Problem with Adorne 3way

[jenfil82]

Hello everyone, 
I'm at a loss, I purchased 2 Legrand Adorne switches ( https://www.legrand.us/wiring-devices/designer-switches-and-outlets/light-switches/adorne-touch-switch-white-with-microban/p/asth1532w2?srsltid=AfmBOooW8rJi7ce8bUfpvnucVNgCv8ZcegiteMPctICH54zcuvWn7TF0 ) and wired them in a 3way setup.
When I use them without the Shelly, everything works perfectly, but with the common wire of the switch going to the SW of the Shelly, the switches don't control the light anymore.

My gut feeling is that the common wire always has some phantom current on it, sometimes 14v, sometimes around 60v, depending on the traveler wire. So I m wondering if the SW contact doesn't sense that voltage as a signal to turn lights on and messes everything up.
What I tried:
- swap traveler wires, doesn't change anything 
- removed the embedded locator lights, no difference 
- connected the common wire to the I contact instead of the SW. Here the Shelly behaves strange and I can't turn it off or the lights don't work anymore

I saw another post with what looked like a similar issue where the recommendation was to add a resistor between the L and the SW, but I'm not sure if this is the same.

Open to any suggestions...

[thgoebel]

This seems to be a multi-cause issue:

(a) Obviously, orientation light is incorporated, and

(b) it is an electronic appliance without neutral wire.

Thus, connecting with a Shelly actuator is difficult but not at all impossible. To evaluate this, I would need such a Legrand switch at my workbench. And we would need some more information about the scenario (classical 3 way switch? In which switch-box shall the Shelly be mounted? Is neutral wire available in this box?).

[arisoft]

43 minutes ago, jenfil82 said:

My gut feeling is that the common wire always has some phantom current on it, sometimes 14v, sometimes around 60v, depending on the traveler wire. So I m wondering if the SW contact doesn't sense that voltage as a signal to turn lights on and messes everything up.

SW input senses current sourced from the SW towards the Live to activate the input. It can also sink current for protection reasons. I have seen that it can keep a relay activated if the relay coil is installed parallel to the SW input, when trying to sense the status of the relay. Using an additional relay between an unknown circuit and the shelly is the easiest way to make it work properly without any side effects.

Edited April 20 by arisoft

[thgoebel]

2 minutes ago, arisoft said:

Using an additional relay between an unknown circuit and the shelly is the easiest way to make it work properly without any side effects.

Fully agree with this workaround, despite the wiring preconditions are still unknown.

[jenfil82]

Alright, I did some additional trial and error by getting rid of the 3way and just hooking up 1 switch. Same issue and I narrowed it down: the switch outputs 120v when on and...75v when off. When playing with the SW contact I can see the Shelly works correctly. Attached a video of the switch turned off, and the light going on and off by inserting and removing the wire on the SW contact.

I attached a pic of the switch as well, there is a neutral wire on it.

Would an additional relay be the right approach? Or is there another workaround I haven't thought off yet by connecting L, I and SW differently

[arisoft]

27 minutes ago, jenfil82 said:

Same issue and I narrowed it down: the switch outputs 120v when on and...75v when off. When playing with the SW contact I can see the Shelly works correctly.

The manual has diagram where the SW input is connected by a switch to Live. A relay can offer a such switch and makes the Shelly happy.

SW input is not activated by AC voltage at all. It is activated by changing resistance between SW and L contacts. That is why using almost anything else fails with some unforeseen effect.

[jenfil82]

49 minutes ago, arisoft said:

The manual has diagram where the SW input is connected by a switch to Live. A relay can offer a such switch and makes the Shelly happy.

SW input is not activated by AC voltage at all. It is activated by changing resistance between SW and L contacts. That is why using almost anything else fails with some unforeseen effect.

Would you mind educating me a bit more on this? Would this be using option A of what thgoebel has explained here? If so, how would this be achieved? (his option B seems easier to implement in my mind for the moment)

 

[arisoft]

25 minutes ago, jenfil82 said:

Would you mind educating me a bit more on this? Would this be using option A of what thgoebel has explained here? If so, how would this be achieved? (his option B seems easier to implement in my mind for the moment)

Option B is a hack. It misleads the Shelly circuitry in a smart way. But it is not any more forgiving than the original is, if used in a wrong way. It does not damage the Shelly if you try it first. If there is no success, then you can use the relay option.

[thgoebel]

The resistor in the SW input circuit (the „hack“) disables the ability of the input circuitry to detect live potential. Thus, the input reacts on neutral potential only. This  inverts the switching logic, which can be amended by setting „reverse input“ in the configuration menu.

