I'm adding a Google Home Mini to an antique floor radio. I don't want to toss the original components, so I'm planning on wiring the speaker to the Google Mini and want to add lights that come on in the radio's display when sound is played (Mini in use) but not sound-activated. I want a constant solid on while it's playing music or talking and off when it's not. That way it's not flashy and looks like it would if the radio was on.

Is there a way to have a switch or relay detect voltage on the speaker wire and turn on an LED? I was planning on connecting the LED to the 5V power as a main source and not be driven from the voltage on the speaker. I'm also going to see if I can detect any other places that have voltage only while on.

  • \$\begingroup\$ There should only be a voltage on the speaker wires when the speaker is actually producing sound. \$\endgroup\$ Commented Jul 21, 2022 at 19:23
  • \$\begingroup\$ (er.go., since speaker signals are alternating above and below zero volts, this won't work very well for switching anything.) Might be able to build a 555 monostable timer to "keep enabling" on sound signal detection (and keep a relay pulled in, lighting the radio.) Would recommend a solderless breadboard for testing such a design before committing to some kind of more permanent circuit board. \$\endgroup\$
    – rdtsc
    Commented Jul 21, 2022 at 19:59
  • \$\begingroup\$ How old is this radio? Be mindful that old tube based radios can run at VERY high voltages and be a serious safety hazard. The good news is that the old stuff usually had schematics inside the housing and those should list the output voltages of the transformer. \$\endgroup\$
    – Bryan
    Commented Jul 21, 2022 at 20:15
  • \$\begingroup\$ The radio is 1949. And something isn't working inside but don't want to damage anything so plan is it not even plug that in and only plug the google home mini in. \$\endgroup\$ Commented Jul 21, 2022 at 20:21
  • \$\begingroup\$ Those old radios, many are known as "All american 5".. They used 5 tubes with heaters specifically designed to total 110V i.e. each tube is electrically tied to the AC wall power line. (saved cost of transformer) Further, often the metal chassis was electrically live. THESE ARE SUPER DANGEROUS IF YOU DON'T KNOW WHAT YOU'RE DOING. For real, there are voltages in there that will turn your straight hairs curly and your curly hairs straight. To say it can kill you is not rhetoric. \$\endgroup\$
    – Kyle B
    Commented Jul 21, 2022 at 22:57

4 Answers 4


A simple way to implement this would be this:

Sound Actuated LED simulation

I show a 100 Hz AC signal 1 volt peak (0.7 VRMS) applied to the base of a common 2N3904 transistor Q1. When the voltage is high enough to turn it ON, it charges the capacitor C1. After about 0.3 seconds it reaches a steady state where its voltage is high enough to drive the LED with about 7 mA. When the signal is removed, the capacitor discharges through the LED and it remains visible for another 0.3 seconds or so.

How well this works depends on the voltage from the speaker of the Google Home Mini. It must be at least about 0.8 volts peak to turn on Q1. If it is an 8 ohm speaker, 0.7 VRMS corresponds to about 61 mW, which can be quite loud in a quiet room. To work it lower volume, an amplifier would be needed, unless you can find a germanium transistor. :)


You could apply your AC signal to a current buffer. The output of the buffer would be a replica of the signal which you could process without removing power from the original AC signal itself. Pass the replica through a rectifier (or absolute-value circuit) and apply the result to a low-pass filter. If there's a signal, the voltage across the filter will be higher than if there isn't one. Let the output voltage from the filter a comparator. If the average voltage is high enough, the comparator will output a high signal. Apply that signal to an indicator light to show when there is a sufficiently large signal. You can lengthen the time constant of the filter to avoid having the light flash for short-duration signals.

Additional Information in Response to Your Comments

Initial Amplification Subcircuit

In the first stage of the partial schematic drawing above an Audio signal is applied to the non-inverting input of an operational amplifier. The very large voltage gain of the operational amplifier tends to make the output labeled Replica match the Audio input signal.

The next four stages of the circuit are identical. Each uses an operational amplifier in an inverting configuration and multiplies the voltage of its input by a factor of \$-10\$, achieved by my choice of the ratio of the feedback resistor to the input resistor. Plots of Audio, Replica, A1, and A2 appear here:

Outputs from Early Amplification Stages

Note that the plot for A2 is actually shown as a negative of the signal of interest. This shows more clearly that all the signals are identical in shape. They only differ in magnitude and algebraic sign.

