A bug in the Open Air

[pcag]

Or, more precisely, lots of bugs!

Achim was right:

Accuracy comparison of PMS5003 with large interval values

I never saw 8000h in any official specs. We have Plantower 5003s happily running since 2019. We do not see any systematic failures. The most common failure is insects getting into it which the sleep mode might make more attractive because the fan is not spinning.

So now I want to put some mesh over the air holes to keep insects from returning, but I don’t want to throw off the readings. What material and mesh size would strike a good balance?

Nylon mesh is cheap and easy to come by, but I don’t know whether it will build up a static charge that skews the passage of small particles?

Stainless steel is also available, but I’m not sure what size mesh would work. 100-micron mesh only has about 50% open area. In theory 500-micron might be more open, but most things I’ve found are actually lower: probably due to thicker wire.

And I suppose fiberglass might be an option?

Thoughts? Recommendations?

[MallocArray]

I think anything you do it likely to impact the accuracy of all of the sensors, as you will be changing the airflow.

I’m sure mine has bugs in it as well, but I haven’t taken it apart and it still works, so I’m just leaving it be.

[Siriel_AirGradient]

Thank you for the picture! Did you observe weird readings with that Open Air?

The finer the mesh, the more likely it will impair the measurement. There are 2 concerns:

• The mesh might obstruct the air flow
• The particles might impact on the mesh, resulting in underestimated readings

Based on my experience in an aerosol lab:
I would assume that a 100 micron mesh does not drastically impair the flow or cause drastic impaction of PM2.5 given the relatively low flow of the Plantower sensor. Nevertheless, some of the particles will deposit on the mesh. I suggest doing a quick experiment by holding the mesh in front of the inlet for some minutes and then taking it off: the ‘jump’ in PM2.5 shows you the amount of deposited PM2.5. It is possible that there is no visible jump at all. You might need to use a candle or spray to increase the PM2.5 level if you are in a clean environment.

It is an individual assessment if the introduced inaccuracy justifies the gained protection.

[pcag]

@Siriel_AirGradient

Did you observe weird readings with that Open Air?

Yes, I did! I opened a support ticket about strange transient spikes, and at their suggestion checked on the Open Air.

Dozens of bugs scattered when I opened it. There even were half a dozen in (at least) one of the Plantower sensors! (Fragmentary remains of a dead one are still stuck in one of them.)

Here’s an example spike:

It’s pretty clear when the critters were crawling around inside the PM sensor… Some days I’d get no glitches, other days multiple, and only at night or early morning.

Your idea for a test is great. I’ll experiment with a few meshes.

Long-term, it would be nice for AirGradient folks to do their typical rigorous testing to find an optimal mesh along with any necessary compensation calculations…

[Siriel_AirGradient]

Thank your for sharing the glitches! That’s valuable as it helps others (including us) to better detect these infested monitors. Based on how big the issue is, we can then work on solutions.

[Narongkij_Tejasakuls]

In terms of mesh filter size, anything larger than PM 10. I would also consider correct placement for the units and disabling any lights to prevent insects from getting attracted to the device.

[Guilherme]

This is very interesting. After less than a week with an outdoor monitor in my balcony, today I already observed similar spikes on the PM time series:

Now seeing how similar this is to the graphs reported in this topic, I begin to wonder if some spider already got into my PM sensor.

Would you guys recommend to always open the monitor enclosure after observing such spikes?