Error reading sps30. about sps30

[paulvha]

Hi,

Without additional information it is hard to give any advice.

What kind of board / hardware setup do you use ?
How do you connect?
How do you power ?
Does it work with a standard sketch?

regards,
Paul

[roberbike]

Hi,

Without additional information it is hard to give any advice.

What kind of board / hardware setup do you use ? How do you connect? How do you power ? Does it work with a standard sketch?

regards, Paul

Hi Paul.
The hasdware uswd is TTGO T-display (esp32) + sps30 by i2c+ AM2320 by i2c, with a battery connected to the board and the usb charguing and powering the hardware.
The power for sps30 is 5V, supplied from a step up, fron 3.3 V to 5 V.
The firmware used is https://github.com/kike-canaries/canairio_firmware that use the sensors library https://github.com/kike-canaries/canairio_sensorlib.

Your library works , but if debug is activated, the error messages displayed before, are displayed by serial debug.
When this happen , the values appearing are the same than the last value measured.

Thanks and regards.
Robert.

[paulvha]

Hi Robert,

When a protocol error occurs either we did not get enough bytes or had CRC failures. Then the data that is provided should be neglected.

I would suspect a hardware issue:
It could be the power not being stable enough. I got very weird results in the past when connecting to 3V3 instead of 5V. Not only needed for the fan, but also the internal regulator. Not sure how well the step up is working or how stable. Maybe adding a capacitor on the 5V would help. Also check that it is a clean power signal so it does not have spikes etc

The other aspect is pull-resistors on the I2C line. The AM2320 does not have those (according to the datasheet) and the SPS30 has very weak (~50K). Both SDA and SCL should have pull resistors to 3V3 ( each resistor 4K7 to 10k)

It could also be bad line connections. Seen that before with others that I have helped.

regards,
Paul

[paulvha]

closed