I need to put an alarm siren close to the location of one of my audio speakers. But I don't have a dedicated cable towards this siren. Because the alarm control unit is close to my speaker amplifier I want to use the speaker cable for transmitting a trigger signal for the siren as well.

At the alarm-control-unit/amplifier side I can use an audio relay to connect the speaker cable either with the amplifier or with the siren signal based on whether the alarm is armed or not. (When armed, I don't need audio.)

But on the speaker/siren side I also need to drive a relay to switch between siren and speaker connection. I was thinking on using an Arduino to detect a DC signal on the speaker cable for this purpose, via one of its analog input pins.

Any suggestion for the electronics to use, in particular on the speaker/siren side? Or other cheap and simple mechanisms to multiplex the siren trigger signal on the audio signal?

My first thoughts: Reduce the siren signal level via a voltage divider to a lower level that still can be detected by the Arduino but does not cause a "plop" on the speaker. And a diode in series with a zener diode in order to clip the AC audio signal between 0 and 5V, before feeding it to an analog input pin of the Arduino?

  • \$\begingroup\$ what about using above audible frequencies? \$\endgroup\$
    – PlasmaHH
    Jan 26 '16 at 14:29
  • 5
    \$\begingroup\$ Ye gods, pull a wire and be done with it. Captain Overcomplicating to the - oh, it's too late for rescue again? \$\endgroup\$
    – Ecnerwal
    Jan 26 '16 at 14:39
  • \$\begingroup\$ What @Ecnerwal said... what are you trying to save here? Time? Money? Effort? I don't see benefit for any of these. \$\endgroup\$
    – Eugene Sh.
    Jan 26 '16 at 14:41
  • \$\begingroup\$ I am totally in agreement with the first comment. Besides that, the "bright ideas" that you are scheming up here will likely end up having some design or wiring problems whilst you try to make it work. Those mistakes could easily cost you the amplifier, speaker or your Arduino. \$\endgroup\$ Jan 26 '16 at 14:56
  • \$\begingroup\$ Belief me, pulling another wire is simply not an option. That's the reason for this question. \$\endgroup\$ Jan 26 '16 at 23:33

Nutty application, but here goes ...


simulate this circuit – Schematic created using CircuitLab

Figure 1. Carrying DC current on speaker pair.

Figure 1 shows one way of doing this.

  • In normal operation audio is passed through (large) capacitors C1 and C2 to the speaker. L1 and L2 present a high impedance to audio frequencies and prevent the audio getting back or loading V+ and prevent it reaching the relay coil.
  • When SW1 is pressed the DC current rises (hopefully slowly enough to be inaudible) and shows up on the relay coil.
  • D1 freewheels when SW1 is opened and allows the current to drop gradually.

I'm not working out component values for you!

  • \$\begingroup\$ Thanks! Did you intentionally left out the option to use a relay on amp side? I have a signal available from my alarm unit to decouple the amp when the alarm is armed. \$\endgroup\$ Jan 26 '16 at 23:49
  • \$\begingroup\$ I really couldn't figure out what you were doing other than trying to superimpose DC on your audio. Sure, replace SW1 with a relay contact. You could replace my relay with a low-power DC alarm / strobe too. \$\endgroup\$
    – Transistor
    Jan 27 '16 at 0:24
  • \$\begingroup\$ I tried your scheme in falstad.com/circuit. It looks like I need a high value (1H) for L2 to suppress the audio signal sufficiently towards the relay. I used 20V AC source at 65Hz to simulate "worst case" audio amp signal, and a 70 ohm, 71 mA relay. Is this not too high/am I doing something wrong in the reasoning/calculations? \$\endgroup\$ Jan 27 '16 at 13:32
  • \$\begingroup\$ Calculations are probably OK. Check the inductance and resistance of the relay. It may be high enough to act as its own inductor and not load the circuit. \$\endgroup\$
    – Transistor
    Jan 27 '16 at 19:17
  • \$\begingroup\$ I tested with this circuit. It has some modifications: using RC filter instead of large inductance, using relay to switch between sources. I added waveform signals to mimic dynamic switching for simulation purposes. Looks ok now. Thanks for hinting me in this direction! \$\endgroup\$ Jan 29 '16 at 18:10

Although I don't disagree with the "pull another wire, it's easier" comments, I think you could also accomplish your goal by using a dc bias (blocked from all audio sources/sinks by filter caps, of course) on the speaker wire to trigger detector circuit and throw a relay to disconnect the speaker & connect the alarm siren. However, as mentioned above, this seems to be overcomplicating things, and introducing needles danger to your components under most "normal" circumstances, so this post is likely more academic than immediately useful. The exception being if there is significant reason why running another wire isn't a good option (transatlantic cable/wire was poured into a solid concrete wall/etc.).

  • \$\begingroup\$ ok, looks like transistor beat me to the "post" button & included a schematic of a similar, yet simpler, approach. \$\endgroup\$ Jan 26 '16 at 18:53
  • \$\begingroup\$ I would not have raised this question if pulling another wire was an option, obviously. \$\endgroup\$ Jan 29 '16 at 18:12

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.