My interrupt driven low power design uses the MMA8451Q orientation interrupt to wake the design from sleep. The configuration and interrupt functionality work perfectly, with the MMA configured and verified to be working at the lowest sample rate in the lowest power mode. The device is connected per application diagram Figure 4 in the product datasheet. In my application, VDDIO and VDD are tied (and unfortunately not separable in this layout for independent current measurements).
I'm measuring sleep current using an Agilent 34405A.
No MMA in circuit (just sleeping MCU) <3uA
MMA installed, sleeping, 1.56Hz, LP Mode 30uA
I'm expecting closer to 10uA total system current consumption in this configuration. I've replaced the device once, and confirmed the same behavior with the second device. Any ideas on how to track down this extra current draw?
Another finding is that when I disconnect power from the MMA leaving the I2C lines connected, the current drops to expected levels. Noteworthy is that the MMA configures and functions as normal with no VDD/VDDIO connections (parasitically powered through I2C lines). It generates the orientation interrupts fine, and wakes the MCU for processing. Measuring VDD/VDDIO at the MMA with a scope confirms an expected unstable voltage signal due to charging of VDD/VDDIO bypass capacitors through the I2C pull-ups, then discharging when the MMA wakes and I2C traffic resumes. In this invalid and certainly unstable configuration the design draws 2.8uA while sleeping.