[Karawebnetwork]

While it may not be the best, it still does the job.

My partner smokes indoor when I am not present and there's no smoke odor when I get home.

According to the chart on their app, the peak low air quality from several cigarettes goes away within an hour or so. https://imgur.com/a/0qgwT1F

[ortusdux]

I've never claimed that it does not filter, just that it is a fraction as effective as the alternative at 5x the price. We heat with wood in the winter and some smoke escapes when you open the door to tend the fire. My $100 unit filters the living room back down to the baseline in ~15 min.

[jamiek88]

Can you let me know the unit you have please?

[ortusdux]

https://breathequality.com/winix-c545-review/

There are definitely better units out there, but not for $100. That was the sale price at Costco when I got mine.

[IgorPartola]

I recently got into building DIY air quality sensors. Any kind of smoke hangs around in a room for many hours well past when you can smell it. My house periodically spikes in air pollutants (still looking for the cause but we are moving soon anyways). The one night when PM2.5 spike from safe levels of below 12 ug/m^3 to over 1000 for one night was the worst sleep everyone in our house got. It seems that it was the equivalent of smoking something like 44 cigarettes in a day. Don’t smoke, especially indoors. It’s a bad time.

[Karawebnetwork]

> Any kind of smoke hangs around in a room for many hours well past when you can smell it.

But once the sensors indicate that the PM2.5 is gone, everything is fine, right?

[IgorPartola]

If that's sarcasm, you aren't providing enough context for me to pick up what you are putting down. If not, then the honest answer is "I don't know for sure, I'm still researching it."

From what I can tell, PM2.5 isn't the end-all be-all metric and sensors can give false readings. That's why I'm building multiple devices with different types of sensors and monitoring multiple metrics. PM10 and PM1.0 are also concerns, as is CO and CO2. I am researching which VOC sensors to get because unfortunately not much info is readily available. But from the research I've read PM2.5 is the most prevalent and damaging in typical households (CO being a big exception, but that's also monitored by regular household CO alarms). I can tell when someone has been frying something in the kitchen from my bedroom sensor for example, and it lingers for a while. I can also tell when outdoor air quality is poor, and then my indoor air filter and closed windows do help. I am still learning about all this, but in general this data has been helpful to figure out when to open windows to avoid headaches.

I plan on publishing my findings at some point, but currently I am still waiting on parts and PCBs and working out the software to make it more usable without having to run to grab a USB cable to flash new firmware on the sensors. I was inspired by https://www.airgradient.com/diy/ but those plans are outdated and the dashboard is proprietary so less than ideal. I am working with simply integrating my sensor network with Home Assistant so I have to do very little with frontend stuff. It was very quick to set up notifications to my phone so I don't have to look at sensor screens all the time.

[adolph]

Are you looking at BME680 vs CCS811 vs SGP30 or something else?

https://www.jaredwolff.com/finding-the-best-tvoc-sensor-ccs8...

[IgorPartola]

I am using the PMSA003 for the PMx sensor, the DS-CO2-20 for the CO2 sensor, and was about to order this sensor breakout for CO and other gasses though currently it looks out of stock: https://www.seeedstudio.com/Grove-Multichannel-Gas-Sensor-v2...

I might play around with the BME680 but it’s VOC detection is very rudimentary and it does eCO2 which is basically a guess. There is also the MICS-6814 breakout for relatively cheap on AliExpress but reviews on it are mixed, saying it uses the wrong values for its resistors and capacitors and it’s not as accurate as the Grove one. The Grove board is the only one that measures separate gasses separately instead of measuring total gasses and then using a formula to try to divide the number down to expected parts.

In practice, who knows. I need to lay my hands on all of them. If I could get all this into a controlled environment to test it all that would really be ideal.

Thanks for that link, it was a great read and the comments were helpful as well.

[Karawebnetwork]

> but those plans are outdated and the dashboard is proprietary so less than ideal

Have you looked at open source smart vivarium projects? Several seem to have good front-end.

This may seem weird, but the reason I first started to explore air quality and sensors was to build a smart vivarium for a reptile.

I didn't go through with that project but it was a waking call for me. Animals need so much environment control to thrive and humans are animals.

So instead I focused on adding sensors and filters to my own home. Nothing fancy so far, just a mix of product that work with google home and smart sockets.

[IgorPartola]

I have not but it sounds interesting! Do you have any links I could use as a jumping off point?

[ahaucnx]

Thanks for mentioning our DIY sensor building instructions. Can you please let me know what exactly is outdated so that I can update it. Also I am more than happy to give people trial access to our dashboard. Just PM me via the support form on our website.

[IgorPartola]

Hey! I actually emailed you when I started but haven’t heard back so I assumed the project was not being worked on.

The biggest thing I found was that the PMS sensor has seen two generations of revisions. I built an original AirGradient and it’s sitting on my night stand currently but for new builds I am using a PMSA003 instead of PMS5003. Also I am substituting a DS-CO2-20 from Plantower for the CO2 sensor. That CO2 sensor seems to be well regarded and is slightly cheaper. And lastly I rewrote the drivers to be a bit different and a little more organized/readable/efficient. It doesn’t add much in terms of functionality, was just an excuse to write some C but I think it came out nice.

I redesigned the PCB to tie the PMS sensor’s SET and RESET pins to 3.3V per the data sheet. It seems to work fine without that with the PMS5003 but I haven’t yet received the new PCBs so I don’t know how it’ll work out. The PMSA003 has a different, slightly more convenient 1.27mm 2x5 connector so it’s easy to mount and get all the connections onto a PCB but isn’t breadboard friendly.

One other difference is that I’m not using a the Wemos OLED shield in favor of a stand-alone 0.96” OLED screen for double the resolution horizontally (your instructions show that one but not in the PCB build). And I wrote a different UI for it, plus added a button the PCB to do things like reset the sensor settings.

There are probably other things I’m forgetting but my email is in my profile if you want to move this conversation to that medium and once I polish up the code a bit more I am happy to share it. I will likely continue using these sensors with HA but I see no reason that you can’t use some of this in your work if you find it useful. I was planning on publishing the code and plans under a BSD license.

Overall, I wanted to thank you for the amazing work you did on AirGradient. I wouldn’t have been able to start on this project nearly as easily without your plans and code. It’s really impressive work very polished.

[Normal_gaussian]

are you able to share anything / willing to work with anyone? I'm just getting through the research phase and am looking to coat my house with sensors - its been a struggle finding quality writing in the research and it looks like a graveyard in the world of pcb design.

[IgorPartola]

Absolutely. See my reply to the sibling comment and shoot me an email (in my profile). Happy to share what I found out.

[Karawebnetwork]

Not sarcasm. Thank you for answering!

[IgorPartola]

:)