High Voltage Mode has high current consumption on custom nRF5340 board

[Rory Morrison]

I've been trying to track down why we have higher than expected current consumption in our application. This is on custom hardware using an nRF5340 module, powered in High Voltage Mode (i.e. power supply connects to HVCC and not VCC) at 3 volts. Also worth noting we are powering some external chips off VCC, but they should be drawing only a few uA each.

Eventually I found that we have ~800 uA more current on our hardware than we do on a Dev Kit, even if we do very little in terms of code running on the device. The difference being, it would seem, that the Dev Kit is in Normal Voltage Mode, and our board is in High Voltage Mode. By connecting HVCC and VCC (as far as I can tell, this is all I need to do to put the chip into normal voltage mode) I get very much lower current consumption - our actual code goes from ~1300 uA to ~500 uA, and the test code I've been using goes from ~900 uA to ~100 uA. So it seems like the VREGH is somehow drawing 800 uA.

Is this expected, perhaps because we're powering at only 3 volts? Or have I missed something simple that we need to do in configuration to correctly enable High Voltage mode? 

I believe the module we're using has all the inductors and capacitors needed for High Voltage Mode, but perhaps we missed something and it doesn't - could that be the cause?

Or does the nRF5340 not support external components powered via VCC? I did see something earlier about an nRF52 chip that didn't, which made me wonder.

The obvious fix is to just use normal voltage mode instead, but I'd be happier with this if I could explain why there's such a big difference.

[Simonr]

Hi

This sounds like the custom board not being set up correctly to me. Have you tried setting the nRF53 DK to VDDH mode by I.E. powering it with USB by setting SW9 to USB (rightmost position) and powering the board using the J3 nRF USB connector.

You can also upload your schematics here so we can do a quick review to make sure that everything is set up correctly. If you don't want to share your schematics publicly, let me know and we can set this case to private mode so only yourself and Nordic engineers will be able to view your ticket. An increase of 800µA in what I assume is a very simple application is not expected when powering a board with 3V in VDDH mode.

Best regards,

Simon

[Rory Morrison]

Hi Simon,

We're using the ublox nora B106 module, would you be able to advise about the layout for that or would we need to contact ublox themselves?

Otherwise, can you confirm the nRF5340 does support using VCC as a power supply for external components?

And can you confirm the only difference between High Voltage mode and Normal Voltage mode in terms of layout should be that in normal voltage mode, the HVCC and VCC are both connected together and to the battery/power source? (I'm still assuming the VREGH filtering inductor and capacitor are present in the module, as the datasheet for that says no additional components are required).

Thanks,

Rory

[Simonr]

Hi Rory

The modules from third parties generally already have an approved schematic from the nRF point-of-view, so the schematic review might not be necessary here. 

Yes, VDD/VCC should be fine to use to power external components as well.

In normal voltage mode VDDH and VDD should be connected together indeed, the filtering inductor and VDDH capacitors can be removed (see image below) when only using normal voltage mode: 

Best regards,

Simon