ESPectre Interference with PIR Sensors

[bugensui2022]

I deployed the Spectre project using my ESP32s3 at home, and it performed very well, exceeding my expectations and being very accurate. I added sample code in ESPHome to implement it, so it does not affect the functionality of other devices on ESP32s3. The principle of this project should be to detect the waveform of WiFi. I don't know if it caused WiFi storms or interference, but it can lead to frequent misjudgment of the human infrared sensor on my other ESP32 development board. If I delete the code of this project, the misjudgment situation will disappear. It's really strange. I'm not an expert in WiFi, and I don't know if you have encountered it before. My Spectre project is deployed in the living room and kitchen. I use ESP32 for my ESP32 development board. We have deployed millimeter wave radar for detecting human presence and human infrared sensors, which are also implemented using ESPHome to detect human presence and absence, Spectre can interfere with the frequent misjudgment of human infrared sensors in the kitchen, but it does not affect the presence of human bodies in millimeter wave radar, and devices in other areas are also not affected. I don't know what happened?

[francescopace]

Hi @bugensui2022, thank you for the detailed report! This is a very interesting observation.

What you're experiencing is most likely RF interference from ESPectre's traffic generator affecting the analog circuits of your PIR sensor.

ESPectre needs continuous WiFi traffic to extract CSI (Channel State Information) data. By default, it sends 100 UDP broadcast packets per second (traffic_generator_rate: 100) to stimulate the WiFi channel. This RF activity can interfere with nearby analog sensors.

You may ask why only the PIR sensor is affected?

• PIR sensors use high-impedance analog comparators that are sensitive to EMI (Electromagnetic Interference). Low-cost PIR sensor often lack proper RF shielding.

• Millimeter-wave radars operate at 24GHz or 60GHz and have dedicated RF filters that reject 2.4GHz signals - that's why your mmWave sensor is unaffected.

• Proximity matters - the interference decreases with distance (inverse square law), explaining why only the kitchen PIR is affected.

You can try to:

1. Physical separation

If possible, increase the distance between the ESPectre device and the PIR sensor. Even 1-2 meters can make a significant difference.

2. Use different WiFi channels

Try changing your router's WiFi channel. Some frequencies may cause less interference with your specific PIR sensor's circuitry.

3. Reduce traffic generator rate

Lower the packet rate to reduce RF emissions while maintaining motion detection:

espectre:
  traffic_generator_rate: 50  # default is 100

The motion detection will still work, though with slightly reduced responsiveness.

4. Add software debounce to your PIR sensor

In your PIR sensor configuration, add filtering:

binary_sensor:
  - platform: gpio
    pin: GPIOxx
    name: "Kitchen PIR"
    device_class: motion
    filters:
      - delayed_on: 100ms      # Ignore short spikes
      - delayed_off: 5s        # Keep motion active for 5s

5. Hardware shielding

If the above don't help, you could add a small metal shield around the PIR sensor or its wiring.

---

Please let me know if reducing the traffic_generator_rate helps. I'm curious to know what value works for your setup.

This is valuable feedback - I'll add a note about potential EMI interference with sensitive analog sensors in the documentation.

[bugensui2022]

Thank you for your professional response, which has clarified my confusion. After two days of observation, I have adjusted the value of traffic_generator_rate from 50 to 40 as you suggested. Now, the pir motion sensor still triggers falsely, but the frequency has decreased. It triggers falsely 3-4 times every 7-8 hours, which I can currently accept. The situation has indeed improved.

Currently, it only occurs in the kitchen, not in other areas such as the bathroom or study. My ESPectre project is only testing on the esp32s3 device installed in the living room. Why is the pir human body sensor in the kitchen triggering falsely? The esp32 devices in the bathroom and study areas are the same as those in the kitchen, both being esp32 development boards connected to millimeter-wave radar human sensors and pir human body sensors to determine whether there is someone present. The sensors and esp32 development boards are the same, as are the esphome codes. The pir human body sensors in the bathroom and study areas do not trigger falsely.

Why is it that only the PIR motion sensor in the kitchen is interfered with? I speculate there are two possibilities: 1. The PIR motion sensor in the kitchen area may not be of good quality, or 2. The distance or space between the living room and the kitchen may be the cause. These are just my guesses for your reference.

[francescopace]

Hi @bugensui2022,
Your hypotheses are both correct:

1. PIR sensor quality variation - Even sensors from the same batch can have slight differences in shielding and component quality. The kitchen PIR may simply be more susceptible to EMI.

2. Distance and spatial layout - This is likely the main factor. The kitchen is probably closer to the living room (where ESPectre is running) than the bathroom and study. RF interference follows the inverse square law: doubling the distance reduces interference by 4x.

Additionally, the kitchen environment itself may contribute:

• Metal appliances (refrigerator, microwave, oven) can reflect and concentrate RF signals
• Power wiring in walls can act as antennas, coupling interference to nearby sensors

The fact that reducing traffic_generator_rate to 40 helped confirms it's RF interference, not a sensor defect.

If you want to eliminate the remaining false positives, try adding the delayed_on filter to your kitchen PIR configuration:

binary_sensor:
  - platform: gpio
    pin: GPIOxx
    name: "Kitchen PIR"
    device_class: motion
    filters:
      - delayed_on: 100ms

This will filter out short RF-induced spikes. See ESPHome Binary Sensor Filters documentation for more details.

[bugensui2022]

Thank you, this code is very useful