Here are the outputs of the later stages in this sequence:

Output from Later Amplification Stages

Here you can see that A3 is 10 times as big as A2. I did not plot the negative of A3 as I did previously for A1. Signal A4 is 10 times as big as A3, but you can see here that the amplification is so great that amplifier U5's outputs have saturated at \$\pm15\text{ V}\$, the same as the voltage sources attached to its power supplies. This is not desirable for audio amplification, although for the purpose of seeing whether there is or is not an audio signal, it does not matter much. (Generally, though, driving amplifiers into saturation is inadvisable, if only because it takes time for the amplifier to get out of that state.) You also can see the absolute value of A4. The explanation of how this signal is created is next.

The additional signal processing is done by the circuit shown in this schematic drawing:

Final Stages of the Circuit

The first stage is constructed from two operational amplifiers. It computes the absolute value of the input. I took the circuit from this link:

Understanding absolute value circuit operation

which, in turn, took it from a Texas Instruments publication.

The next stage uses operational amplifier U8 to filter the result. For low frequencies the capacitor looks like an open circuit, so the operational amplifier simply inverts its input, giving the negative of the absolute value. For high frequencies, the capacitor looks more like a short circuit, and the operational amplifier tends to attenuate them. The overall effect is to give the negative average of the absolute value. By taking the absolute value, negative values of the audio input are made just as important as positive values for determining whether an audio signal is present or not. The average value of the absolute value will be non-zero if there is any audio power present.

The final stage applies this negative average to the inverting input of comparator U9. This device yields a high output if this input is lower than the \$-2~\text{V}\$ applied to the non-inverting input. Otherwise it yields a low value. The overall effect is to set the On_Off signal high if the Audio signal input has enough power in it—is strong enough—and to set it low otherwise. The results of the last few stages of the circuit are shown in these plots:

Plots Showing Audio Input and Detection of a Signal

The filter output is more negative for stronger audio signals. When it is negative enough, the on_off signal goes high. While I don't show how this could be used to turn on a light here, suffice it to say that you could use it to turn on a transistor that would let current flow through, say, and LED to show that there is an audio signal present. The comparator can output negative output voltages. For the purpose of turning on an indicator light, this isn't necessary. Replacing the voltage at the comparator's negative input with a zero voltage (ground signal) would eliminate this trait, which might be objectionable in whatever follow-on circuit is used to power the indicator light.

I set the \$-2\text{ V}\$ comparator threshold pretty arbitrarily. Note that it is low enough to detect signals in the \$[0,10]\text{ ms}\$, \$[30,40]\text{ ms}\$, and \$[50,60]\text{ ms}\$ intervals, but not in the \$[20,30]\text{ ms}\$ interval, where there clearly is a weak audio signal present, and certainly not in the \$[10,20]\text{ ms}\$ and \$[40,50]\text{ ms}\$ intervals, where there is no audio signal.

One note about the audio signals. I used the LTspice simulator to synthesize them artificially. They include sums of sinusoids at 100 Hz, 1 kHz, and 10 kHz, all within the typical audible range of human beings, which runs from around 50 Hz to about 20 kHz.

  • \$\begingroup\$ Hmm I've seem to have stumbled upon the part of the internet where I'm no sure what any. Of that means :) I'll have to look up basically everything you said. I do appreciate it though! I might have better odds probing the google home for places to tap into. Unfortunately it's all smd \$\endgroup\$ Commented Jul 21, 2022 at 20:24

I believe the Google Mini has lights that turn on when it is active. If it does, you could use a light sensing circuit to detect that and turn on a light in the radio. Something using a photo-diode or photo-transistor should work, I don't know if an LDR type would be sensitive or fast enough.

Look for something like 'light activated relay module'.

  • \$\begingroup\$ The lights only come on when it talks. If I play music on it then the lights are off. Was going to rewire those lights so the cabinet display has moving lights when it talks. But I also want a constant light when it's on and not just talking or listening to commands \$\endgroup\$ Commented Jul 21, 2022 at 21:12
  • \$\begingroup\$ Do a search for "Sound Actuated Switch". I found several products on Amazon as well as various circuits. \$\endgroup\$
    – PStechPaul
    Commented Jul 21, 2022 at 21:32

So after realizing I only have basic soldering and electrical skills I'm opting to add an inline power switch to the google and led light. I will mount it in the existing hole on the radio and it will kill power to the google and led completely. This way the radio will kind of act functional in appearance. I may also convert the capacitive touch controls back to mechanical and see if I can use a rotary switch for volume. I appreciate everyone's input and time.


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.