I'm using an ADXL362 on a custom PCB along with an nRF52840 and a bit of extra periphery. The schematic & layout can be seen here: https://easyeda.com/mhx77/nrf52-lightmeter
I've been trying to optimize the circuit for low power consumption and was quite surprised to see the actual consumption being off by about 100µA from what I thought it would be. After disabling most other components, only the ADXL362 remained as the consumer.
Interestingly, I noticed that touching the SPI debug pads I had one the prototype board caused the current consumption to decrease. So I went on to solder 2M pulldown resistors to MOSI, MISO and SCLK. This actually reduced the current consumption of the whole board to 7µA, of which I reckon around 2µA are consumed by the ADXL362 (the remaining periphery uses at least 4.6µA).
Going further, I was able to remove the pulldowns from the MOSI and SCLK lines, and increase the value of the MISO pulldown resistor to 11M. If I remove the resistor, the current instantly jumps from 7µA to more than 110µA.
Connecting anything (scope probe, multimeter) to the MISO pin has the same effect, i.e. it will cause the overall current consumption to drop down to 7µA.
I've looked through the ADXL362 threads here and while I've seen high power consumption be an issue, I've not seen it brought um in relation to the MISO line.
While the pulldown certainly fixes the problem I'm seeing, I don't quite understand what the root cause could be. Is there anything that I've missed?
Sorry for having overlooked this post. We'll get back to you as soon as we have an update. Best regards,John
My colleague says that there's not much issue on the schematic although this still seems as a system level issue. Have you tried solder reflow on the PCB ? Also, have you tried to put on a different part so that we could cross-out if it is a part related issue.
Also, you shouldn't rely on adding pull down on the SPI lines to bring the power down as this would actually add extra power consumption.
Hope this helps. Best regards,John