Power-on Reset problem with REG0

[thibv]

Hi,

I'm working on a custom board with a nRF52840.

I want to use REG0 and REG1 as mentionned in Figure 13 of the datasheet : with VDDH, and both DCDC. My PCB is powered by a LiPo battery, from 4.2 to 3.3V. I want VDD to be 3V.

After several weeks developing without this problem, I encountered the following one :

My board doesn't start, because (I think) VDD is not set properly. In the datasheet, I can read "The system is held in a reset state until the power supply has reached the minimum operating voltage and the internal voltage regulators have started." and I think it's the problem.

- VDD is at 560 mV, while VDDH is at 4.2V (DC power supply, currently).

- DECx voltages are not set properly neither

I'm using Circuit Configuraiton No 4. To start my PCB, I can temporarily set VDD to VDDH (manually, with an external wire) and the MCU keeps running after that.

My question is : is there a way to find why power supply doesn't reach the minimum operating voltage ? (If it's really the problem). 

Thanks for your help,

[Einar Thorsrud]

Hi,

Sadly, the DC/DC of REG0 is not functional, so your main options are either to use REG0 in LDO mode (and take the current consumption hit) or use an external DC/DC. has several errata. This is documented in erratum [197] POWER: DCDC of REG0 not functional. You should also be aware of erratum [202] POWER: Device does not start up in high voltage mode in case you want to use REG0 in LDO mode.

[thibv]

Hi,

Thank you for your fast response.

Erratums 197 and 202 are not present on revision QIAA-AA0, which is the version I use. But there is Erratum 140 which suggests to use DCDC, which is contrary to Erratum 197.

My VDD bus is connected to an IMU to ensure power supply. I cannot modify it in my PCB. Consumption hit is not a problem for this version. Is there a way to use VDDH (DCDC or LDO), taking account that :

- I cannot unconnect my IMU, who could consume more than 1 mA

- My revision is QIAA-AA0

Thank you for your help,

Thibv.

[Turbo J]

Did you de-solder the chip from a prerelease DK? These chips where not available otherwise to my knowledge...

Triple-check the revision number, please.

[Einar Thorsrud]

Hi Thibv

Erratum 197 and 202 have not been confirmed for The Engineering 0 revision (QIAA-AA0), but you should consider that it is no longer being tested as it has been obsoleted by three subsequent revisions. It will have issues that are not documented for that revision. The DC/DC is non-functional with REG0 on all engineering samples as well as the only production version, revision 1 (QIAA-Cx0).