VERSION 3: wiring self-test, corrected frequency and automatic receive scan.

Open Efergy8266.ino in Arduino IDE; keep EfergyDecoder.h alongside it.
Choose NodeMCU 1.0 (ESP-12E Module), upload, then use Serial Monitor at 115200 baud.
Wiring is unchanged: VCC=3V3 GND=GND SCK=D5 MISO=D6 MOSI=D7 CS=D0 GDO0=D2 GDO2=D1.

Startup statically toggles the CC1101 GDO0/GDO2 output pins separately and checks
they appear on D2 and D1. Both should report PASS. If either fails, check the
pin labels and jumper wiring; the sketch stops reception. This test does not transmit.

The old firmware inherited frequency word 0x10ABD1 from a project whose log
said 433.55 MHz. At the assumed 26 MHz crystal, that word actually tunes to
433.450104 MHz. V3 corrects it to 0x10ACCD (433.550079 MHz).
The crystal assumption must match the physical CC1101 module.

V3 listens 15 seconds each at 433.550, 433.500, 433.450, 433.600, 433.650,
433.700, 433.400 and 433.350 MHz. A full cycle takes roughly 2 minutes.
It holds the current frequency after two valid readings from an ID previously
observed in your Flipper capture (47183 or 26649). Otherwise it repeats the scan.
At each frequency change it prints TIMINGS: a sample near the longest matching
bit run, with positive=high, negative=low, durations in microseconds.
These are diagnostic samples and are not necessarily Efergy transmissions.
Share the entire 2-minute startup and scan output for analysis.
Attach the 433 MHz antenna and put the powered Efergy transmitter 1–2 metres away.

The decoder remains checksum-validated; Wi-Fi remains off.
Validation: offline replay decodes the known Flipper capture; corrected frequency
words are checked against the TI CC1101 formula. GDO test uses the documented
0x2F constant-low output and inverted 0x6F high output. New firmware compilation
and physical operation have not been verified here.

References:
https://www.ti.com/lit/ds/symlink/cc1101.pdf
https://github.com/merbanan/rtl_433/blob/master/src/devices/efergy_e2_classic.c
