Gary Turner

Finally! I built & tested an open-source concept model air purifier that should be good for schools. Lots of air (CADR > 420 CFM), quiet (<45 dBA), uses <9 W, and ‘cause it’s wall-mounted & down-blowing, it moves air in the right direction & takes up no floor space.👀my ALTs🧵⬇️

@open_erv I wonder why they used heavy wires but only used 20/30. I'd have thought just use multiple lighter wires to share the load. That would be more flexible, too.

@NukitToBeSure Brilliant! Just f*in' brilliant!

Nukit PC Fan Gauge is live. Learn about it here: github.com/opennukit/Nuki… Test it here: filterboxbuilder.com/fangauge/ You can use these readings to test before you 3D print a plenum: Ambient pressure (hPa) 996.658 Ports closed 996.550 1 port open 996.570 2 ports open 996.612 It also makes for a good, low-cost STEM-Ed project, so please share.

@open_erv @gothburz isn't, in fact, a farmer or the owner of a $400,000 combine. He's a humorist, and a good one. I recommend his pinned post, which is a hoot.

Ok but without blaming the guy, we have to ask quite seriously why people buy them to begin with, when we know this is how it is. Because therein lies the answer with why John Deere keeps getting away with it. There is a whole community of people that tell people not to do that, that design better machines. People still do it. They don't buy the better machines, either. They don't support the development of better machines. For whatever faults JD has, people flock to them because, I suppose, they believe their tractors are superior in some way. So they do actually do a good job at something, then. I hope it was worth it.

@Engineer_Wong Thank you for sharing this, Adam. I am so very sorry for your loss.

My 85‑year‑old grandfather passed away recently. I flew back to the city where I was born to attend his funeral. I grew up with him, so we shared a very deep emotional bond. When he was admitted to the hospital, imaging scans showed that his lungs had turned white. Given the ongoing COVID‑19 pandemic, it was almost certain that he had contracted the virus. The hospital said his body was too frail to survive, so they only provided oxygen support; he died two days later. He had been diagnosed with stomach cancer three years ago and underwent a total gastrectomy. Without a stomach, he became extremely emaciated. The cancer recurred six months ago, but because of severe malnutrition, the tumor grew very slowly. This time, however, the COVID‑19 infection overwhelmed his already weakened immune system, and he simply could not fight it off. He is the first person in my family to have died from COVID‑19.

@open_erv WRT motors, I was thinking your vendor might ship direct 2 your customer, higher cost than a batch shipment 2 you, but only 1 shipment per customer. I was imagining the molded fan might be quieter, since that's the component liable to turbulent flow where smoothness matters.

Yes, you could buy a random NEMA23 2 Nm motor and that would work, but unfortunately only 2 out of 6 motors I tested were actually completely silent. The others made various buzzing noises. So you might end up with inferior quality, or have to shop around to get a good motor. I even got some motors that were stuck, they were DOA. The molded fans are better mostly because they look better and will be physically stronger, hopefully enough that even dropping it isn't a big deal. I mean it's plenty strong even right now if it's treated reasonably but for true reliability making it really tough is a good idea. The main reason for molding is to increase production rate, though, it's going to be only 1 kit every 3 days or so I think, the way things are looking. Molding can up that to 5 per day or maybe more no problem, even without moving to a new premises.

