IIS2DH Reset

[iaeinfach]

I have an issue where the IIS2DH accelerometer gets into a state where I get no output form the axis registers. How can the device be reset without removing power? The IIS3DSH used to have a register that could be toggled for a hard reset. The IIS2DH only appears to have a soft reset with:

CTRL_REG5 (24h) : BOOT bit.

The BOOT does not reset the device to a working state. Only a POR.

[Eleon BORLINI]

Hi, the BOOT bit on the IIS2DH reboot the original calibration memory content, it doesn't turn off the device, let me say. But which is the symptom of your issue? After certain time do you get FF or 00 data on the output? Are you using SPI or I2C for the device communication? Especially in case of I2C, at which Vdd / VddIO are you working?

regards

[iaeinfach]

Indeed I am using I2C and yes the symptom is 00 data. The voltage is nominally 3.6V for both Vdd and VddIO.

On a future revision of the board I plan to use SPI. I assume this is the preferred method.

[iaeinfach]

Does not exceed 3.6V I should say.

[Eleon BORLINI]

yes, SPI is -let's say- less dependent to the "environmental" condition than I2C, and doesn't need pull-up resistors. Vdd 3.6V condition is quite at limit... could it be possible for you to work at lower Vdd?

[iaeinfach]

Perhaps, but trying to avoid excess components. Its a very small area. I also noted in earlier development that voltages below 3.3 would cause issues with readings. The low threshold varied with different chips, but transient dips would definitely cause me issues. Since I went with a higher voltage those issues have gone away.

[Eleon BORLINI]

​I understand... which pull-up values are you using?

[iaeinfach]

6.8k.

Correction, I was mistaken before. Earlier version was 3.6V direct to the IIS2DH. We now are in fact using 3.3v supply and I added caps to deal with the transients.

[Eleon BORLINI]

On how many samples did you faced the issue? Is your Vdd stable?