Help with strange npm1100 performance

[thejeffcooper]

Hi,

I'm working on a board that uses an npm1100 (QFN) in the following configuration:

When testing my board, I populated it with only the components shown in the above schematic and supplied 5V to VBUS (no battery attached). For some reason, VSYS reads only 0.5V and VCC (the output of BUCK) reads 4.9 - almost the same as the input voltage. My power supply also indicates that the npm1100 is drawing ~5mA in this configuration (with no load), far above its rating.

Is something wrong with my schematic? I can't find any behavior in the datasheet that would explain any of this behavior. I suspected it might be a short, but manual probing of the board shows that everything is connected as the schematic indicates, and no part of the board is getting hot. I did have a mistake in the original board layout, but I've confirmed that my cut traces and bodge wires match the above schematic.

(As an aside - does the QFN24 center pad need to be connected to ground, or can it be left floating? The datasheet doesn't specify, but some of the example schematics show a "pin 25" listed as a second AVSS and connected to ground)

[ketiljo]

You seesm to be missing the caps on VBAT, VBUS and VSYS at least, see https://infocenter.nordicsemi.com/topic/ps_npm1100/chapters/hw_layout/ref_circuitry/doc/frontpage.html#config_3

The center pad must be connected to ground. 

The schematic looks correct as far as I can see. 

[thejeffcooper]

I do have caps on the VSYS line that are located fairly close to the npm1100; populating them doesn't seem to change anything. I'll include additional caps on VBAT and VBUS on the next iteration of the board, as well as grounding the center pad. I'll update this question when I get the next iteration of the board and have a better idea of what's going on.

It would be great if the product specification could be updated to include an explicit acknowledgement of the center pad grounding requirement (in addition to correcting the other errors in the pin list table, specifically the one about pin 4 being AVSS rather than NC)

Thanks!

[thejeffcooper]

I just wanted to close the loop that grounding the center pad and adding capacitance did indeed fix the problems I was having. Thanks!