Smart Air

How I accidentally started a global air purifier company (and why I didn’t tell my PhD advisor!). quora.com/How-do-people-… @Quora

More: smartairfilters.com/en/blog/how-to…

Vacuuming a HEPA filter may help a little, but the benefit is usually small. 👇

@VentiloAngel Interesting! The Sickleflow 140 scored higher subjectively than the Sickleflow 120? Based on the sound graph, I would have guessed the 120 would do better. And sure! I'd enjoy hearing the sounds. I'm curious if my preferences would be correlated with yours. My guess is yes.

Yes, partly! Silverstone outperforms mainly because of its much higher flow (71.9 cfm in the box). But for sound pleasantness, the Sickleflow 140 actually scored highest subjectively — despite a less 'beautiful' spectrum. I seem to prefer deeper sounds — flat and high from 10-100hz, then steepest possible drop toward high frequencies, minimal spikes. Would you like the audio links?

The Starvind box Silverstone was not planned. It started with a simple question : which PC fan sounds the most pleasant at 36 dBA? After testing 6 fans, the data led me to the Silverstone. Then I tried to find something better — tested 2 more fans. Nah. Silverstone holds its ground.

@VentiloAngel Right on--there's a lot of room for error in calculating CADR from flow + datasheet filtration %. 👍 For my devil's advocate, I was assuming CADR came from the official CADR test (or a close approximation).

The answer is in tweet 14 — two reasons: 1. some people estimate CADR from flow + datasheet filtration % rather than chamber testing (because it is more accessible and quick test). Since real filtration is, more often than we think, far below spec, that estimated CADR will be too often overstated. 2. too much person still believe in 2026 you’re safer next to a “True HEPA” purifier because of the high % of filtration, but I what to make them think again: are you certain the air exiting has >99% PM1 filtration? Without measurement, that’s not a safe assumption

🧵Thread: Filtration % is NOT a filter property — it's a system property 1/ 94% PM1 filtration on a MERV 13 build. Only 59% on a build with an H11 HEPA filter. Same particle size range. Real chamber measurements. The % filtration is not a property of the filter. It's a property of the entire system: filter + fan + plenum + seals + geometry etc.

We’re proud to share that we’ve removed 99% of plastic packaging from all our products. 🌱

Another place is joining our clean air network! 🌿 With our AIR+ Clean Air Inside program, we help keep indoor air clean and monitor its quality in real time.

@VentiloAngel It's neat to see the tent setup! Can I ask what you bought and hear more about pros/cons of the setup?

15/ A note on my protocol: % Filtration = CADR / Flow CADR: Measured in test chamber. Validated at start of each series against an AHAM/Energy Star/Intertek-certified reference machine. Weighting applied if needed. Target: ≤5% deviation. Flow: Testo 405i anemometer. 10 sec/measurement, 40 measurements/fan (14 fans). Weighting factor of 0.9 based on comparison with Cybenetics independent tests on Noctua G2 and P14 Max. Test chamber Data Protocol details & raw data available on request (please be specific). Protocol essentially unchanged since spring 2023. docs.google.com/document/d/1eQ…

@VentiloAngel I deserved to get called out for only reading the headline tweet there and assuming I understood the thread! 😅 I just read it in more detail, and you were making a different, interesting point.

Thanks! But this thread addresses something different — not the flow effect caused by the filter (which has been well covered), but the % filtration itself varying depending on the build. Same filter, different builds, different settings → very different filtration %. That's the new part. Worth a deeper look at the data 🙂

@VentiloAngel Interesting! So the company sells the same filter with the same particle % rating in two different thicknesses?

10/ Aerostar (MERV 13) — rated ~50% LagoStyle Aerostar P14Max → 57.5% (velocity: 36.47 ft/min, filter area: 5.46 ft²) Luggable Aerostar 7 Sickleflow → 90.3% (velocity: 27.10 ft/min, filter area: 10.92 ft²) Same filter. 57% vs 90%.

@VentiloAngel One context where I've seen people really care about this is for portable air purifiers to prevent infection, although even there, people's interest seems to be based on intuition and logic, rather than empirical data. In my mind, I'm still uncertain it matters much above CADR.

6/ So how do you actually know the filtration % of a build? One solid approach: % Filtration = CADR / Flow Where CADR is measured in a test chamber, and flow is measured with an anemometer — not pulled from a spec sheet. Not the only way. But it gives you real numbers on the actual system.

Installing air filters in classrooms improved student test scores: math up by 0.20 and English by 0.18 standard deviations — equal to about 2.5 EXTRA MONTHS OF LEARNING.

@BCarfree @farhadali Oof, yes 🫣

@farhadali @SmartAirFilters We bring little bits of the street home in our brain.😬

Spending time in PM air pollution can make people more distracted. The interesting part is that it may not show up right away — the effect was seen a few hours after people had already left the polluted area. More: nature.com/articles/s4146…

@DrMLivingston @Qantas That's annoying! Honestly, my approach would be to just try to make sure they don't see it. 🫠

Was giving the Smart Air QT3 a go on @qantas flight to Brisbane, cabin crew initially thought it’s a small heater as it’s cold this morning. Explained what it was. Just had them taken off us @Qantas by cabin crew. Huh? First time ever on any flight either domestic or international always had HEPA allowed on any flights previously since 2021. They are hand held and can easily be held for take off and landing.

@unify33333 That's right! 👍 It's the same with MERV filters and N95 masks. Here's a test of a whole bunch of MERVs, from MERV 16 all the way to MERV 5. The particle capture results show that same rebound upward for smaller and smaller particles.

@SmartAirFilters Same with MERV 13?

HEPA filters don’t stop working below 0.3 µm. 🔻

Perry!

MARCH RECAP 🩵 59 CR boxes granted to help provide cleaner air to schools and community spaces. That's clean air for approximately 235 children & community members! 🙌 Please DONATE to help us continue: https:// corsirosenthalfoundation.org/donate