nPM1300 - No stable VSYS when supplied only with mirco USB

[Uzay Yavuz]

We observed a problem with our three different prototypes:

When there is only 5V-VBUS connected, the npm1300 does not ramp uo its voltages properly. The voltage rises up for a short period of time periodically, after which it drops to 1V-2V.

A similar problem was reported here:  Incorrect output from nPM1300 when powered through VBUS 

We believe that this problem may be caused by a too high current at start-up, and the inrush current seems to be limited to 100mA in our case.

I have measured the voltages before and after a 10R-Shunt, the voltage dropoff over the shunt is ~1V / 100mA.

VSYS correctly ramps up when:

   1) Connecting VBUS to VBAT

   2) Briefly shorting VBUS and VSYS 

   3) Connecting a Battery (may be removed after)

I measured the current draw when VBUS was shorted to VBAT at 350mA-peak during start-up. Quiescent current after start-up is 10mA. 

We believe our problem is caused by micro USB (no CC lines) somehow not allowing a higher current draw.

We tried the following things, but they did not help:

   1) Removing mostly everything from VSYS apart from two 10µ caps

   2) Setting VBUSINILIMSTARTUP and VBUSINILIM0 higher

   3) Connecting a CC-pullup resistor to the pcb, we even checked USBCDETECTSTATUS to see if a higher current was set.

Schematic for reference:

[Tomi Koskela]

Hello,

Default VBUSINLIM is 100mA. You can set this higher with VBUSINLIM0, but note you can write to those registers only after startup and the registers are volatile memory. And the current limit will be reset when you unplug the USB. So the initial startup should be limited to 100mA max current from the system point of view if VBAT is not present.

Is there anything else loading VSYS on your board?

In the schematic buck and LDO outputs are shorted together with 0ohm resistors. Are those resistors assembled?

VBAT seems to be missing a bypass capacitor. Should not affect this issue, but it is recommended to have 2.2uF capacitor on VBAT. 

Can you share your layout to see is there something missing there?

Thanks.

[Uzay Yavuz]

Hi,

VSYS is not significantly loaded elsewhere.

The Bucks are not shorted the the LDOs.

For the layout you can see the attached pictures.

I did not put the 1µF on VBUS on the pcb when I designed it, however, I soldered a cap to the pcbs on T6 to GND.

I managed to "fix" all three pcbs for now, but the fixes seem quite random. The bug can be triggered on one pcb by removing the 1µF vbus cap. It works with the cap attached. The other pcb can be fixed by removing the 0R on the Vsys pin and replacing it with 10R, this pcb did not need a 1µF cap.

I will try to find out more in a few days when I have time for it.

Also I have a technical question. Would the 100mA limit until configuration also apply for USB-C with CC lines connected to the PMIC or would it be able to draw 500mA instantly?

[Tomi Koskela]

Hello,

Yes the 100mA applies until you change the limit with TWI write. You can check the CC detection/max current level with software, but the actual current limit change must be done manually. It does not change it automatically.

Please also check our nPM1300 and nPM1304 Hardware Design Guidelines and our nPM1300 reference design for guidance on the PCB layout requirements.