No neutral power supply/relay board for in wall switch
License:
Attribution-ShareAlike (CC-BY-SA)
Created:
5 years ago
Updated:
4 years ago
Views:
12815
6 Collect
0 Comments
Share
220 Download (738.75 kB)
Make a donation to mtiutiu

Intro

This is the power supply board (including relays) for the touch switch plates that I made for Livolo.

The project complements the previous touch switch plates that I made a long time ago providing a full solution for creating a custom in wall RF switch which works with no neutral line.

The project is based on the DER-622 design from Power Integrations which is free to download and study in order to understand how it works (schematic is available also). There's also another one called DER-832 which is more recent and a little bit improved.

And now the why

Why I did this? I wanted to see that it can be done and not impossible to make it work and have a full solution for my home when there's only the live wire present. And if Livolo or other companies did it why shouldn't I? More than that I wanted to learn more about electronics in general (especially designing custom ac/dc power supplies and such).

And now the how

All I did on my side was to customize it a little bit by changing the stand-by power supply with another one based on the LNK364 **chip which is a solution provided also by **Power Integrations. The PCB was made so that it has the same dimensions as the one from the original Livolo switch except the screw terminals which are not aligned/positioned the same way. The pin header is also positioned so that the custom touch plates that I made should fit and uses the same pinout.

**V_SENSE **functionality is there but not used on the custom touch plates as it wasn't planned at that time (the only thing that it does is to signal whenever the switch changes the internal power supply used or when the series MOSFET regulator kicks in).

And yes, it took some effort to finish the custom board as it has to be pretty compact and also to take care of keeping as much as possible the minimum distance between tracks where high voltages are implied. Also for high current paths to make sure that the traces are more wide and mirror on both sides of the PCB (using via stitching).

PCB's arrived and did some final testing. Everything seems to work fine so far.

Note:

  1. Although there is a possibility to use loads that draw several amps I don't advise you to do so because it wasn't quite designed for it. Saying this because of the heat dissipation and pcb traces. Max current that can be drawn should be limited to 1.5 - 2 amps.
  2. The touch sensing board including radio/mcu must not draw more than 8-10mA on average. This can be achieved theoretically via the built-in dc-dc converter of the NRF52xxx IC.
  3. The output capacitor (C3) of the 3.3V LDO must be increased to 220uF (case B tantalum - yes, it can be soldered over the 0805 footprint) if instability is observed. Also a 0.22 - 0.47uF AC rated capacitor (275Vac) must be placed in parallel on one of the light bulbs (can be avoided only if the front board draws 1-2mA max - this can be achieved only if used with a Bluetooth LE project - I'm sorry but these are the laws of physics).

Next step is to design the enclosure which will be 3D printed with PETG. The front side will be a glass plate which I'm thinking now how to make it look and fit better.

** Tested and it works! **

Revisions:

  • 0.1 - initial design
  • 0.2 - fixed/optimized some traces
  • 0.3 - added mounting holes (for M1.4 small screws), updated schematic