.00V it is not because nPM1300 pulls current.

So the problem appears to be upstream of the nPM1300, not within it.

Could there be a possible short or excessive loading somewhere in the circuit draw curremnt from VBUS?
Or issue with PSU ? 

Also, 

Disabling DCDCs didn't help:

Rules out excessive DCDC inrush current as the root cause

Bypass test failed:

Pre-charging capacitors via VBUS-VSYS short didn't create lasting fix ,confirms it's not just a startup transient issue

This extra 4.7 uF capacitance at VSYS looks reasonable and shouldn't cause the issue. 

Regards 

Tharaka

[ryeanne]

Hi Tharaka, thank you for the ideas.

- When testing VBUS using a power supply connected directly to the rail, there do not seem to be any signs of high current draw causing the power supply to shut down or limit. The PSU current remains near-zero but sometimes pulls a couple mA.

- The only thing connected to VBUS is an ESD protection diode placed near the USB port.

- I will double check tomorrow whether I successfully turned off BUCK1 and 2, so I will give an update on that.

- If the power source seems to be the issue, why would it function normally once bypassing the regulator temporarily with the VBUS/VSYS short?

- Would you have any other suggestions for testing that I could do for a better idea of the root cause?

Thank you 

[ryeanne]

- After investigating today, you were correct in assuming that VBUS collapsed due to the power supply. The power bank would anomalously 'turn off' and run at 2V after a period of time. I decided to test again using an oscilloscope and power through a USB cable connected to a PC. When powering through a PC, VBUS no longer collapses. However, VSYS sits at 1.8V with small periodic voltage spikes. On the other hand, when powered through VBAT, VSYS sits at VBAT stably.

- I am unable to turn off BUCK1 and 2 when powered through VBUS and shorting VSET1 and VSET2. However, when powered through VBAT, I am able to turn off the bucks as intended.

[Tomi Koskela]

Have you tried disconnecting the buck converter TPSM82821S you have in the schematic? Maybe it has too high inrush current at startup that will exceed the current limit. 

[ryeanne]

I haven't tried yet because both the VSYS input and EN pin for the TPSM are hard to access, and I want to keep desoldering the entire module to the last resort. Could it be though that the inrush current for the buck is higher than a SYSREG output limit but not a Charger limit?

[Tomi Koskela]

Yes, default discharge limit for the battery is 1460mA and default VBUS current limit is 100mA

[ryeanne]

Thanks for the clarification. If I were to write to change the default VBUS current limit through devicetree, would it then be able to work through VBUS as intended?