wanderinggnome648
commented
1 month ago Cool. No worries, I'm actually still learning these interfaces, too. Few thoughts.
- My early designs put the fan inside, where it's also convenient for the fan to rest against. I changed it after some early designs that seemed like they should have been good weren't that good using the salt crystal particle counting method (shoot pure salt water in the air with a nebulizer and then track with a particle counter how quickly your purifier removes it in a confined space like a bathroom. I am not a fluid dynamics expert, but after many iterations and comparing against other purifier designs, I learned that the interior size needed to be minimized. Also, I learned that I couldn't just deck out the box with fans, but in certain positions (e.g., across from each other), they interfered. I also found a few papers and tweets on similar topics. On the fan placement inside vs. outside, I guessed that outside might be superior based on the Air Fanta 3Pro's design. My suspicion is that pulling the air out is better than pushing it out, but I haven't confirmed this in a head-to-head test. But that's why I decided to default to outside and use the 3d printed holder on the interior. To be clear, it's a guess.
- Again, earlier in my tests I did a lot of sealing with tape and foam when I believed that was a problem. Turned out, it was my designs that were bad and the sealing didn't matter. At the moment, I don't think I'm losing a ton of air through the small gaps. However, when I do my close-to-final test, I'll probably try that, too, to see if it makes a difference. Other makers in this space seem to have the same belief -- as long as the design is decent, the small gaps don't seem to matter.
On the monitoring/speed control, I love programmable microcontrollers and I debated on whether or not something like that was useful. For my specific use case of something for classrooms, I've avoided adding anything extra due to cost. What I may consider down the road is LoRaWan for monitoring. My rationale is that for schools, I want to know that it's being used, but I can't trust (or expect) the school to give my purifier wifi access. So, instead, I can just broadcast the stats via LoRaWan and have a nearby parent save and relay the data. So, my only thoughts for microcontrollers are just thinking through your use case and whether you might want to scale it up and just make sure it makes sense.
On the multiple fans, my advice is to mock up different designs with cardboard, fans, filters, and duct tape. And do quick tests with salt particles to see how well it works. It was humbling for me to see good design ideas I had didn't work well in practice. The Corsi box design (and AirFanta 3Pro) works well (to my current understanding) because it's a box with a lot of filter area, somewhat minimized interior space, and all the fans on one side pulling in the same direction. The Tempest 6 fan design (3 fans at right angles) seems to work well because it's a good amount of filter area, minimized interior space, and 6 fans pulling in non-competing directions.
Hope that helps and good luck with your builds.
raccettura
This looks pretty good.
Wasn't sure the best way to provide some feedback, so figured I'd just open an issue... I've been looking into building something similar but perhaps 8 fans and using an esp32 running @esphome to make the fan speed something that could be controlled and monitored via @homeassistant. A few thoughts popped up aside from sizing: