I'm trying to build a speaker enclosure with a lot of LEDs on the outside of the case that react to the sound, I've played around with smaller amplifiers (30 watt) and would only get 3.8v at maximum volume from the sub which was great but not enouph from the side speakers (2v-2.5v). With a bigger system I'm getting upwards of 60v from the sub woofer and 30v from the left and right speakers at maximum volume.
I'm not sure how often the sound system would be turned up to it's maximum volume, at half volume I was getting 5v max from the sub and 3.5v from the speakers which was great but I'm looking for something that would be able to scale the power to a maximum of 3.5-4v regardless of what the volume was at.
If you want the LED's to flash to the music use a resistor inline with each LED and wire two reverse parallel in series with each resistor. The resistance value must be able to keep the LED's under there maximum MA rating at peak voltage.
an example of reverse parallel? And is there a way to calculate the resistor needed? I've tried LED calculators but they only have a steady power supply as a power source
figure H is perfect.
You will just have to know the LED's forward voltage drop and the amplifiers peak voltage to figure out the voltage difference and calculate for the resistor value to give you the right maximum current for the LED's.
As ling as you dont get too carried away withthe number of LED's this will work fine.
If you use an opto-coupler in place of one of the LED's you can then use that to drive a mosfet or transistor to flash all the power you want using an independent power source.
I did this back in high school using a triac opto-coupler and power triac from an old light dimmer to make two 250 watt flood lights flash to the music! It works great too!
If you take the opto-coupler concept a little further by putting diodes or lower voltage zeners inline with the resistors to give several different voltage triggering points you can then have different things turn on at different power levels too!
You need a peak detector circuit to drive an LED driver circuit.
Your vision's response to brightness is slow so the peak detector holds a peak turned on long enough for it to be seen. Without the peak detector circuit then LEDs will be just a dim blur.
True.
but the average intensity varies with the power of the music. Still it gives a cool looking and proportional light effect!
The actual on/off speed of the LED's is way to fast for our eyes to respond to but the PWM effect does give it a perceved flare and flicker related to the instant power of the signal going to the speaker.
Try it with the two LED and one resistor setup on any audio source with a peak voltage higher than that of the LED you are using.
A LM386 amplifier IC with a microphone input and the two LED setup will do the same thing with local ambiant sounds. Try it!
Or use an LM386 with a small solar cell with a 10K resistor in parallel for the input and a speaker for the output. You can get the audio signal back out of the light. (And yes it sounds terrible but thats not the point.)
Or just listen to what light sources sound like!
Its very interesting to "hear the sounds" of light that changes at rates far faster than our visual senses can detect.
CDS photo cells work pretty good too, much more compact than a solar cell. The two or three I've yanked out of various devices are smaller than my pinky nail.
Sure. Those would work well. I was just thinking about how I made one of those devices when I was akid. I used a solar sell out of an old solar powered calculator with a 10k resistor across it to get the needed pull down effect to make a crude AC signal for the LM386 input to work right.
A bright sunny day sure sounds wierd! Point it at a large reflective object with heat waves between you and it. Very unusual.
I gotta try that, I've always been curious. It's kind of like sferics (the warbles ticks blips and wines that come from lightning or car/interference signals at audo RF frequencies) Only optical.
I fuigured it was worth mentioning. The OP is trying to turn sound wave forms into light so I figured why not tell him how to turn light back into sound.
Changing one sensory source so it can be detected by another does give some really unusual results some times. Do it with a low IR detector and listen to heat on a hot day!
sounds like a fun project to try one day... and I will give this reverse parallel hook up a try, but in regards to what forumlicker007 said, would it be better to just split the audio input to the amp and a new circuit with say a LM386 chip?
I'm looking at possibly 40-60 LEDs and would hate to loose power in the speakers because of all of the lights
this also isn't 40-60 LEDs on 1 channel, but split between the 3 in a 2.1 system
The problem with having many LEDs and speakers on an amp is not a loss of power.
The amp might blow up if it does not have an over-current protection circuit.
A little LM386 amp can light only a few LEDs if it does not drive a speaker.
I would just try running a few sets at a time and see what your sound quality does.
If you do not notice any change I would not worry about it. Ten sets of LED's with the resistors is still going to be a fairly high combined resistance in comparison to to the speaker.
And typical speaker resistance ratings can be fairly far off from what they say and the amplifier will still not have a problem with it. I have seen many 8 ohm speakers with as little as 5 actual ohms run on 8 ohm rated amplifiers and never have a problem.
10 sets of LED's and resistors are going to probibly have an equivilant resistance of 30-40 ohms or more. Your amplifier will probibly not even know they are there.
If it does give you problems then just use the opto-coupler method and an external power source.