I am going to start only engaging for short periods once per week on twitter, posting updates more like a newsletter for the BQF project. When I look at th 14,200 tweets I have supposedly made, only a very small fraction of that actually led to a useful outcome, and I need to run a tighter ship here with my time. Regarding the recent drama, I have not read it, I will not, I am not interested, I said a reasonable thing and I'm leaving it there. It's pretty clear the discussion is not going to go anywhere. I have a lot of real, difficult things to do and don't have time. I have not that much animosity towards Nathalie, I tried to work with her, it wasn't working out so that's the end of that. She can say what she wants, that's fine with me and it is up to the listeners to decide how much they want to listen, but I am not interested, and that's fair enough. This week's update is that the interim output grille solution is done. That was the last piece before I can start with getting the details all sorted out for the supply of the parts/their production. Then I need to make good instructions, and I can start selling kits. It's nylon black fish net, 10 mm holes, clamped down. It reduces airflow by about 2.3%, so not too bad. In an unexpected turn, I boosted the airflow by about 8% further with no impact on noise, by trimming the tips of the secondary. It's difficult to print this geometry which is why I did not try it before. Most of the parts are already sorted out, I got some new boards already, and 15 motors are on the way. Just gotta get the power supplies and shipping boxes sorted out and I'm just a few large format prints and some packing away from a few kits. I got a promising quote for the boxes, and am close for the power supplies. In the meantime I can just use good quality parts off Amazon, the boxes are a little harder. I am pivoting to kits rather than fully assembled fans. I was hesitant to do this at first because I wanted more progress faster, mostly. Assembled units are in general better for large scale roll out. However it's become clear this is going to take longer than that. By focussing on the kits first, I can focus on getting the parts supply, transport, duty/taxes etc all worked out, which is the next stage in general. The constraints regarding cosmetics etc. are less severe for kits, and it fits in with the goal of tremendous performance to cost ratio, and the diy nature of CR boxes. It also takes less of my labor per fan that gets out there, which is a bottleneck, and less space in the premises etc. Fundamentally, it is a form of collaboration with others. Teamwork makes the dream work. For now, everything will be printed, and production rate should be about 1 kit per 2.5 days. Hopefully soon I can get the largest part molded, production can rise to about 2 kits per day, a 5x increase, again bottlenecked by the large printer. With the second largest part also molded, and a few extra small printers, it goes to about 10 per day. Further expansion would probably require moving to a new premises and hiring someone to help. The break even per unit cost, including labor at a living wage, will be hard to meet while also providing spectacular value, mostly because of shipping and taxes. For instance, the motors are $11.8 USD, but after duty and shipping they come to about $53 USD, each. With a suitable sea courier service that's expected to go down to more like $20 USD, but it's still nearly twice the cost of the actual parts. Shipping would be about $4-5 of that. Fundamentally, making ends meet is sort of not a problem because if you compare the capacity and noise of the resulting air purifier (which I am trending towards dubbing an EQ-CR box, for Extra Quiet, preferrably with the 6x filter set (diagonal V in the middle)), it would cost you a few thousand dollars to get it any other way. If the fan was $1000 it would still make plenty of rational economic sense, but we want more progress than that even still. So I don't think things are on thin ice. However, getting the best result possible, which matches the dream to some degree of something more like $140 usd, is not so easy. Unfortunately production with a collaborator in China does not solve many of these challenges, and it also adds new ones. So for now, that's on the back burner again. Producing kits in small scale may seem thinking too small, but it beats just waiting. It does help get the ball rolling/pave the road.

@sdamico Don't limit the filters to HEPA, locking in a technology. High flow through high-MERV furnace filters can give better clean air delivery (CADR) at the same or lower noise level and cost.

This is awesome — but one thing I’ve wondered is if you could make a respiratory illness-free *daycare* by overprovisioning HEPA filters + using far-UVC. Would be an amazing experiment.

@Froglet1980 @NukitToBeSure Done it. Good luck, Froglet.

@NukitToBeSure you have more reach in our circles can you help me reconnect 🙏 you know my email to verify this

To answer some questions - nope, have no idea what happened, no reason given. Yup, files appeal, instant response denying it withing seconds. And no, I'm not ok. You can verify this is me by seeing update on fundraiser. gofundme.com/frog2026

@SmartAirFilters 1. High-MERV furnace filter instead of HEPA. 2. Wall-mounted, down blowing.

What would you like to see from Smart Air next? We want to make our products better for you.

@SmartAirFilters Even simpler: "just a filter and a fan."

Clean air is simple. Just a HEPA filter and a fan.

@Christo60498776 @NukitToBeSure How about a PVC tube, close fit for a ping-pong or foam ball, with a wide vertical slot for high flow at low pressure. DUT blows up into a cone feeding the tube. Calib should be pretty reproducible, mostly driven by slot width.

@NukitToBeSure That would be great. I have a few old mismatched PC fans laying around which wait for a second live

CADR calculator seems to be working and reasonably accurate. Also gives total current draw and total noise output along with other relevant stuff. Will try to finish and integrate it into the Builder at some point, but still mostly busy with the Torch 2. housefresh.com/nukit-tempest-…

@Christo60498776 @NukitToBeSure Replied above. Meant here: I'm wondering how to use the data to optimize. You'd have a flow-pressure curve for the ping-pong flowmeter (calibrate 1, make many) & each fan (of e.g. 6) would give one point on that curve. Then what? Just reject any below a set ball height?

@NukitToBeSure I'm wondering how to use the data to optimize. You'd have a flow-pressure curve for the ping-pong flowmeter (calibrate 1, make many) & each fan (of e.g. 6) would give one point on that curve. Then what? Just reject any below a set ball height?

I'd like to figure out a cheap DIY tool for pressure and flow testing scavenged fans. Some people spend crazy money on high-end RGB PC fans for these, and that's just not practical for most places. I had people in SE Asia making the Open Tempest from cardboard and old PC fans, along with whatever filters they could find, but there's no easy way to tell at a glance whether the old PC fans had enough pressure to be useful. With a cheap testing tool, I could set it up to maximize performance with mismatched, random fans.

