Many people have come up with calibration models to make the data output match popular units that are used in research and academia. I have chosen to make my units report in micrograms per meter cubed (ug/m^3).
To do this I compared the output of the sensor’s ratio data with a Dusttrak II Model 8530 air sampler with a PM2.5 impactor attached. I sampled the Dusttrak and my sensor at 1 minute intervals.
For my PM 2.5 (particulate matter 2.5 um) source I used matches and candle smoke. Since smoke is one of the major components that I am interested in this should provide my project with the best calibration.
The calibration curve I came up with is:
PM25= .1809 * (ratioP2)^2 + 3.8987 * ratioP2, (+ 2.5003)
R^2=.9858 from a PM2.5 concentration of 0 up to 600 ug/m^3
I removed the 2.5003 from the regression because I wanted a zero ratio to read zero ug/m3. It did not change the R^2 value much at all.
Keep in mind that my regression curve is different than other published curves because I changed the threshold voltage by using a resistor across Pin 5 on the senor and reading off of P2 (pin 2) on the sensor instead of the P1 (pin 4).
Laboratory Evaluation and Calibration of Three Low-Cost Particle Sensors for Particulate Matter Measurement – You have to pay for the full text or find a database you have access to that can get you the full copy.