[ale] Sensor nuts

[Alex Carver]

It's an interesting device but I think it's lacking for what you're
really attempting to do.  The data sheet from Sharp does not make any
claims about the final particle size that it can potentially detect (not
measure, that's an important distinction).  It mentions just "smoke" in
the datasheet which is probalby going to be around 1 micron diameter for
average smoke.  However, there are a range of particulates in the air
that are smaller than 1 micron (the EPA tracks PM10 and PM2.5 which is
10 microns and smaller and 2.5 microns and smaller).

For a short list:  Household dust has an expected range of 0.05 - 100
microns, atmospheric dust (important if windows are open) ranges 0.001 -
40, wood smoke is 0.2 - 3, rosin smoke (from solder) is 0.01 - 1,
tobacco smoke is 0.01 - 4, and anything combustion related including
cars is 0.01 - 2.5.

The problem with this detector is that it can't distinguish size, only
density.  So 10 particles at 10 microns diameter is the same as 1
particle of 100 microns or 100 particles of 1 micron (roughly).  It's a
reflectance measurement so it relies on the total cross section of all
particles in its field of view and makes an aggregate measurement.

This sensor will basically be able to tell you "this room is % dusty"
but it can't say "there's lots of floating hair combined with very fine
pollen or soot" or "your chimney flue is not open enough and causing
some smoke to back up into the house".  It's also not a forced air
sensor so sampling is not going to be right since it will depend on dust
settling into the detector opening.

What you actually need is a forced air, laser based sensor that can
distinguish sizes.  There are many around that could work within a
budget and give you some better numbers.    Here is one I found with a
quick search of "laser particulate sensor":

http://www.dfrobot.com/index.php?route=product/product&product_id=1272&search=SEN0177&description=true



Or you can get more commercial versions that have more bins for the
particle sizes:

http://www.nano-sense.com/en/P4000/p4000-fine-particles-probe.html

http://www.sens2b-sensors.com/en/item/pm2-5-laser-particle-sensor-module-for-air-purifier-application

Just keep searching for other options but make sure it bins into various
sizes instead of lumping everything into a single measurement.

On 2016-06-08 09:11, Chris Fowler wrote:
> Any sensor nuts here?
> 
> Yesterday Amazon dropped off a Sharp GP2Y1010AU0F and I have it working with an UNO.
> 
> Communications and code I can do.  I have no clue about measuring air quality. 
>   I don't even know what numbers are good or bad.
> 
> We want to make changes to the house.  One change includes pulling up all carpet 
> now, living with the subfloor for a month or so, and then laying the wood floor. 
>   Pulling the carpet now addresses allergens.  I need time to source the wood. 
>   I told the wife that we can't just make all these changes without knowing 
> those changes had a positive effect.  Measure, adjust, and measure.
> 
> I need to validate the code, the values, and sensor placement.  I also need to 
> remove the chains of USB so I can easily transport this device around the house. 
>   I do not have wifi shield.  I have many of the $3 UNO tiny devices.  I'm 
> thinking the FASTEST path of breaking that chain is to pair up the UNO with a 
> NodeMCU and use the uart of the NodeMCU to read the values and then use wifi to 
> submit them.  rrdtool will do the rest.
> 
> Eventually I could just move the code to the NodeMCU and I will not need to tiny 
> UNO.  The requirements of sampling are stringent.  I wanted to try this out with 
> firmata to demonstrate firmata capabilities with our device, but I'm thinking 
> the requirements are too strict.
> 
> In honor of all federal agencies are now prefixed with "Obama's"  I'm calling 
> this device "Obama's EPA's Air Quality Meter" :)
> 
> Yep, print it on white letters and post it to the nicely black painted scrap 
> wood I'll use to mount it.  You should see the "fake TV" I made and use when we 
> go on trips.
> 
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo
>