Idiots guide to getting an esp32 to control AC Infinity Cloudline EC fans esphome/HA

lemo-uguu said:

I started in Fritzing too and felt its limits fast. KiCad has a steeper learning curve but it’s completely free, rock-solid, and way more powerful for real-world PCB work. Once you get comfortable with schematics, footprints, and the 3D viewer, you’ll never look back. And the community library support keeps growing, so you rarely have to build a part from scratch.

Click to expand...

I needed a footprint for something and it wasn't in the library. I started Googling and found a lot of things about, "DO NOT ASK US TO DESIGN FOOTPRINTS!" and realized that a lot of people must have expected to get footprints designed. I looked up info on that and found it was really easy to do it. Once you get used to KiCad, it makes sense. So many times technical programs are so tough to get anything done on them unless you have spent years using it, but KiCad's UI is well designed.

lemo-uguu said:

Yes, you can absolutely hack a D1 Mini to bit-bang ~12–14 kHz using a timer interrupt—it’ll spin the fan fine for bench tests or a single-fan breadboard prototype. In my testing, I scoped the stock AC Infinity controller’s D– line and saw right around 13–14 kHz, so I matched that with a hardware timer on the ESP8266.

Click to expand...

So you are doing well with testing your device hooked up to the fan? That's always an exciting and satisfying step - when you get your designs to just "drop in" to another system and to work well with them!

lemo-uguu said:

• Heads-up: at that rate you’re servicing an ISR ~24 000×/s, so Wi-Fi and HA MQTT can stutter under load. It works for a prototype, but for Rev 2 I’m moving to an ESP32-C3 so I can offload the PWM to the LEDC hardware PWM peripheral—no CPU hit and rock-solid timing.

Click to expand...

I'm not sure, but I *think* that ESPHome uses ledc for PWM, so wouldn't that be doing the same thing?

lemo-uguu said:

Running a small OLED (SSD1306) over I2C on an ESP8266 is doable (plenty of examples in ESPHome), but if you want a full-color TFT + touch you’ll want the extra RAM and SPI channels of the ESP32. I’d prototype on the D1 Mini for the PWM & tach logic, then switch to a C3 or C2 dev module when you add the UI layer, but since you're already using an ESP32, I wouldn't worry about going backwards to an ESP8266. It's just what I had lying around and I've got a ton of them. lol

Click to expand...

I don't need full color - B&W would work. Of course, color is nice. The issue I'm having is that I need a good scope, since my cheap one doesn't always show waveforms like it should. But, hey, it was a $45 o-scope! Considering that, when I was learning basic electronics, scopes were those big things on carts, like the old Tektronics ones, I'm just amazed there can be ANY kind of o-scope that cheap AND small!

Anyway, there are other ESP32 projects that are better at working with screens than ESPHome - but they don't output PWM that I can verify is good and useable with my fans (without a good scope). So I can either use a good scope and adjust the frequency in the source code and recompile whatever project I use, or I can use ESPHome and deal with the pain it'll take to get it working with a good touch screen I can get easily. A side note on this is also that a lot of screens take a number of pins and I still need 4 GPIO pins I can use to control the fans.

lemo-uguu said:

I've done extensive searching but unable to find anything to just 'plugs in' to their EC fans and allows control via ESPHome / Home Assistant. You’re right to be cautious. I’d brand it “Universal EC Duct Fan Controller” or something generic, and in the fine print note “compatible with AC Infinity Cloudline.” Avoid using their name in your product title or logo to stay under the trademark radar. Worst case, they may send a Cease & Desist, but keeping things generic and noting “for 10 V PWM EC fans” usually keeps you in the clear. Plus does AC Infinity really want to spend the resources and lawyers for a $20-30k market? lol

Click to expand...

I think we're in sync with those issues. I haven't seen anything, but I suspect it's because nobody wants to market their device for a lot of reasons. (Especially liability.) In my case, my device works, but I also would not market it without adding a touch screen. Beyond that, though, I cobbled it together. I'd have to pay an electrical engineer to review the design and verify I didn't do anything stupid and that it will work as I think it does before I'd market it.

And if I did, I would be doing just what you said. Advertise it as a generalized device, but include that it works with AC Infinity fans of certain models. (And I'd find other fans to find out which other ones it could work with - I think that's increase the market for it. It'd be nice to market it and say it works for half a dozen to a dozen fans. (That would be after testing it with them all, of course.) I'm sure it'd be possible to include a few things like DIP switches or jumpers so people could configure it for different fans. (For instance, other fans may use a different PWM frequency. I can think of several ways to use DIP switches to set the frequency.)