RAK10702 - PM Sensor Not Working

I recently purchased several RAK10702 IAQ kits with the PM+ VOC sensor + e-ink screen. I have not been able to get any readings to show/report from the PM sensor (RAK12039). I have tried using the original firmware shipped on the device as well as through using the following repository: GitHub - RAKWireless/RAK10702-Indoor-Comfort: RAK10702 Indoor Air Quality Sensor

I have also tried increasing the delay before the if (!PMSA003I.begin()) line, in case the sensor has been slow starting up. I have tried powering the device using USB as well as using the 5-24VDC connection.

I have tried all of this on two different assembled sensor kits.

I suspect this is a firmware issue, but am now at a loss of what to try. Suggestions are appreciated!

Welcome to the forum @buildings

Did the kit come with pre-installed modules or did you install the modules by yourself?
Are the modules in the slots they should be?

When powering up the sensor, does the fan start in the PMSA003I start to work?

I installed the modules myself, but I double checked the installation instructions.

The fan does not start at all, however, the sensor does appear to have power when checked with a multimeter.

(Thanks for the response!)

I’ve done a bit more digging and Wire.beginTransmission(0x12) is returning an error of 2 (“received NACK on transmit of address”)

Is it possible the address of 0x12 is incorrect? When I scan the IC2 I see devices at the following: I2C device found at address 0x10 !
I2C device found at address 0x38 !
I2C device found at address 0x50 !
I2C device found at address 0x51 !
I2C device found at address 0x52 !
I2C device found at address 0x53 !
I2C device found at address 0x59 !
I2C device found at address 0x61 !
I2C device found at address 0x70 !

Can confirm the addressing in the firmware is not correct for at least some sensors. I.e. the lux sensor says it should be at 0x44 in the code but appears to be at 0x38.

Help with the correct addresses would be great.

The I2C address found are correct.
I2C device found at address 0x10 ! ==> RAK12010 Light sensor
I2C device found at address 0x38 ! ==> RAK14001 RGB LED
I2C device found at address 0x50 ! ==> RAK15000 EEPROM (on base board
I2C device found at address 0x51 ! ==> RAK15000 EEPROM (on base board
I2C device found at address 0x52 ! ==> RAK15000 EEPROM (on base board
I2C device found at address 0x53 ! ==> RAK15000 EEPROM (on base board
I2C device found at address 0x59 ! ==> RAK12047 VOC sensor
I2C device found at address 0x61 ! ==> RAK12037 CO2 sensor
I2C device found at address 0x70 ! ==> RAK1901 Temperature sensor

The device (should) come with the RAK12010 light sensor (0x10), not with the RAK1903 (0x44)

But your I2C scan is missing the RAK12039 (0x12), which means

  • not powered up
  • not connected correctly
    • check IO module connector
    • check flex cable for correct orientation
    • check flex cable if it is plugged in fully in the connector on both sides.
  • damaged module or damaged sensor
    • any visible damage on the module or sensor?

Hmm, OK got it thanks. I can see it does have the RAK 12010 and not the RAK1903.

No visible damage on either the module connector, flex cable or sensor. I have triple checked orientation and connections & disconnected/reconnected each of these a few times to ensure all is good on both units.

Checked the continuity of the connections with multimeter and that looks ok to the sensor connector that the PMSA003I gets attached to. However, I don’t see 5v at the PMSA003I sensor at any point during the sensor start up cycle etc.

Do you get 5V on the modules connector?

What voltage do you get at these resistor pins? It should be ~3.3V when the sensor is enabled.


Was able to do a few more tests in the light of day. The flex cables provided with the kit are shaped backwards. To get readings + power to the board they have to be flipped over which causes wire routing issues, but does get PM measurements!. I didn’t notice originally as the issue is also present with the front sensor package and it still read values (albeit not correct ones of course) with the upside-down connector.

Thanks for the help.

:astonished:
When did you buy that kit?
There should be a 90 degree angled cable for the PM sensor, not a straight one.

With that cable you could not plug it the wrong way.

Just bought them May 31st! I have 8 more kits I haven’t opened yet, but assuming they are the same.

They are angle cables included, but they are the wrong orientation (e…g., if you plug them in the way they have to go to fit properly, the connectors are upside down)!

Can you please open a ticket at Contact Us

I am checking with production and R&D why the the cables are wrong.

Done! Sorry to be slow