H3LIS - low power optimize impact or impulse detect

[JKeat]

I want to optimize the low power mode of my H3LIS331DL to detect an impact. I've attached an image of a scope capture of the current draw of my board during sleep mode. I've determined that the large current draw pules are the accelerometer turning on. The low power frequency in the capture (10Hz) gives me the most responsive detection. However, that pulse is huge! 40ms wide and 300uA is way too much for this application. What's the best way to optimize the power in this case? What strategies can I use? Is 300uA really the baseline current draw when the accel turns on? Or am I missing a setup parameter that can lower that baseline draw?

[Miroslav BATEK]

The current consumption of this sensor in low-power mode is typically 10uA.

Can you share you schematic and sensor configuration?

[JKeat]

I'm working on getting the schematic from the contractor. In the meantime, what is the expected current consumption during the 'Turn on time'? What is the expected current consumption in sleep mode?

[Miroslav BATEK]

I don't have the value during turn on, but as the sensors starts to power down move, the current consumption wont be significant. As the power consumption in power down mode is 1uA, it doesn't make sense to turn off the sensor and the again turn it on and configure again.

Please be aware also "When no communication is ongoing, data on CS, SPC, SDI and SDO are driven by internal pull-up resistors."

[JKeat]

When the device is in low power mode, it looks like it has two states. In state 1, it is drawing <10uA of current. However, it periodically wakes up to check for any interrupt events. At that point, state 2, it 'turns on' and draws more current. Then it returns to state 1. It transitions into state 2 depending on the ODR set for Low Power Mode. For Low power mode at 10Hz, the device wakes up once every 100ms. Am I correct that this is the behavior of low power mode? If so, what are the expected current consumption values for state 1 vs state 2?

Also, is there any way to disconnect the CS, SPC, SDI, and SDO pull-up resistors?

[JKeat]

Also, I should mention that the part number is actually H3LIS100DL.

[Miroslav BATEK]

Yes, you understanding is correct.

I did a measurement and my device draws ~270uA in active state (state 2) and ~1uA (state 1), this is in line with expectation. The average value for low power mode 1Hz ODR is ~10uA.

There is no way how to disconnect the internal pull-ups.