Complex calibration formula for pm2.5 sensor

[Alan Stankevitz]

Newbie here...

I am using the G1000 driver for Ecowitt sensors. Works fine but need to apply a complex calculation to correct the pm2.5 sensor data.

The equation would be something like:

Cpm2.5 = (pm2.5/(1+((.25*RH)*(.25*RH))/(1-RH)))

RH = Relative Humidity
pm2.5 is raw pm2.5 data

Cpm2.5 is corrected data

Is this possible to apply this formula and would the equation be placed under the StdCalibrate section of the config file?

Thanks!

[Graham Eddy]

in [[Corrections]] section of weewx.conf you can do something like

pm2_5 = 0.52*pm2_5 - 0.085*outHumidity + 5.71

slightly off-topic, does anyone know if the purpleair correction applies to other sensors, or just the purpleair?

[Alan Stankevitz]

Thanks for the info. Regarding your slightly off-topic question, purpleair and ecowitt use similar light-scattering detectors made by plantower. The purpleair has two sensors that average out the readings and I believe are doing some internal coding to rectify the humidity problem. The purpleair unit has a humidity sensor and barometer with the two pm sensors and I believe they are there for calculation purposes. The formula I have been testing was actually written for the purpleair unit by the University of Colorado during a field-test study.

The purpleair unit also detects pm10 as well. I have been attempting to make corrections to the Ecowitt unit cuz it's much less expensive than the purpleair sensor. The guts are very similar, so I figured I should be able to correct the humidity error by using math!

[Graham Eddy]

i have long thought my pair of ecowitt units read too high, but i have no reference hereabouts to calibrate them.
does anyone know of a correction algorithm for them?

[vince]

How would you know how to calibrate them without a known-good reference device ?

How would you know how much to calibrate for humidity anyway ?

FWIW, I've been using weewx to save my PurpleAir data for over 27 months of data into a separeare db if you guys want a copy in order to do some checking of whether your correction algorithm lines up with what you 'think' PurpleAir is doing when they calculate AQI that they report.   Email me if you want a copy.