DUALR3 (non-Lite) – shutter moves on its own with long cable run. Need tips to stop “ghost” commands

Hi everyone!

I’m trying to tame a rather stubborn setup and would love some feedback.

Hardware

  • Controller: Sonoff DUALR3 (full version, not Lite) – firmware 1.6.1

  • Mode: Motor (Shutter) – calibrated, works fine via app

  • Motor: 230 V tubular shutter motor (standard up/down leads)

  • Wiring:
    • Push-buttons → S1/S2 → about 7 m of in-wall cable → DUALR3
    • DUALR3 → motor (L-Up / L-Down), about 10 cm cable

Issue
After a random amount of time (seconds to hours) the shutter rolls up or down all by itself. Logs in eWeLink app show real “button” events, so S1/S2 are definitely seeing a pulse.

I have tried:

  • Swapped this DUALR3 with another from a different room → the problem stays in this location, so the issue is the line, not the device.

Sonoff support also confirmed the DUALR3 does not have galvanic isolation on the S1/S2 inputs, so a long cable can act like an antenna or somehow create interferences that signal S1/S2. The manufacturer requires an isolated relay for this motor – my mistake for not checking earlier.

Questions for the community

  1. Has anyone successfully run S1/S2 over more than 5 m without false triggers? Any shielding or ferrite tips?
  2. Is there a hidden parameter (Tasmota style) to extend debounce?
  3. If I keep the DUALR3:
  • Would an opto-isolated interface module between S1/S2 and the wall switch solve it?
  • Any recommended EMI filters or MOVs you’ve used with shutter motors?
  1. If I replace the DUALR3 for this location, which Sonoff, Shelly, or Finder module with dry-contact relays would you recommend for a 230 V shutter? (Power and energy metering would be a plus but not relevant for me)

I can post pictures of the PCB and wiring if that helps.
Big thanks in advance – any trick or pointer is welcome!

Hello, bringing the topic back up in case anyone has any advise.

Thanks!

I have for my venetian blinds about 8 m between dual R3 and wall switch and work with no problem

I used to do research into cardiac electrophysiology. We did all our experiments in faraday cages and spent ages dealing with noise from earth loops. I would just either swap the cable for a grounded one or create one. Guessing wrapping the cable with tinfoil and earthing one end will do the trick.

My electrician was also sceptical of the hypothesis of the length of the cable being too long. But it was sonoff costumer service to bring up this possibility first, and it was validated when I swapped the two relés with each other…

I can’t think of any other reason either

Thanks for sharing your experience!

Covering the cable for 6+ meters within the wall is not an option, as there is no room… Instead, I wonder if having a cable or more diameter would work?
Are there cables built with some sort of shield? Or even better, would it make sense to add one resistance or diode of some sort in series with the wall switch?

Thanks again for your input!

There are loads of shielded cables out there. The higher category Ethernet cables have to have it for example. Some USB C cables. Pretty sure you can buy mains cables with it too, but it tends to just be foil or braiding and costs more, so I’d be tempted to wrap a normal wire in foil and then gaffer tape over it to make it easy to handle.

If you want to go the complicated route have you attached an oscilloscope to the switch output? Sounds like you could probably just filter off the noise with a capacitor, but I’m not sure it’s worth it when you could just replace the cable or bodge some shielded cable. Maybe just check with Sonoff the switched current is low and use an old bit of Ethernet cable you have lying around?