Let’s make a list of things that can be made better:
[x] Move C3 design here once it’s validated
[x] Add ESD protection diodes to USB
[ ] Research about how to add optocoupler to check that light is really consuming in main power PCB
[x] Reroute things more wisely in 1.1 S3 and further C3 revisions
[x] Make average consumption tests with and without WiFi in both PCB models (Rainmaker version)
[ ] Make sure deepsleep is a < 1.5 mA deepsleep in worst case scenario. This idea is meant to be plugged 24/7 to 220AC so we are not going to be demanding in this topic. But considering someone might use it all day connected to USB we should keep that in mind
S3 test scenario with DS3231 RTC is 50 mA avg with very short 90 mA peaks when SPI transmission takes place:
Further notes:
In master now, S3 pro version, the DS3231 backup is not anymore capacitive as @Uspizig mentioned it will only keep the time a few seconds so why care spending 2 u$ in capacitors?
This is updated by a standard and available at around 0.9$ CR1220 battery connector. In hardware branch there is still the old version in case there is the need to compare something.
Let’s make a list of things that can be made better:
S3 test scenario with DS3231 RTC is 50 mA avg with very short 90 mA peaks when SPI transmission takes place:
more tests added in the WiKi
Further notes: In master now, S3 pro version, the DS3231 backup is not anymore capacitive as @Uspizig mentioned it will only keep the time a few seconds so why care spending 2 u$ in capacitors? This is updated by a standard and available at around 0.9$ CR1220 battery connector. In hardware branch there is still the old version in case there is the need to compare something.