A custom PCB with an esp32 I am working with seemed to crash whenever WiFi.begin() was called. The Vcc has been measured and it seems to be as stable as a ESP32 DevkitC V4 running the same sample code. When using the Espressif IDF WiFi sample, when Tx power is set to 15-20 in the PHY tab of the configurations, the device will crash when attempting to connect to a WiFi AP. When the Tx power is set to 10-14 the device will not crash but will fail to connect to the WiFi AP in the specified number of tries.
We have completed multiple tests of the 3.3V power supply using an oscilloscope, when the trigger is set to 2.8v at a faster sampling rate, no voltage spikes are observed. The lowest spike observed when WiFi.begin() is called is approx 3.1V with a 3.2V trigger. The reset pin (EN) has also been tested with an oscilloscope in various situations including the WiFi crashes, but seems to be very stable.
- We are using a raw chip (ESP32-D0WDQ6) with our own PCB based antenna (with a ground free plan around the antenna) that seems to work in another similar version of our PCB.
- The PCB uses a 4 layer stack up to reduce the ground bounce.
- RF traces have been routed as per ESP32 Hardware design guidelines and we should have maintained enough clearance/protection for RF trace.