[Solved] How to modify smart switch into dry contact mode (detach mains from relay)

Good evening, I have this Smart Switch Zigbee version, I hope I'm not off topic, I would like to ask if anyone has tried to modify it to have a clean contact at the output?

What do you mean by clean contact? A relay contact that is separated from mains voltage?

Thanks for the reply.. dry contact sorry, but without adding an external relay..

The following modification is very simple to do but it's hard for me to tell you the exact steps as I don't know the PCB layout of your device. Maybe you can also give us a photo from the above so I can see where the relay is?

The following modification requires cutting the traces. Futhermore, it may be problematic because that kind of devices are using a common zero rail so you might need to do some cutting trace and soldering wire tricks, depending on the board.

I attach the photo above where the components are, it is mirrored to better understand.. in the past I have already modified a similar device, brand Sonoff, following the tutorial, I am able to weld and cut tracks if necessary, thanks for the advice

Can you also include the description of screw terminals of the product?

Again, it's very easy to do when you have a device in front of you, but hard to do remotely.

So far I can see:

It seems that right two terminals (when looking from the bottom), marked also with green arrows, are controlled by relay (it can close and open circuit here) and the pins marked with blue arrows are a relay coil (that is powered by 5V to turn on the relay).

I am unable to see yet where you'd have to cut something. Are you saying that the most right two terminals (or one of them) are connected also somehow to mains input?

This is the description
Sorry for the sketch, I do it with smartphone while I am driving;) however your analysis is correct, I think I have to remove the Resistance for separate N, but then I don't know continue

Please don't play around with phone while driving, IoT "hacking" can wait!

So the relay switch is here (unless I'm much mistaken), it connects L-in to L-out:


Ok, so if you markings are correct, then I'd say that L-in has to go somewhere to the power supply circuit. We know that L-in is here:

The question is - where does the L-in branches to the power supply to make 5V for Relay and later 3.3V for WiFi module?

Well, there are are two options:

if it branches where the red arrow shows, at this joint, on the top of the PCB (top layer), then you can easily cut the trace to disconnect relay from L.

The option two is:

if it branches where I drawn the blue arrow (at the Relay pin, branches at the top layer of the PCB), then hacking this relay is hard, because cutting the trace won't just do the job...

At the moment I am not able to tell more. There are multiple ways to proceed from here. You can:
- just try looking at PCB?
- just give it a try, cut of the trace going from L-in to Relay switch and check if board still powers on (if not, then that's not good, restore the trace, board is not easy to modify)
- just desolder relay with flux and solder wick...

it seems that from the L-in input a power branch goes to RH3, could this be the 5V power supply and then 3.3V?

Sorry but I can't see well on that photo. For me it looks like the trace goes only to RH3 pad and RH3 is not soldered. Or is the trace going somewhere further there?

you're right man, the track stops there and nothing is soldered, tonight I'm unsoldering the relay and I'll try to find where L branches, thanks for your time, I'll keep you posted;)

@quasar87 maybe it would be faster to just cut the track, then you have 50% chance of getting it right, and if the device won't power, you can just reconstruct the trace with several axial resistor leads (so it can take the current)

hello man, then the hack was successful! the power track was hidden under the terminal block,

I cut it, I separated the two neutral pads, and I reconnected the power supply with a wire,

I hope it's not too thin, however I plan to power the switch with a maximum of 9/10V.. Thanks for your support, you've been very helpful

Good job. I think it's also worth to note that it has turned out the L power trace is on the two sides of the board. Manufacturer at least tried to handle higher currents.