Comment by schiffern | Hacker News Reader

Comment by schiffern | Hacker News Reader

Parent Full thread

Very cool project. Thank you for sharing your work.

  > Also... useful to regulate flow, any extra flow in either direction hurts efficiency.

I think this factor is what ultimately tips the balance in favor of a centralized "smart" auto-calibrating unit[0] (using simple and reliable pressure sensors) over distributed ductless. However centralized units do have other advantages.[1][2]

I wouldn't get too discouraged just because you had difficulty selling an early unit. Marketing and sales is hard. Reaching the right people and then effectively communicating a product's value is both an art and a science.

[0] https://www.youtube.com/watch?v=LGB83pSLMiI

[1] https://news.ycombinator.com/item?id=38743885

[2] https://www.youtube.com/watch?v=TChYcSSNXrI

I can actually get small thermistor based hot wire anemometers for about a dollar, so I can just use those with a PID control loop to control fan speed. Then the flow is quite well regulated. However high winds will overwhelm the fan, a problem you don't have so much with centralized units, however that's because the fans are so powerful in centralized units. This implies quite high electrical energy consumption and also noise production. The noise production is not as bad because the fan is in your basement or whatever, but it's still serious and not as good for the most part as a well designed, flow compensated decentralized unit, even for the flow vs noise ratio. I added an automatic valve which only costs about $30, which closes in a storm. So as long as most of the time the wind isn't crazy everything works pretty good with the decentralized approach.

Reply on news.ycombinator.com