IIS3DWB consumes an excess of 1mA when it is communicated through SPI, only falling back with power cycling.

[MAnil.1]

Hi, I have been investigating a current leak in my circuit.

I triangulated it to my IIS3DWB, when the accelerometer is not communicated through and all pins are pulled up by the MCU, I get a system current consumption of ~20uA (expected)

But if I ever communicate with IIS3DWB, even for checking the WHOAMI register, my current floor is set to 1mA. (starting accelerometer adds 1an extra 1 mA)

This persists even if I set the lanes that drive SPI to GPIO and pull them up. (or down)

Persists if I make a software reset. It only resolves with a power cycling of the board. It is as if in IIS3DWB something latches and sinks current.

I read the other post, cleaned the flux and tried SW RESET. I also have ties the VDDIO to VDD, since the comm voltage is the same with VDD.

What might be the culprit? VDD is 3V6 (But the current persists with 3V or 3V3)

What might be the culprit?

[Eleon BORLINI]

Hi @Community member Anil​ ,

please note first that the consumption when ODR is at 26kHz on 3 Axes is 1.1mA.

When you read the WHOAMI register, are you setting the IIS3DWB in power down mode?

The device is set in PD mode when CTRL1_XL (10h) XL_EN[2:0] = 000b.

On the other side, if you are continuously monitoring the current consumption, you should measure a consumption peak during the SPI/I2C communication, due to the commutation of the digital buffer on the dataout line.

If the current consumption is constant, we should probably focus on the hardware: can you please share the schematic of your board? Is it in line with the STEVAL-MKI208V1K one?

-Eleon

[MAnil.1 · Accepted solution]

Thanks Eleon,
I have realized that my microcontroller has an errata about this issue. Today I was able to confirm it.

[KMeld.1]

Dear @Community member Anil​ & @Eleon BORLINI​ 

I find myself experiencing the same problem with current consumption. I my case the excess current consumption is around 250-1500uA.

I'm using the STM32WB55CE MCU running a BLE application which controls the IIS3 though SPI1. While developing I'm using the STEVAL-MKI208V1K kit.

The application is using FreeRTOS tickless idle in STOP2-mode (I know from the MCU manual & ST presentation that the SPI is affected by this, but control haven't been an issue yet.

Relevant part of schematic:

(ACC_INT2 is not connected!).

I've made a few measurements using the X-Nucleo-LPM01A board:

Consumption of board without the IIS3:

Consumption of board + IIS3-kit - IIS3 in full power mode:

Consumption of board + IIS3-kit - IIS3 in low power mode:

Consumption of IIS3-kit alone - IIS3 in full power mode:

Consumption of IIS3-kit alone - IIS3 in low power mode:

 

The Ampere increase when turning on the IIS3 fits the datasheet very well, but I'm confused what is drawing ~440uA in the 3rd graph...

My best guess is something SPI related, but not sure what...

Any ideas or hints?

[MAnil.1]

In my case, shutting SPI peripheral down AND executing any code in the Errata worked. Shut SPI down?​

[KMeld.1]

Shutting down SPI (HLA_SPI_DeInit() ) tkaes it down by ~90uA, so that will definitely be part of my solution as well. I will keep digging and get back asap.