We are using Apollo3blue for wearable production and a few of the boards are found brown-out issue. The reset register value is 0x484 which indicates there was a BLE buck brown out. While measuring the VDDBH voltage, some of the boards have 100mV~200mV voltage dip.
Could you explain Apollo3blue BLE buck brown-out mechanism?
BLEBUCK brownout is flagged when the blebuck_comp1 comparator (compares VDDBH to a reference voltage) asserts continually for a programmable period of time. We have this trim set at 0xF which equates to comp1 asserting for 81us. BLEBUCK_COMP1 compares attenuated VDDBH to a 580mV reference.· The brownout circuit does not look at voltage, it senses how long the comparator continually determines VDDBH is lower than its trimmed value.· 100~200mV may or may not cause a brownout. Brownout depends on time duration of dip not magnitude. 100us with a 10mV dip may brownout, but 10us with 100mV should not.
Is the voltage dip on VDDBH normal? Why does the dip happen?
A small dip may be seen on the BLEBUCK when a change in load is seen. This is because of the time constant of ton adjust loop means it may take a short period of time for BLEBUCK to respond. On most chips a dip may not be seen, on some it may due to a combination of zero-crossing trim offset, inductor quality and load dynamics. A 100mV dip will not cause any issues to the circuits running on the VDDBH domain. There is headroom there and it's ok to use.
What is the recommendation of Apollo3 BLE buck inductor?
For operation across full voltage range (1.755V - 3.63V), recommend to use high saturation current (>800mA) inductor. E.g. 1uH with less than 20% loss at 1A current preferred. Inductor's max DC resistance should be less than 0.55 ohm and inductor's operating frequency should be more than 20MHz.