I’ve been trying the Moes Knob for a couple weeks with Zigbee2MQTT. It is replacing a Lutron dimmer. While I will rate the Moes dimmer higher than the Lutron offering, neither are anywhere near the function or form of an old school analog dimmer knob. This is an adventure in frustration!
Directly to your question about ‘smoothness’ of the dimming of the Moes. So far, I would not call it smooth. That said, this is a ‘function’ of the function you have between the rotation messages coming in and you light devices ‘step control/ability’. I think it is possible to smooth control with both, there needs to be a maths function in between the dimming messages and light brightness or color steps. I think it needs to be a non-linear function, both on speed of ‘twist’ and level position (for brightness some type of log function to match our different perception of light levels at the dim and bright ends). I say, I think it is possible, because I tested the Lutron device with a Hue bulb and Hue hub ( am not sure if the dimmer was talking directly to the bulb or via the hub ) and the experience of operation of the Lutron device and light level perception in this scenario was much smoother.
On the physical aspect of the two dimmers, the Moes has slight ‘bumps’ that you can physically feel as you rotate, the Lutron does not. I am not sure which gives the better ‘feel’ and control. Back to the old school dimmer, these were smooth moving. However the data coming out of the Lutron is a nightmare. Both devices send out too much extraneous signal quality and battery level messages during device operation, I think these make interpreting the level movement messages harder. These status messages should only be sent when the device has been idle for a period IMHO. The MOES dimmer message stream is easier to work with from my experience so far.
Back on the physical aspect of both, the Moes ‘sticks’ out from the wall plate about 35% more than the Lutron, I am not sure that is good. And the Moes needs a better (read stronger) connection, it is magnetic and with the device sticking out as much as it does, it is somewhat easy to knock off the wall plate. On the other side, you really have to pull to get the Lutron off (only reason to take them off IMHO is to change battery), and you almost feel like you are going to break the Lutron.
Another problem for both is that there needs to be a better way to distinguish between a click and a turn. It is too easy to ‘click and slightly turn’ both devices when you are just trying to click. Maybe there is a filter of some type that can be applied to the message stream to tune this sensitivity.
While the set of messages and format of them that comes from the Moes is a big improvement over the Lutron. However, the design of the messages that come from both is odd. The Moes has two distinct modes, and generates different set of messages in each. And it odd how they are bucketed. For example you can only get a ‘double click’ message in one of the modes, that function does not seem to overlap with any function in the other mode and yet not available. And to make it more difficult on the ‘significant other acceptance factor’, if you ‘triple click’ the Moes, it flips between the two modes. And now all you messages are different, so you need to check for this mode change or the significant other will be clicking and twisting to no avail. And since the Moes is easy to remove from the wall and makes a very solid ‘throwable weapon’, I can see danger ahead .
And my final comparison between, with a plus for the Lutron and minus for the Moes, they both have LED’s that light for a moment when you click or move. The Lutron has a white LED on the front of the ring that is visible while using. The Moes has a green LED on the back side of the unit that is covered by the magnet plate that holds it to the wall, WTF? End of rant, I am going to get out my fidget spinner now… Good hunting, hope you find a good solution.