1
\$\begingroup\$

Refactoring the question as requested:

I'm doing a small home automation project. The circuit I'm trying to control using an ESP8166 is this:

Two 3-way switches controlling a lamp

LOAD: 220VAC / 1A

To be able to control the circuit using the ESP and the mechanic switches, I need to plug a 4-way switch in the middle. Something like this:

Two 3-way switches + One 4-way switch controlling a lamp

My idea is to use a latching DPDT relay with crossover NO/NC, like this:

Two 3-way switches + mimicking a 4-way

So, in the end, I'll have something like this: (minimal schematic, missing protection circuits, external source for relay, relay in's, ..)

schematic

simulate this circuit – Schematic created using CircuitLab

My main question is: Is this correct ?

My second question is: Is it possible/better to use a mosfet+bridge rectifier instead of the relay since 1. a mosfet switches faster, 2. the relay wears-off faster

\$\endgroup\$
  • \$\begingroup\$ Instead of trying to describe your circuit(s) with words, just draw us a diagram. If you edit your question, there's a button which will let you do just that. \$\endgroup\$ – brhans Mar 21 '17 at 11:55
  • \$\begingroup\$ Say what you are functionally trying to do rather than describing possible circuit solutions. \$\endgroup\$ – Misunderstood Mar 21 '17 at 12:06
0
\$\begingroup\$

Decent sealed relays are typically guaranteed to withstand something like 100K cycles with a resistive load, which means they can switch on and off every hour and still have a 10+ years lifetime. Unless you need to switch your load repeatedly each minute or more often, I'd advise you to stick with relays.

You also say that MOSFETs switch faster than relays, which is true. However, I'm having a hard time picturing a load for which a delay of 50ms or less would be problematic.

Finally, MOSFETs + bridge rectifier will work with an incandescent lamp, but more sophisticated loads may require proper AC waveform to function correctly. Commutation of a sinusoidal signal with MOSFETs will be much more complicated than with relays (you'll probably end up rectifying AC to DC, doing commutation in DC and install inverters at each load which needs AC).

\$\endgroup\$
  • \$\begingroup\$ The delay could matter if by some chance I got a signal from the ESP and someone switched one of the wall-switches at the same time. The lower the delay the less % of something wrong happening. Nonetheless, I agree with you. Its probably negligible. I'm sticking with relays then. Latching relays would be more apropriate to save power right ? \$\endgroup\$ – nip Mar 21 '17 at 13:03
  • \$\begingroup\$ Also, to know whether the lamp is on/off I was thinking of using an optocoupler. Do you agree this is the best/a good solution ? \$\endgroup\$ – nip Mar 21 '17 at 13:05
  • \$\begingroup\$ @nip A human pushing the wall-switch will still have a delay of about 300ms between an event and a response, even if you use MOSFETs. Regarding optocouplers, there are many ways of using these to detect mains voltage / current, both good and bad. For lamps specifically, using photoresistors or photodiodes instead of optocouplers seems like a sensible choice. \$\endgroup\$ – Dmitry Grigoryev Mar 21 '17 at 13:18
  • 1
    \$\begingroup\$ The issue with photoresistors/photodiodes is that I'd have to get them near the lamp because I have other light sources and sun light. So this is not really an option. I was thinking either voltage sensor, current sensor or opto. From what I understand, the opto is the best solution. \$\endgroup\$ – nip Mar 21 '17 at 13:45
  • 1
    \$\begingroup\$ Regarding power consuption I realize it's not really much but since I've to buy the relays anyway, I'll decide based on the price. \$\endgroup\$ – nip Mar 21 '17 at 13:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.