In this post, we shall explore the working principle and output signals of particulate matter sensors, and explore how the PSLab can be used as a data acquisition device for these.

Working Principle

A commonly used technique employed by particulate matter sensors is to study the diffraction of light by dust particles, and estimate the concentration based on a parameter termed the ‘occupancy factor’. The following image illustrates how the most elementary particle sensors work using a photogate, and a small heating element to ensure continuous air flow by convection.

Occupancy Rate

Each time a dust particle of aerodynamic diameters 2.5um passes through the lit area, a phenomenon called Mie scattering which defines scattering of an electromagnetic plane wave by a homogenous sphere of diameter comparable to the wavelength of incident light, results in a photo-signal to be detected by the photosensor.  In more accurate dust sensors, a single wavelength source with a high quality factor such as a laser is used instead of LEDs which typically have broader spectra.

The signal output from the photosensor is in the form of intermittent digital pulses whenever a particle is detected. The occupancy ratio can be determined by measuring the sum total of time when a positive signal was output from the sensor to the total averaging time. The readings can be taken over a fairly long amount of time such as 30 seconds in order to get a more accurate representation of the occupancy ratio.

Using the Logic analyzer to capture and interpret signals

The PSLab has a built-in logic analyzer that can acquire data signals up to 67 seconds long at its highest sampling rate, and this period is more than sufficient to record and interpret a dataset from a dust sensor. An inexpensive dust sensor, DSM501A was chosen for the readings, and the following results were obtained

A short averaging time has resulted in large fluctuations in the readings, and therefore it is important to maintain longer averaging times for stable measurements.

Recording data with a python script instead of the app

The output of the dust sensor must be connected to ID1 of the PSLab, and both devices must share a common ground which is a prerequisite for exchange of DC signals. All that is required is to start the logic analyzer in single channel mode, wait for a specified averging time, and interpret the acquired data

Record_dust_sensor.py

from PSL import sciencelab   #import the required library
import time
import numpy as np
I = sciencelab.connect()           #Create the instance
I.start_one_channel_LA(channel='ID1',channel_mode=1,trigger_mode=0)  #record all level changes
time.sleep(30)   #Wait for 30 seconds while the PSLab gathers data from the dust sensor
a,_,_,_,e =I.get_LA_initial_states()      #read the status of the logic analyzer
raw_data =I.fetch_long_data_from_LA(a,1)  #fetch number of samples available in chan #1
I.dchans[0].load_data(e,raw_data)  
stamps =I.dchans[0].timestamps    #Obtain a copy of the timestamps
if len(stamps)>2:   #If more than two timestamps are available (At least one dust particle was detected
		if not self.I.dchans[0].initial_state:   #Ensure the starting position of timestamps
			stamps = stamps[1:] - stamps[0]   # is in the LOW state
	diff = np.diff(stamps)   #create an array of individual time gaps between successive level changes


	lows = diff[::2]      #Array of time durations when a particle was not present
	highs = diff[1::2]    #Array of time durations when a particle was present
	low_occupancy = 100*sum(lows)/stamps[-1] #Occupancy ratio
print (low_occupancy) # datasheets of individual dust sensors also provide a mathematical
                      #equation to interpret the occupancy ratio as concentration of
				#particulate matter

Further Reading, and application notes:

[1] LED based  dust Sensor application note