NRF52 DK: ~60uA extra current draw when changing P0.27 to input!?

RE: NRF52 DK: ~60uA extra current draw when changing P0.27 to input!?

martijnthe — Fri, 18 Jan 2019 20:02:23 GMT

Thanks. I ended up checking on a bare board with the same chip and didn't see any issues with P0.26/P0.27.

RE: NRF52 DK: ~60uA extra current draw when changing P0.27 to input!?

hmolesworth — Fri, 18 Jan 2019 16:21:28 GMT

Hmm, edited as I see the upper schottky clamp only apples to channel 1, not the 3 data channels. That only affects P0.17.

"should disconnect P0.26/P0.27 from the I2C pull-ups (R29 and R30)" .. ah, if only life were that simple. R29 and R30 can be clamped by internal schottky protection diodes within U7 as power switch voltage on the switch IC can rise above (VEXT+500mV) via (say) R29+R32 in series; if VEXT is Gnd that limits those voltages to 500mV-ish and allows current to flow between all those pins. To verify whether this is the root cause requires removing (say) R29 and checking the current draw again.

Also note P0.17 can possibly power up VEXT if driven high by the 0.5V schottky internal clamp on the NX3DV2567 switch itself. Perhaps start by ensuring P0.17 is floating or actively driven low (see edit note above).

RE: NRF52 DK: ~60uA extra current draw when changing P0.27 to input!?

martijnthe — Fri, 18 Jan 2019 08:42:53 GMT

Thanks for the hints.

It looks like P0.26 and P0.27 are connected to U6 (a NX3DV2567 switch IC).

But when SHIELD_DETECT is not connected to anything, it's being pulled up by R39, which, in theory should disconnect P0.26/P0.27 from the I2C pull-ups (R29 and R30). I measured P0.26/P0.27 and I don't see them being pulled up. I also tried toggling the switch by connecting SHIELD_DETECT to GND and only then they are indeed pulled up.

At any rate, I wonder if the switch is somehow causing the leakage.

I will try measuring on a board that doesn't have the switch connected.

RE: NRF52 DK: ~60uA extra current draw when changing P0.27 to input!?

hmolesworth — Thu, 17 Jan 2019 22:24:19 GMT

This is perhaps covered in the nRF52832 Development Kit v1.1.x User Guide v1.2 - see Table 5: pins P0.26 and P0.27 are used as i2c pins and have 4k7 pull-up resistors. This will introduce leakage into the input buffer when connected.

RE: NRF52 DK: ~60uA extra current draw when changing P0.27 to input!?

martijnthe — Thu, 17 Jan 2019 20:54:26 GMT

P0.26 also seems to exhibit a similar problem.

I'm seeing an even higher current draw, more like ~70uA extra.

I also tried P0.24 and P0.25. They don't have the problem.

What's so special about P0.26 and P0.27 that causes the current draw to become so high when used as input?