Particulate Monitor Design Initial version: Neilh 2012Apr17 I wonder is anybody interested in kicking around a Particulate monitor design in detail? Seems like there are some knowledgeable users interested in a open design particulate monitor. The design looks to me like it requires a combination of talents from the final air quality metric, to electronic to mechanical issues. From Tim Dye of Sonoma Technology Table

Worldwide

Time Resolution

R, H, V R, H, V H, V

~0-300 g/m3 ~0-500 g/m3 ~0-50 g/m3

35 g/m3 24-hr 150 g/m324-hr None6

1-5 g/m3 ~10 g/m3 1-5 g/m3

1 min to 1-hr 1 min to 1-hr 1 min to 1-hr

      Year-round     Year-round      Year-round

Indoor

Accuracy7

Outdoor

Health Standard3

Rural

Health Concern1

Atmospheric Concentration Range2

Urban

Pollutant Particles PM2.5 PM10 Black Carbon

Near Trans.

Location

Season

There is also a PM1.0 (a fourth range) Looking at the Sharp GP2Y1010AU0F Dust Sensor. It specifies a minimum Sensitivy of 0.35 V/(0.1mg/m3) and an output voltage range of 3.4V

The measuring algorithm is to pulse 100 times a second for 320uSeconds and sample output 280uS after turning on Led. The GP2Y1010AU0F units seem a bit strange as clearly it doesn’t have a m3 inside its package. ON its Internal Schematic is says “Dust through hole”. So as a digital snapshot it can provide a response to any particles that are there – but the IR reflection is likely to be dependent on many factors and might only be found by experimenting. And then a m3 of air has to be moved past the hole. However no specification for the depth of the field that dust can be detected – is it 1mm or 10mm

For the electronics the real challenge is the lower end of PM2.5 scale 35 g/m3 24-hr with an accuracy 1-5 g/m3 So this 24 hour measurement needs to be distributed over the total number of samples to be taken. The key it would seems is the accuracy of 1-5 g/m3 What would GP2Y1010AU0F output look like if it was detecting a particle of 1-5 g Well its specified for 350mV/(0.1mg/m3) so maybe this can be 3.5mV/1.0 g/m3

Which at least on paper looks doable If a 12bit ADC is used with a 4V range – then it can sample at 1mV. So that looks doable. Now moving a m3 in say useable time, say 1hr over the front of the hole is a mechanical challenge and possibly some more math. The hole is 8mm in diameter – but probably not all useable. If it is approximated to 4x4x4mm – then this is a volume of 64 mm3 So at a sample rate of a 360,000/hr this is or 0.023 m3 which over an hour is a significantly less than the target of 1 m3 (any holes in my maths). So this might require some rationalizing on getting to the air flow of 1 m3 / hour. So the next challenge would be how to differentiate between the three types of particulates – BlackCarbon, PM10 PM2.5 Presumably BlackCarbon is the larger particle – so the process may be about trying to filter out the three particle sizes and do it with a mass of air flowing over it, and then make three separate measurements. Any ideas? The other design areas that are a challenge are the * Moving air over the sensor – can probably be done with a low cost fan, but needs careful design to avoid particulate being blown into the sensor (see data sheet). All doable in detail design. * Environmental considerations – should survive temperature extremes outdoors from Lima to Patagonia, Jakarta to Bejing, Arizona/Florida to the Canadian North. * Should survive humidity fluctuations. ? Humidity not specified on data sheet. This is a red flag. May not work in humidity, or the very least design should be so condensing moisture drains away. “When inside of the sensor is moisturized, this product does not keep its proper function. Please design the application so that moisturization of the sensor does not happen.”