Edited April 20 by thgoebel

[jenfil82]

17 minutes ago, arisoft said:

Option B is a hack. It misleads the Shelly circuitry in a smart way. But it is not any more forgiving than the original is, if used in a wrong way. It does not damage the Shelly if you try it first. If there is no success, then you can use the relay option.

So would a relay like this do the trick:

https://a.co/d/h7qzEAg

and wire like in this attachment? Don't pay too much attention on how the switches' travelers are wired, that is what Legrand is asking to do and it works perfectly

[thgoebel]

Circuit diagram is correct and should work!

A solution with resistor and capacitor to maintain the function of orientation light is described here:

https://shelly-forum.com/thread/28200-schalter-taster-mit-orientierungslicht/
 

Edited April 20 by thgoebel

[jenfil82]

1 minute ago, thgoebel said:

Circuit diagram is correct and should work!

And would your "hack" work as well in my case? I like yours a bit better as it will take way less room in my electrical box, a relay will take quite a bit of room

[thgoebel]

As I mentioned above, I’ve no deeper knowledge of the Legrand switch. To guarantee the proper operation with the resistor or the resistor/capacitor combination I would have to test the circuit. Unfortunately, the Legrand switch is not as cheap as it would allem to buy one for a test scenario.

Thus, if you need a diagram which is working with a very high likelihood, you should go for the relay solution. Otherwise I could send you a resistor/capacitor combo for a test…

[jenfil82]

51 minutes ago, thgoebel said:

Circuit diagram is correct and should work!

A solution with resistor and capacitor to maintain the function of orientation light is described here:

https://shelly-forum.com/thread/28200-schalter-taster-mit-orientierungslicht/
 

So like this with your solution?

[thgoebel]

Exactly!

[jenfil82]

Can I buy one of those resistor/capacitor combo from you? I'm not sure which rating I would need for the capacitor and it looks like you used 2.

I'm going to try both solutions, the relay and yours.

[thgoebel]

I‘ll send you gladly a resistor/capacitor combo! Please drop me a PM with your address. I’ll keep you informed about the costs and the shipping. You’ll get this part:

Ready for installation…

 

[wooly]

About touch switch: an interesting addition to shelly line, rather than the display panels or the TRV, that are source of troubles for everyone, could be a touch switch, similar to this one, that is one of the options on the esp32, having two touch input and two relay (actually a 2PM with 2 additional touch inputs), or just two touch input and the same input for the UNI (to connect a real wired thermometer). this device could have also a 4 digit led display that can be controlled as an output (component: paneldisplay) . Useful is used as a thermostat.

Edited April 21 by wooly

[jenfil82]

So good news, with the relay it worked flawlessly, see attached video

Then just for giggles, I tried to combine a Touch switch and a Dimmer, with the dimmer linked to the Shelly and the Touch on the Hot side. Well, the Dimmer doesn't have that residual voltage and really outputs 0 volts when lights are off. So that combo works without any modifications...

 

[wooly]

@tvbshelly why that strange face when proposing a shelly that has two touch input (an ESP32 may have up to 6) plus two relay output or the four other inputs of an UNI+ ?

[tvbshelly]

@wooly The smiley is intended to express confusion. I found your mixing of TRV, wall display and then touch button a bit confusing. Nothing more, did not intend to annoy you 😉 

[wooly]

5 hours ago, tvbshelly said:

smiley

for me TRV and wall display, from where i see reading the post are thwo defices that, not being "true" shelly are the source of a lot of discontent [given also the huge price they have]. The touch button on the other way, if it were a true shelly, would be an useful addon. (not the BLE button, but a true gen 3 device ! it might have a bigger energy consumption, but much less trouble and much more power)

[jenfil82]

On 4/21/2025 at 2:01 PM, wooly said:

About touch switch: an interesting addition to shelly line, rather than the display panels or the TRV, that are source of troubles for everyone, could be a touch switch, similar to this one, that is one of the options on the esp32, having two touch input and two relay (actually a 2PM with 2 additional touch inputs), or just two touch input and the same input for the UNI (to connect a real wired thermometer). this device could have also a 4 digit led display that can be controlled as an output (component: paneldisplay) . Useful is used as a thermostat.

Would you have a link to the esp32 with a touch function?

[wooly]

4 hours ago, jenfil82 said:

link to the esp32 with a touch function

https://docs.espressif.com/projects/esp-idf/en/stable/esp32/api-reference/peripherals/rmt.html

 

[tvbshelly]

42 minutes ago, wooly said:

https://docs.espressif.com/projects/esp-idf/en/stable/esp32/api-reference/peripherals/rmt.html

 

Quote

The RMT (Remote Control Transceiver) peripheral was designed to act as an infrared transceiver. However, due to the flexibility of its data format, RMT can be extended to a versatile and general-purpose transceiver, transmitting or receiving many other types of signals. 

What the heck is this about touch?