I know how to install and set up a “no neutral switch”. What I’m asking here is how these switches work internally. They need power to run the LED and zigbee radio and they have to work if the switch relay is open or closed for a long time.
So if I were to smash one of these switches and look at the circuit board, trace the connections, what would I find?
My guess is that this switch is never really closed but only in a very high impedance state so that some current always flows and then maybe there is a transformer to “steal” a few microamps.
We can speculate about how it COULD work, but does anyone really know? I’d like to find a schematic.
It treats the load as a resistor back to the line source, and pulls a small amount thru. Hopefully not enough to make the light shine. When the bulb load and the switch aren’t good enough matches, you add a bypass cap or resistor to leak a little more around the bulb.
But the question is the details. How to effecently convert to leakage current when the bulb is off and the full current when on to the required 3.3 volts DC. I can think of a few complex and expensive ways but the entire switch sells for $13. So there must be an inexpensive trick.
This is where engineers earn their money. Anyone can find expensive ineffcent solutions, minimim cost reliable solutions are harder.
It’s what Troon said. As you’ve observed, it’s not trivial to do efficiently though.
Perhaps this will satisfy your curiosity.
Note that the smart switch itself still requires a neutral. It’s just that you don’t need a neutral return from the load.
Think of it this way: normally there will be two circuits involved, especially with smart switches with a dry contact relay. One circuit with a live and neutral powers the switch, while the load has its own live and neutral and goes via the relay. If you remove the neutral at the load, you still need to complete the circuit, so you need to get to the smart switch’s neutral instead. That’s why the smart switch will now be inline and won’t have a dry contact relay. Now you should see why there will be this always-on stray power through the load: it’s needed to keep the smart switch powered.
Yes, we all know WHAT these swiches do. The question is HOW,
Even the linked artical does nt answer: It says (quote) The Difficulty To Create No-Neutral Switches. The true difficulty of creating a no-neutral switch is in minimizing that wattage required in its OFF mode. This is a problem facing manufacturers, who have been trying to perfect this technology for years. (end quote)
No much of an answer, they only say it is a hard problem.
I guess that I will have to sacrefice a switch and smash the plastic case and look inside.
Look at the diagrams in the article and my explanation. It tells you how it’s wired. There’s no other trickery. It’s high school circuitry – that’s how it’s wired. The extra tricks needed isn’t to get this working in principle: It’s to improve power consumption, or flicker, or a dimly lit light – and those things will depend on the specific device. Open one up if it fancies you.
Sir Goodenough already pointed out there could be a bypass (Shelly does it this way), like a cap or resistor, to let more of the current flow through there instead of the light. That’s one way, but it’s not necessary to get it working in principle. The trade-off is between having the lowest possible power to keep the smart switch’s electronics on without turning the light on.
What you’re quoting is an optimisation problem – not the principle to get it working.
I have this exact issue. I have a single Aqara no-neutral wall switch and a single hue bulb. The Aqara switch only works when the hue bulb is on, even when the Aqara is in decoupled mode.
If I turn the hue bulb off (via the bulb not switch) the wall aquara flickers and eventually turns off. This cuts power to the hue bulb and means I am no longer able to use the bulb or the switch.
What resistor or capacitor should I be using to allow enough current to pass to keep the switch on?
And it seems like working, lights are on, the first relay is clicking one or two times, but then the switch turns off completely, and to get it to work, I need to break the circuit and connect it again.
As I understand, the capacitor might not draw enough current, or the system equalises itself as soon as the capacitor is loaded.
Is there a better workaround to get this switch working without an actual load?
Or is this a dead-end, and I should just use a neutral-required switch in such a setup, to make it work as a virtual switch?
Use a device made for that purpose (load) like the ones I linked above, not some random device.
Also I would think if you put a device as an actual load there, so when it turns on it stays on, that would help.
