After much testing and getting abysmally low values (for once i wanted more pollution!!) on the plots we have cleared some design flaws, which can be simplified here:
- Current flow to SPS30 could have been lower than expected. So upgraded a capacitor in the path to 1000uF from 470uF.
- Upgraded firmware library from Paulva‘s fantastic one (sorry Paul) to Sensirion’s recommended one with hopes of being in compliance with the company and then checking if all works better. Didn’t help much though!
- Thanks to my colleague Dr. Sumithra’s (SSLA, Pune) help, upgraded sampling and reporting method from median to average, and soon will shift over to mode.
- It is recommended that find dust samplers must 16.67 L/min of flow rate past the sensors to ensure high sensitivity. I could not measure, but i added a powerful 12VDC fan (0.6 A) to the system to emulate that effect.
So the outcome is better due to the above steps. How good and how close to actual, we can only test that with a calibrated instrument and i don’t know how we will get there.
Here are some pictures.
And the data can be seen here – https://thingspeak.com/channels/841841