Question: Sanity check for air sensors

Ag8n asked on August 31, 2018 16:58
87 | 0 answers | #17032


One sanity check often used for particle sensors was to block the input with a good quality air filter and make sure the number of particles dropped to zero. If it didn't drop to zero, probably the particle counter was messed up.

Is this approach viable for pm2.5 and pm10 sensors? What kind of filter would you use?

Thank you!



13 Comments

I can try this right now, at the Providence Purple Air sensor:

https://www.purpleair.com/map?#12/41.8091/-71.4266

Is this a question? Click here to post it to the Questions page.


DSCN0015.JPG

OK, it's every 10 minutes, and I just covered it:

Screen_Shot_2018-08-31_at_12.58.57_PM.png

Is this a question? Click here to post it to the Questions page.


OK, I'm removing the cover now, wow that seems to have worked!

Screen_Shot_2018-08-31_at_2.02.59_PM.png


@warren Cool experiment. So your dip in the top graph--do you think that is expressing less traffic due to start of holiday or is it raining there?

Is this a question? Click here to post it to the Questions page.


Nice "sanity check" for pm "real time" monitors. I'd suggest it is always good practice when testing to make use of original data, in the case of purple air that would mean the logged concentrations. Below is a graph using the recorded concentrations that I downloaded. There is certainly a drop in concentrations but it is not to zero. Sorry I don't understand how to put an image here. image.png

image.png


you should be able to just drag in an image to the text area!

On Tue, Sep 4, 2018 at 6:05 PM \<notifications@publiclab.org> wrote:

Hi! There's been a response to a discussion you're involved in. Do NOT reply to this email; click this link to respond:

https://publiclab.org/questions/Ag8n/08-31-2018/sanity-check-for-air-sensors#answer-0-comment-20572

jeffalk wrote:


Nice "sanity check" for pm "real time" monitors. I'd suggest it is always good practice when testing to make use of original data, in the case of purple air that would mean the logged concentrations. Below is a graph using the recorded concentrations that I downloaded. There is certainly a drop in concentrations but it is not to zero. Sorry I don't understand how to put an image here.


Look like spam? Mark this as Spam

Reply at: https://publiclab.org/questions/Ag8n/08-31-2018/sanity-check-for-air-sensors#answer-0-comment

Report abuse to: moderators@publiclab.org

Check out the blog at https://publiclab.org/blog | Love our work? Become a Public Lab Sustaining Member today at https://publiclab.org/donate

Is this a question? Click here to post it to the Questions page.


I agree that the original data should be used. But, if the sanity check is not zero, that result should also be recorded, or at least noted.


I tried to drag an image, tried copying and pasting, tried putting the image into a WORD document and dragging or pasting, you see the results above.


@jeffalk does it work if you convert image to jpg format?

Is this a question? Click here to post it to the Questions page.


I'll try to unpack a couple of things here.

The use of zero filters is standard practice as a quick check of particle instruments. It is used to test for leaks in a sampling line easier than with a vacuum test. It is also used, if you're sure that there are no leaks, to test if there is any dirt in the optics of laser particle counters.

During a measurement campaign, many years ago, I was using an optical particle counter that seemed to report too large concentrations and a distribution that was too uniform and it turned out that a little fibre (likely plant based) got caught in the optics chamber and the laser was scattering of the fibre which meant that the instrument was reporting particles that weren't there.

Now, to ask the original question, any "respirator" kind of filter will remove ALL particles larger than 200nm (the existing laser based particle counters cannot see particles smaller than that because of the wavelength of light used). But the setup is not necessarily that simple.

Placing a bag over the sensor can seem tempting but there are risks with that approach. First there is heat management that without air flowing, components will get warmer. However, the more important consideration is about exhaust management. It is "conceivable" that the fan generates particles (very tiny) that when kept in an enclosed volume will coalesce and grow until they are large enough to be measured by the sensor.

So, what I would suggest to "zero-test" is to blow clean air over the sensor for a few minutes until the readings go to ZERO. Small laser particle sensors usually have a fan and fans are good at pushing air but are rubbish at sucking air so they are not able to suck air through a filter (unless it is a very low pressure drop filter). Also, these sensors have a "preferred" air path but since they are not sealed, air can get through the instrument joints.

Now, it is important that the readings go down to ZERO, no "near zero" or "very low" but actual ZERO because what you're testing here is that the instrument doesn't report particles that are not there. If the instrument is telling you that it sees particles even if you're only feeding it clean air, then there is something either in the optics (dirty sensor or bug in the instrument ... it has happened!) or, more serious, the signal from the light detector has a bias. If the sensor is dirty, just clean it but if the electronics have a bias, get a new sensor.

Having said all that, the tests reported by warren seem OK as a first pass and I would label those sensors as "not insane" 😃


I'll try placing the graph again. I'm sometimes clutzy at this stuff! I'm trying to use jpg format. public_lab_purple_test.jpg

public_lab_purple_test.jpg

public_lab_purple_test.jpg


Hi, Jeff - if you can't get it to work, please feel free to email it to me at jeff@publiclab.org and I can try. 

On Wed, Sep 5, 2018 at 1:34 PM \<notifications@publiclab.org> wrote:

Hi! There's been a response to a discussion you're involved in. Do NOT reply to this email; click this link to respond:

https://publiclab.org/questions/Ag8n/08-31-2018/sanity-check-for-air-sensors#answer-0-comment-20584

jeffalk wrote:


I'll try placing the graph again. I'm sometimes clutzy at this stuff! I'm trying to use jpg format.


Look like spam? Mark this as Spam

Reply at: https://publiclab.org/questions/Ag8n/08-31-2018/sanity-check-for-air-sensors#answer-0-comment

Report abuse to: moderators@publiclab.org

Check out the blog at https://publiclab.org/blog | Love our work? Become a Public Lab Sustaining Member today at https://publiclab.org/donate

Is this a question? Click here to post it to the Questions page.


This is a great thread! Excited to find this simple experiment. We might consider recommending this as an activity for people to try as they set up their own particulate matter sensors!


Log in to comment

0 Answers

Sign up or Login to post an answer to this question.