Add features via additional sensors to MJXDD001 / MJXDD01SYL (yeelink.light.ceiling22)

[B48D81EFCC]

Hello,

I want to put some I2C sensors into the lamp housing. Sensors to measure temperature, humidity, air pressure and/or brightness.

Unfortunately, the built-in board has very limited GPIOs, but for I2C just two additional GPIOs are needed.
I am wondering if the the TX/RX pins could be used for I2C.

Alternatively, a pin-compatible replacement board could be used. Any suggestions?

[B48D81EFCC]

Okay, I looked a bit closer to the PCB and I found 3 GPIOs labeled with R G B.
I guess these pins are not used in this lamp. I guess I could use these as well.

[syssi]

Good catch! I guess you are on the right track. Just assign the i2c component to these GPIOs.

[syssi]

Most Yeelight devices come equipped with a EEPROM accessible via i2c. In other words: There is a i2c bus already. If you are able to identify the EEPROM you could attach the additional i2c devices here.

[syssi]

This is what I'm talking about:

https://github.com/syssi/esphome-yeelight-ceiling-light/blob/main/yeelight_light_lamp9.yaml#L123-L128

[B48D81EFCC]

Oh interesting. But does the lamp use this bus already?
Is I2C Multi-Master capable? Can the bus have multiple masters?

[syssi]

No. The stock firmware uses the i2c bus to store some encryption keys. The ESP32 is the master and most of the time a single device (the EEPROM) is available on the bus. Just the bslamp2 is a bit special here because it has some touch controls etc.

[B48D81EFCC]

Okay, got it! Thanks.

[B48D81EFCC]

Some findings so far using GPIO26, GPIO27 and GPIO33.

When I use GPIO33 (SDA) and GPIO26 (SCL) for the bus, apparently some devices are found.

i2c:
  sda: GPIO33
  scl: GPIO26
  scan: true

[10:39:05][C][i2c.idf:075]: I2C Bus:
[10:39:05][C][i2c.idf:076]:   SDA Pin: GPIO33
[10:39:05][C][i2c.idf:077]:   SCL Pin: GPIO26
[10:39:05][C][i2c.idf:078]:   Frequency: 50000 Hz
[10:39:05][C][i2c.idf:084]:   Recovery: bus successfully recovered
[10:39:05][I][i2c.idf:094]: Results from i2c bus scan:
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x0E
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x12
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x17
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x20
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x24
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x3C
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x3D
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x42
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x49
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x58
[10:39:05][I][i2c.idf:100]: Found i2c device at address 0x74

Any other combination of the 3 GPIOs did not find any devices:

[10:45:05][C][i2c.idf:090]:   Recovery: failed, SDA is held low on the bus
[10:45:05][I][i2c.idf:094]: Results from i2c bus scan:
[10:45:05][I][i2c.idf:096]: Found no i2c devices!

Is this the EEPROM? Is each I2C address a storage register?
I have never worked with a EEPROM and I am wondering why so many devices were found.

[syssi]

Interesting. Normally there is just a single i2c device at a single address. The short answer to

Is this the EEPROM? Is each I2C address a storage register?

is no! :-)

[B48D81EFCC]

Good news. I was able to connect a BME280 (temp + humidity + pressure) and a BH1750 (lux) successfully to the built-in esp32.
Next step is to make the wiring nicely.

[syssi]

Good job! Did you use the existing i2c bus or a second one?

[B48D81EFCC]

I used a separate bus.
The existing one looks suspicious with all these i2c devices on the bus

[B48D81EFCC]

I finished the upgrade.

I replaced the built-in board with a Seeed Studio XIAO ESP32S3.
I attached a SHT30 (Temp. and Humidity) and a BH1750 (Illuminance).

[syssi]

Good job! Enjoy your unique ceiling light! :-)

[B48D81EFCC]

Hi,

I wanted to increase the brightness a bit and changed max_power: 0.88 to max_power: 1.00.
Will this operate the LEDs outside their specs and decrease the lifetime significantly?

[syssi]

I don't know. I measured the voltage limits of the original firmware just to make sure to operate the device in the same range as the manfacturer.