Trying on ESP32-S2 QT Py & Pico W (3 of each to run each port with each power setting overnight). Not yet exercising the new get/set power APIs. Consistently getHard fault: memory access or instruction error somewhere right after the wifi connection (after printing the IP address. Code runs w/o safemode on 9.2.7.

adafruit-circuitpython-adafruit_qtpy_esp32s2-en_US-20250420-main-PR10271-f8ed46a.uf2
Adafruit CircuitPython 10.0.0-alpha.2-22-gf8ed46adce on 2025-04-20; Adafruit QT Py ESP32S2 with ESP32S2

adafruit-circuitpython-raspberry_pi_pico_w-en_US-20250420-main-PR10271-f8ed46a.uf2
Adafruit CircuitPython 10.0.0-alpha.2-22-gf8ed46adce on 2025-04-20; Raspberry Pi Pico W with rp2040

importtimeimportosimporttracebackimportsupervisorimportmicrocontrollerimportwifiimportsocketpoolimportadafruit_requestsTEXT_URL="http://wifitest.adafruit.com/testwifi/index.html"pool=socketpool.SocketPool(wifi.radio)requests=adafruit_requests.Session(pool)defconnect():whilenotwifi.radio.connected:try:wifi.radio.connect(os.getenv("WIFI_SSID"),os.getenv("WIFI_PASSWORD"))print(f"{wifi.radio.ipv4_address}")exceptExceptionasex:traceback.print_exception(ex,ex,ex.__traceback__)time.sleep(5)# uncomment one of the three as appropriate, or leave all commented for default MIN# wifi.radio.power_management = wifi.PowerManagement.NONEwifi.radio.power_management=wifi.PowerManagement.MIN# wifi.radio.power_management = wifi.PowerManagement.MAX# increase delay between requests each timeinterval_sec=0increment_sec=60timer_sec=time.time()-interval_secwhileTrue:iftime.time()-timer_sec>interval_sec:try:connect()print(f"Fetching URL...")withrequests.get(TEXT_URL)asr:print(f"{interval_sec:>5}{wifi.radio.power_management}{r.status_code}{r.reason.decode()}{r.content}")interval_sec+=increment_sectimer_sec=time.time()exceptExceptionasex:traceback.print_exception(ex,ex,ex.__traceback__)time.sleep(5)# supervisor.reload()# microcontroller.reset()

(edited to match proper API)

The crashes@anecdata is seeing are totally unrelated to this PR. I am seeing the same crashes on the tip ofmain right now, somain is busted in some way.

Testing now on all six boards. If it behaves like it used to (on Pico W), the more aggressive power management settings may causethe wifi connection to drop unrecoverable[Errno 2] No such file/directory as the time between requests increases.But, the field values for each setting are quite different than they used to be, so they may all work. I'll update tomorrow, once they have run long enough to confirm whether wifi stays connected at each power management level (it used to fail between 5-10 minutes on Pico W with increased power management).

Interim Update: Both ports, at each power management level, have run so far with 20-minute intervals between requests, with no wifi disconnections. So it's looking good for no regressions. There are a number of requests exceptions along the way (more onespressif than onraspberrypi):

Traceback (mostrecentcalllast):File"adafruit_requests.py",line658,inrequestOSError: [Errno116]ETIMEDOUTTheaboveexceptionwasthedirectcauseofthefollowingexception:Traceback (mostrecentcalllast):File"code.py",line37,in<module>File"adafruit_requests.py",line711,ingetFile"adafruit_requests.py",line671,inrequestOutOfRetries:Repeatedsocketfailures

butcode.py retries recover, and wifi stays connected.

I've repurposed theraspberrypiwifi.PowerManagement.NONE board to try to reproduce the original 2022 issue by forcing values ofcyw43.set_power_management(). It's possible some other pico-sdk changes have fixed that. It's also possible that network differences account for differences in outcomes.

I have not done any actual power usage testing.

@anecdata thank you for the testing! I'm marking this ready for review now.