Sonoff Basic R3 very slow LED blink, reset button unresponsive, ESP8285 brown-out suspected

User question

Sonoff basic r3 not workink , reset button not warking, led blinking very slowly

Artificial Intelligence Response

Direct answer to the question

• Most likely cause: the internal low‑voltage power supply is failing, so the ESP8285 is brown‑out/reset looping. In that state the firmware never runs, so the pushbutton appears “dead,” and the LED blinks very slowly.
• Key points:
  • Verify the 3.3 V rail first. If it’s below ~3.2 V or unstable, the PSU is the root cause.
  • Power the logic from an external, known‑good 3.3 V source to confirm. If it boots normally then, repair/replace the internal PSU parts; if it does not, investigate firmware corruption (serial re‑flash) or a failed ESP8285/flash.

Detailed problem analysis

• Symptom mapping
  • Very slow LED blink (one brief flash every few seconds) + unresponsive button almost always indicates the ESP8285 never reaches the main application. The usual reasons are:
  • Brown‑out (undervoltage) causing watchdog resets.
  • Corrupted firmware image that can’t pass early init.
  • Less commonly, MCU or flash hardware damage.
  • The Basic R3’s pushbutton is wired to a GPIO (GPIO0 on most revisions) with a pull‑up. If the MCU never boots, the firmware never samples the button → it seems inoperative.
• Power architecture (what typically fails)
  • Basic R3 uses a small offline AC→low‑voltage supply that feeds a 3.3 V linear regulator (commonly AMS1117‑3.3 or similar). Aging electrolytics or a weakened primary side will leave the 3.3 V rail low and/or very noisy.
  • Excess ripple or <3.2 V at the 3.3 V rail puts the ESP8285 into repeated brown‑out resets → slow, “heartbeat‑like” LED.
• What the slow blink is not
  • It is not the normal pairing pattern (which is typically faster). It also isn’t caused by a bad “reset” button in isolation—if the button were open‑circuit but power/firmware were OK, the unit would still boot and LEDs would follow normal states.

Current information and trends

• Field experience reports show a high incidence of Sonoff Basic R‑series units going unstable after brownouts/surges due to dried‑out electrolytics and stressed primary parts. Replacing the 3.3 V regulator’s output capacitor and, where present, the small primary bulk capacitor often restores stability.
• Note: Some web posts mention an “HLW8012” failure, but that IC is used in energy‑monitoring models (e.g., POW), not the Basic R3. Treat “HLW8012” as not applicable here.

Supporting explanations and details

• Target measurements (multimeter; device open, exercising mains‑safety precautions):
  • 3.3 V rail at the ESP8285 VCC or on the in‑circuit programming header: 3.25–3.40 V DC with minimal ripple. If you have a scope, aim for <100 mVpp ripple. Anything <3.2 V or visibly wobbling correlates with your symptoms.
  • Pre‑regulator rail (input to the AMS1117‑3.3): typically around 4.5–6.5 V (varies by board revision). If this is missing or low, the primary side of the AC/DC supply is at fault (fusible resistor open, rectifier/open, switcher IC failure, or primary cap dried out).
• Quick isolation test (no mains connected during the test):
  • Feed the logic 3.3 V directly:
  • Connect an external 3.3 V, 500 mA capable supply (or a USB‑to‑UART adapter that can provide a solid 3.3 V) to VCC and GND pads of the serial header.
  • Do not power the mains side at the same time.
  • Result A: Device enters pairing/rapid‑blink or responds to the button → internal PSU fault confirmed.
  • Result B: LED still shows the same very slow blink/no response → proceed to firmware/MCU checks.
• Firmware integrity check (serial):
  • With the external 3.3 V supply still attached, put the ESP8285 into bootloader: hold the button (GPIO0 low), then apply 3.3 V, release after ~2–3 s.
  • Use esptool.py to read the flash ID (esptool.py flash_id). If the chip responds and you can erase/flash, firmware corruption was likely. If the tool cannot sync after verifying connections, suspect MCU/flash damage.

