But in short: It’s a rotary dimmer switch to be installed in the wall alongside other installation and fits in a standard 55mm system (like Gira System 55). It’s just a controller so it works with smart lights like hue (and deconz, ZHA, zigbee2mqtt) but doesn’t actually dim or switch a load. I have one prototype running via deconz.
I designed it to be powered by mains so it can act as a router to extend network range but also because I hate changing batteries But creating a battery-powered version shouldn’t be too hard.
Currently I am testing the new ESP C6 chip as a replacement for the TI chip hoping to make the device a bit smaller, cheaper, and simpler (probably mostly software wise).
I am also still looking for alternatives but there are very few and they either don’t fit inside the 55mm system (most dimmers from aliexpress) or are very expensive (this seems to be the closest solution: https://www.tlg.no/de/de/slc-smartone-zigbee-4in1-rotary-remote)
I’m looking forward to your feedback and if anyone else could use such a device or maybe would be interested in helping with further development
Yes, that would be great. In any case I am trying to keep it as open as possible. Maybe even add a USB port for easy flashing of custom firmware like @Tropaion did here: DIY ZigBee DLMS SmartMeter Reader
I haven’t. My main motivation was that I really like the rotary control. It allows to quickly change the brightness (or also color temperature in my case) but also precise control when desired. Works a lot better for me than the four-button hue dimmers I have all over my home.
That said, I’m open to collaborating on different designs as well I had this things as somewhat modular in mind anyway.
How does the dimmer you describe work? Is that a rocker that can be pushed top/bottom and returns to the shown position when let go?
This is exactly what I have been looking for! I might be too much of a hobby engineer to be able to productively contribute, but a big thumbs up for your efforts.
Im also trying to get a diy zigbee enddevice up and running with the TI Code Composer Studio. But i have problems with the setup. Do you share your code?
I have managed to build a prototype for a new dimmer based on the ESP32-C6 chip. I am looking for feedback on the hardware design before having a few prototypes manufactured. Help is much appreciated.
Here are the details that I also posted to the ESPHome discord.
I am trying to design a custom Zigbee dimmer switch based on the ESP32-C6-Mini. I am more of a software guy and still learning a lot on electronics, so I hope some people here can give me valuable feedback on my first design. Here it is:
The gist is: I want to build a dimmer switch with an rotary encoder and switch that fits in my 55mm Gira System 55. As I have Hue bulbs everywhere, I want the dimmer switch to just be a remote (so not actually dimming any load), but still 230V powered.
I attached the schematics of my main board and the very simple power supply board that will supply 5V to the main board from 230V input.
I chose 5V so the board can be powered by the 230V board, via USB for development, or possibly in the future 6V from a coin cell. I also chose 5V to power the WS2812B_5050 LED.
My main question is if there something obviously wrong with my design. I already switched the MCP1700T-5002E/MB and MCP1700T-3302E/MB LDOs I initially used because I read that the provided 250mA might not be enough for the C6. Not sure if the parts I chose are a good choice, I read that the battery can be easily drained in sleep mode with the wrong LDO.
I tried to fit most components on the back of the main board (the encoder needs to be on the back for space reasons/the button being on the front) so I can have it printed and assembled by JLCPCB and only solder a few buttons/LEDs on the front.
The JST connectors are to connect additional external switches.
Nice to see that I’m not the only one with this need.
However, for a single-channel solution, I found a commercial device that fits the bill: Schneider Exxact WDE002961
It only seems to be available in the nordics and baltics for some reason, but you should be able to source it from any online retailer who has it in stock.
I use it already, together with HA, and it works like a charm!
It has a drawback though, and that is that it only has one control channel, or to put it differently; it can only control one zigbee light. I have a constrained space, and need to control up to four different lights. So I came up with this design, which is already proven in 3D-print + laser cut acrylic.
The idea is that the transparent ring under the knob will pass on light from a semi-circle of leds that indicate the intensity of the currently selected light source.
Switching between sources is done via the four buttons that make up the 55mm bezel around the knob.
It may seem like a mechanically complicated structure, but I have already made a proof-of-concept using a standard rotary encoder, two 3D-printed parts (the knob is one part, of course, and the bezel is actually also only a single part) and a laser-cut ring from 5mm acrylic. I also ordered a water-jet cut steel sheet for the mounting frame, and that was probably the most expensive part, but still within reason using JLC’s fabrication services.
The only missing part were the electronics, but thanks to your GitHub project, now I have a decent starting point.
You may wonder why would anyone go through all of this hassle, when there is a HA involved already? Well, I live in a rental flat, and if the day comes that I move out, I’d prefer if the various light-controllers kept working in stand-alone mode, should I decide to not leave my HA behind
First, and most importantly, It differs by me not being aware of the Philips solution
Secondly, and perhaps equally important; the Philips Tap Dial is battery powered and I’m of the same opinion as Daniel, that I prefer my controllers to be mains powered. This may not be an issue to some, but it is an issue for me.
As I wrote above, this has to be a 100% stand-alone, tech-support-free implementation with a user-friendliness and reliability on par with a classic A/C light rocker switch. Quite the challenge, right?
But creating a battery-powered version shouldn’t be too hard.
