Public Lab is launching a new 🎈 Mini Balloon & Kite Kit as part of Kickstarter Gold -- take a look!

Public Lab Research note


  • 6

Dustduino potentiometer testing

by imvec |

What we want to do

We want to verify what resistance value (position of the potentiometers) is needed to measure PM10 and PM2.5 particles using Shinyei PPD42.

Our attempt and results

We've set up two Shinyei PPD42 sensors toghether inside a black plastic box using two different arduinos and two raspberry pi's as datalogers.
We've used the SerialDuino code changing the sample time to 3600000 (one hour).
We've collected 96 hours of data (hourly measurements) of wich the first 48 hours the potentiometers were in "factory position". We didn't touch them.
The next 48 hours we put the potentiometers to zero and continued collecting data.
You can check the data here: stationA - stationB The main thing we've found out is that in "factory position" the sensors give different outputs. After this results (48h) we decided to turn the potentiometers to zero on both sensors and both potentiometers.
Results look more similar in between the two sensors but values are very very low in relation to the nearest public station.

Questions and next steps

Find out the value that the potentiometers might have to measure accurately in relation to public and certified air quality stations.
We'll be testing several set ups next weeks, deploying some devices close to the Poble Nou station in the city of Barcelona.
https://fair.to/TQGm3

Does anyone know the operational resistance values? We changed them not testing the value out of the box.
Anyone having a Dustduino working at home can use a multimeter to check the value of the potentiometers out of the box?

Why we're interested

We've found that the potentiometers come set up differently from factory and give different outputs.



calibration particulate-sensing barcelona spain air-quality activity dustduino dust pm open-air dust-monitoring espana shinyei dust-sensing

8 Comments

Found out that seedstudio folks have this image in the documentation but no value... https://raw.githubusercontent.com/SeeedDocument/Grove_Dust_Sensor/master/image/Grove-Dust_Sensor_cautions.jpg

We're re-setting the two sensors for other 48 hors putting the potes to that position and measuring them


If you want to check the station we are going to be using as reference, its exactly the one circled in the image, you can check it in this link https://fair.to/TQGm3

Captura_de_pantalla_2016-12-12_a_las_17.17.07.png

Zoom into Barcelona (area with more spots)


Here the PM10 (and NO2) Levels this afternoon.

Captura_de_pantalla_2016-12-12_a_las_17.17.27.png


Awesome post and good questions!

I did find this interesting post w some possibly helpful info:

https://publiclab.org/notes/Willie/04-01-2014/field-testing-the-shenyei-pm-sensor


thanks @warren,

that post (pdf of field calibrations) is the reason we changed the sampletime to 3600000ms (1h) as this guys have tested its in hourly messures when Shinyei PPD42 takes similar values to comercial/more expensive dustmeters, will read all this again, its a lot of info and maybe i've missed something...

At this point the only mistery is what value should the pots have, testing two sensors toguether they messure up and down at the same rate but they will not messure the same concentration at the same pot position, i'm sure theres a fixed value for both resistors where they will measure "more correctly" or more alike a comercial/more expensive meter.

We've arranged an station to test this days and see if we can find this value comparing with the local station we talk above, it would be awesome to have access to a more expensive meter to test this (we are working on that), or if someone working with Shinyei PPD42 has calibrated and could tell us the resistors value....

thanks again ;)


I have just found this: http://takingspace.org/wp-content/uploads/ShinyeiPPD42NS_Deconstruction_TracyAllen.pdf

"In the two that I tested, VR3 was set close 50kohm"

"VR1 was set at near 50khom"

Will be testing and posting soon ;)

Captura_de_pantalla_2016-12-15_a_las_9.58.26.png


Please excuse this basic question but what makes you think that all such devices should calibrate with the same resistance? You've shown that for some similar resistances they do not give the same output. There is variability that comes with inexpensive. It may well be the case that each such device must be calibrated individually, especially if you want your data to be trusted. I could not get the data links in your post to work (this may be due to my ignorance) but it is also important to convey just how you are going from particle counts to concentrations. This is not a simple process and margins of error may be such that two different device outputs may be statistically similar.


Hi @jeffalk! Thanks for your help!
That's not a basic question. That's in fact the kind of answer we've received from Shinyei, that every sensor has "different specification". We've asked a little bit more as we do not understand if they mean that the position of the potentiometers (as phisical screws) must be different depending on each sensor or that same resistor of different sensors will give different resistance. Our question is related to a multimeter measurement. Not precise, but yes aproximate.
We've wrote also to Seeedstudio folks since Shinyei customer service has told us that SeeedStudio is not their distributor, so they doubt their source of the sensor...
In relation to the conversion from counts to concentrations, since I'm not a techie maybe can help that we're doing the test using the Dustdino Serial code from Dustduino.org.
https://github.com/dustduino/dustduinoserial

PS: sorry but the links were empty. Data from those 48+48 hours are now availeable ;)

stationA - stationB


You must be logged in to comment.