Ethical and legal aspects

• Safety and compliance:
  • This device operates at mains voltage. Opening the enclosure and probing inside exposes hazardous potentials and energy storage in primary capacitors. Always disconnect from AC and wait a few minutes for discharge, and use an isolation transformer when live‑testing is unavoidable (professionals only).
  • Modifying the internal power supply or permanently bypassing it with an external adapter may void certifications and could violate local electrical codes. Do not deploy a modified unit in safety‑critical applications.

Practical guidelines

• Step‑by‑step workflow
  1. Unplug from mains. Open the case.
  2. Visual inspection: look for a charred fusible resistor near AC input, cracked solder joints around the AC/DC section, bulged/leaking electrolytics (small 10–100 µF low‑voltage cap near the 3.3 V regulator; small primary HV cap).
  3. External 3.3 V test as above. If it boots, repair the PSU:
     • Replace the AMS1117‑3.3 (or equivalent) if it runs hot or the 3.3 V rail sags under load.
     • Replace the 3.3 V regulator’s output capacitor (use fresh low‑ESR 22–47 µF, ≥6.3 V).
     • If the pre‑reg rail is unstable/low, replace the small primary bulk capacitor (µF value and voltage per board), and inspect/replace the input fusible resistor and rectifier components.
     • Reflow dull/cracked joints in the PSU area.
  4. If it does not boot on good 3.3 V, try serial flashing:
     • Wire TX↔RX, RX↔TX, GND↔GND; keep VCC at 3.3 V external. Enter bootloader (GPIO0 low on power‑up) and re‑flash a known‑good image (stock or Tasmota/ESPHome if you used them).
  5. If serial comms fail or flashing verifies but the device still won’t run, the ESP8285/flash is likely damaged → replacement is more economical than component‑level repair.
• Best practices
  • Never feed 5 V into the ESP VCC; it will destroy the MCU.
  • Use a supply that can source at least 300–500 mA at 3.3 V for testing; weak USB‑UART boards often droop under load and mislead diagnostics.
  • After PSU repair, soak‑test on mains for 30–60 minutes to confirm stability (watch LED behavior and 3.3 V rail).
• Potential challenges and how to overcome them
  • No obvious damage yet unstable 3.3 V: ESR of old caps may be high; replace on suspicion.
  • Button still “dead” after PSU fix: verify the button’s continuity and the pull‑up resistor to VCC; trace to the GPIO pad.

Possible disclaimers or additional notes

• Board revisions differ. Pad labels and component values vary across Basic R3 batches; use the silk‑screen where present.
• LED patterns depend on firmware (stock eWeLink vs Tasmota/ESPHome); “very slow” here means markedly slower than normal pairing blinks.

Suggestions for further research

• Look up: “Sonoff Basic R3 serial header pinout,” “ESP8285 GPIO0 flash mode,” “AMS1117‑3.3 stability and output capacitor ESR.”
• If you intend long‑term reliability in surge‑prone environments, consider adding an external MOV/surge protector upstream and ensuring your installation follows local electrical codes.

Brief summary

• Your symptoms match an internal power supply that can’t hold the 3.3 V rail, causing brown‑outs so the MCU never boots; hence the very slow LED blink and non‑responsive button. Confirm by powering the logic with a known‑good 3.3 V source. If it then behaves, repair the PSU (regulator + caps, inspect primary). If not, attempt a serial re‑flash; if that fails, the MCU/flash is likely damaged and replacement is most practical.

If you can, tell me:

• Approximate blink rate (e.g., once every 2–3 s).
• Whether it’s running stock eWeLink firmware or custom (Tasmota/ESPHome).
• Whether you have a multimeter and/or a USB‑to‑UART adapter. With that, I can tailor exact test points and part suggestions for your board revision.