3
\$\begingroup\$

How can I make a capacitive wall touch switch? remembering that the wall standard light switch has just two wires for close the circuit. How can I supply my circuit when the light is on in order to turn it off again?.

PD: I have designed a capacitive touch switch circuit, but it needs an power supply, a wall outlet for example, but is useless in the case exposed before.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
5
  • \$\begingroup\$ I'm sorry, I don't understand your question – could you maybe make a drawing or so to explain your problem? \$\endgroup\$ Commented Jan 15, 2017 at 21:54
  • \$\begingroup\$ I'm trying to make a capacitive wall touch switch like this one: img.banggood.com/images/upload/2014/09/SKU168755.jpg @MarcusMüller, how would you do it?. Consider just two wires. \$\endgroup\$
    – whady
    Commented Jan 15, 2017 at 21:55
  • \$\begingroup\$ Make a drawing. Two wires from where to where? Doing what? edit your question to include the drawing. \$\endgroup\$ Commented Jan 15, 2017 at 21:57
  • \$\begingroup\$ Do you see it now? @MarcusMüller, sorry for the poor info. \$\endgroup\$
    – whady
    Commented Jan 15, 2017 at 22:07
  • \$\begingroup\$ you need a flip flop T type, D /2 or anything that toggles from an edge detected envelope of stray noise or C contact closure or coupling. > it can even be a non-contact line noise envelope detect , wave of the hand. So C switch detect> Schmitt trigger> Toggle function> latched driver (Compartmentalize the design process) \$\endgroup\$
    – D.A.S.
    Commented Jan 16, 2017 at 0:21

2 Answers 2

1
\$\begingroup\$

There are three possible solutions I can think of, assuming you want a switch that works with all kinds of connected electrical devices and not only light bulbs.

  1. Use (at least) a third wire for power supply, if possible.

  2. Use a battery. This has the drawback that the switch might not work if the battery is empty.

  3. Do something fancy and recharge the battery while the switch is on. A slight voltage drop is usually ok also for other devices than light bulbs (see below), but it ensures that not current flows if the switch is turned off. If you want to be on the safe side, use induction*, but that is not something I would suggest for a consumer product without hesitation.

* There is a corresponding paper called "Energy Harvesting from Electromagnetic Energy Radiating from AC Power Lines" by Vikram Gupta, Arvind Kandhalu and Ragunathan (Raj) Rajkumar (https://users.ece.cmu.edu/~vikramg/docs/hotemnets11.pdf).

If you are restricted to classical light bulbs or something similar, you can exploit the fact that if a few mA flow through a light bulb it will not light up. The filament works as a simple conductor. So you have to design a power supply circuitry that has a high voltage drop, but consumes only a few mA.

Then when the switch is turned on you have to switch to a power supply circuitry that has a low voltage drop so there is enough left for the light bulb. For classical light bulbs this will result in a non-perceivable drop in brightness.

Instead of a steady current you can also exploit the fact that very short pulses (either on or off) will also not be perceived, but can be sufficient to power your circuit.

As written above, this is only possible because of the way a light bulb works. If you connect for example a device with a switched power supply you might get really strange effects.

\$\endgroup\$
10
  • \$\begingroup\$ what is the third wire you are refering to? @koalo \$\endgroup\$
    – whady
    Commented Jan 15, 2017 at 22:12
  • \$\begingroup\$ @whady a power supply wire, I presume. \$\endgroup\$ Commented Jan 15, 2017 at 22:12
  • \$\begingroup\$ Yes, of course. \$\endgroup\$
    – koalo
    Commented Jan 15, 2017 at 22:13
  • \$\begingroup\$ notice that you could also build a wall sensor that actually only has two wires used for power supply, and communicates with the actual switching device wirelessly. That would have a number of benefits like making the place of the wall sensor independent from the position of the lamp. \$\endgroup\$ Commented Jan 15, 2017 at 22:15
  • \$\begingroup\$ this is the circuit of the livolo switch, the one that I'd post in the comments. we.easyelectronics.ru/uploads/images/00/30/26/2014/02/28/…, the red part is the power supply of the circuit when the lights are off, and the green part is the power supply when the lights are on. @koalo \$\endgroup\$
    – whady
    Commented Jan 15, 2017 at 22:15
0
\$\begingroup\$

You certainly could create some voltage subtracted from the supply voltage as they did in the circuit you commented on. However, reducing the supply voltage does potentially impact your load especially in the 110/120 V range.

One way to power your circuit when the light is on is to sense the load current with a CT. Since you probably only require a few 10's mW to power your touch sensor this might provide a reasonable solution that would not impact the line voltage to the load. It would also work with a broad range of load currents.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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