I want to power some LED on a passive loudspeaker. That means getting the energy from the incoming, amplified, audio signal.
Of course I understand the LED won't be lit if no signal has been input in the speaker yet of from some time. But I'd expect to be able to store some energy (how?) to smooth out the input voltage variations.
How can I get a stable DC voltage from un unsable AC?
How can I store some energy and smooth out the input variations and the possible absence of signal for some seconds/minutes?
How to make this without impacting the input impedance?
What's the amplifier power, and are you "using" the power?
Higher power makes this easier but it means you also have to protect the LED.
Impedance isn't an issue since you always have a current-limiting resistor in series with an LED, and that resistor is much higher than 4 or 8 Ohms.
Here are a couple of circuits you can try -
The regular diode is a rectifier so the capacitor can charge-up to a positive voltage and it prevents the capacitor from discharging back into the amplifier... The capacitor stores energy and the diode insures that ALL of its energy is discharged into the LED.
The 220-Ohm resistor is the usual current-limiting resistor for the LED, and it also limits current into the capacitor. Otherwise the capacitor would "short out" the amplifier.
A larger (more uF) capacitor will hold the LED on longer, and you might need more than 1000uF. But a larger capacitor also takes more time to charge-up.
With the "low power" circuit, LED brightness WILL vary with loudness.
In the "high power" circuit the Zener diode "clamps" the voltage to 5V maximum. This holds a more-constant brightness and it protects the LED. The 100-Ohm resistor protects the amplifier and the zener.
Yes, the power will be used to feed a coaxial speaker with a crossover filter. So, I suppose I connect the circuit in parallel with the crossover filter?
I don't know the exact power at the moment because it might change according to some listening tests but it will be solid as it should move a 30cm woofer. Therefore, I guess I risk to have too much power rather than not enough.
Should the diode at the entrance of the circuit be rated for the power of the amplifier or can a simple diode function? Which model would you recommend?
Audio power is not a useful value for powering LEDs. The LED voltage must reach at least 2V, that's the minimum audio voltage. Also asymmetric current through the non-linear LED module will affect audio quality.
You didn't say, what is so terrible about that. By the way, it seems like it's been a few days. What have you tried? Did you ever just hook up some LEDs (with resistors of course)? Listen to the sound?
What exact specifications would qualify it as adequate? What was inadequate about the circuits in reply #4?
Your question sits in an awkward place because an experimenter will say, "try it" and an engineer will say maybe, "don't do it" or "use a separate circuit". Because it is a passive circuit as you describe it, there are automatic limitations that no magic solution exists for.