@chrisgj198 @Engineer_Wong I imagine you can, but expect it to be noisier for the same flow, and you still need to protect the hole from the weather.

@Engineer_Wong Can you leave it indoors and reverse the fans? (Maybe also reverse the filters I guess, if they aren't symmetrical.) I have been looking at air-to-air heat exchangers to reduce the heating costs but good ones still seem expensive.

Many people are interested in positive pressure systems. Let me give you an example here. If you have a sheltered platform outside your wall, you can cut a hole in the wall and point the air outlet of the 3Pro toward that hole. This can create a positive pressure clean airflow of up to about 600 m³/h, which dilutes excessive carbon dioxide, particulate matter, bacteria, viruses, VOCs, and other substances in your home that affect your health. The trade-off: keeping your home at a comfortable temperature will consume extra energy. But usually, each person in a room only needs 30 m³/h of fresh air. So you can set the device to a very low level, and that way, both the noise and energy usage will be minimal.

@SmartAirFilters I imagine that's because masks are one-pass, so they need higher fiber density (smaller spaces between fibers) than HVAC filters. So the impact capture mechanism extends to smaller particles and so does the crossover (MPPS).

@DtaGuy Right on! And masks too. 👍 (Although the MPPS seems to be a bit smaller, based on tests like these from 3M.)

HEPA filters don’t stop working below 0.3 µm. Around 0.3 µm is usually the most penetrating particle size (MPPS)—the “weak spot” that’s hardest to filter. Below that, diffusion (Brownian motion) becomes very effective, so HEPA can capture nanoparticles extremely well.

@NukitToBeSure @o_bel_ Nukit's answer here is right on. I think the papers sited, while valent efforts, do too much with minimal data. 2.5 m/s=9km/hr=5.6mph isn't a hard threshold; only ~6 points below half that and accounting for incubation very weak. If you can feel a breeze outside it's prolly good.

To be honest, an outdoor location with a decent breeze, possibly a small rise or hilltop, even a beach? You can scout with a cheap anemometer. Some pavilions would be ok so long as the sides are open. Transmission is still possible of course, but much, much harder than indoors, particularly with good air movement. ≥2.5 m/s will get you better mitigation through diffusion than almost any affordable indoor setup. You should still push masks though. A wind speed threshold for increased outdoor transmission of coronavirus: an ecological study pmc.ncbi.nlm.nih.gov/articles/PMC86… Simple quantitative assessment of the outdoor versus indoor airborne transmission of viruses and COVID-19 pmc.ncbi.nlm.nih.gov/articles/PMC80… Outdoor Transmission of SARS-CoV-2 and Other Respiratory Viruses: A Systematic Review pubmed.ncbi.nlm.nih.gov/33249484/

Please share resources for planning a COVID safe wedding! I am intimidated

@TheMemeticist OK, thanks. I feel slightly less dumb.

@DtaGuy Labels/context were cutoff by the screenshot crop. Redline is cases, black is deaths. You can see the full context in the site at thememeticist.github.io/MemeLabV3/#/si…

Just fired up MemeLab and simulated the CDC’s worst-case for this Ebola outbreak under only 20% isolation. Nailed the same explosive trajectory they warned about. But here’s the twist: my run hits that grim death toll with way fewer total cases. Low isolation = curve still takes off hard, even with quarantine, vaccine/masks, and on/off lockdowns kicking in later. Why fewer cases for the same deaths? I’m running a spatial cellular automaton (N×N grid, local neighborhood spread, finite population). Once local clusters burn through susceptibles, transmission naturally slows even without perfect controls. Looks like CDC used a stochastic branching-process model built for early explosive spread in an unlimited pool. Different tools, different assumptions. That’s the beauty of modeling: - Branching processes shine at “how likely is a massive early blowup?” - Spatial CA sims like mine show the full burn local clustering, real-world constraints, and the long tail. These projections aren’t destiny. Play with the exact setup I used at thememeticist.github.io/MemeLabv3/ tweak the dials yourself and watch the curve bend.

@IntegralAnswers Great thread about great review. Thanks. But I need an article that just proves long covid is real, to convince some technical but conservative friends that just naturally assume people are lazy and want a freebie. Can you suggest something?

1/ Long COVID is one of the most complex post-infectious syndromes ever studied. A new review in Nature Communications Medicine attempts to unify the biology. Here’s what’s established, what’s emerging, and what’s still speculative. 🧵