LED Lighting Setup, along with bass-responsive trigger.

Hey guys,

First off, I've read a couple of the "LED music sync" threads on here, and it's also been a few years since my intro-level EE course in college.

I'm looking to do some wiring in my house to compliment my newly-constructed bar and speaker shelves. Essentially, I want to wire up 2 sets of LEDs, one with a constant-power source, and the other to utilize a phono jack from either an MP3 player or computer to trigger the 2nd set while using a low-pass filter to restrict it to bass hits. Also, I'd ideally like there to be a pass-thru for the 3.5mm phono jack so i could send the audio to the amp without using a splitter.

I'd also like this to have 2 pots to use as "dimmers" for each circuit along with switches...

The LED's I have are:

-24 3.5V blue LEDs (preferably 8 parallel branches with 3 LEDs in series on each branch) for the constant-on set
-16 3.3V white LEDs (8 parallel branches with 2 on each) for the bass-response set

I'd like if both could run off the same power supply, but it's the circuitry and calculation work that I need some help with. I've got a pretty solid concept for the end-product, I could just use some help form you all. I'll attach an image of what I'm kind of getting at.

As far as schematics are concerned, here's what I have so far.... Biggest bit of confusion is implementing the low-pass filter so the pulse isn't "overall-volume" controlled, but rather low-end "thump" controlled...

Outside of the low-pass (dunno whether to go active, or a simple RC filter), can anybody provide any insight as to whether or not I'm heading in the right direction?

Thanks in advance for any assistance.

Why is your schematic a negative picture with a black background instead of a normal positive picture with a white background?

Who says your LEDs are 3.5V? Did you measure each one? That is probably their typical voltage. The max voltage might be 3.8V then 3 in series and a 12V supply will barely light them. If they are 3.2V then they will burn out.
Therefore I think you need more voltage across the current-limiting resistor (fewer LEDs in series or a higher supply voltage) so that the different voltages of the LEDs are less significant.

A simple RC circuit is a very poor filter.

You can set your circuit to work well only at one volume setting. If the volume is turned down then the LEDs might not light. If the volume is turned up then the LEDs might be turned on all the time.
Therefore you need an automatic level control circuit.

Schematic is negative because that's the way gEDA (Linux schematic program) output the image. Original is white bkgnd because it was a quick photoshop job.

As far as the LEDs are concerned, they are listed at 3.3V typical, 3.6V maximum. Having recalculated for the 3-in-series set, I get 100ohm resistors...is this right?

Regarding the one-volume-setting, the idea of this was to operate the computer/ipod at a single volume level (non-maxed out) to avoid clipping and distortion, and the volume is controlled via the amplifier/reciever.

As far as the AGC, where in line would I place it? Before the low-pass so it maintains a consistent pulse voltage to the transistor? And for the Low-Pass, should I run a standard op-amp first-order active filter?

How terrible is the schematic? That is...how far off is it from what I'm looking for?

Your negative schematic is very odd.

If you have a regulated 12.0V supply then three 3.3V LEDs in series use 9.9V and a 100 ohm current-limiting resistor will set the current at 21mA which is good.
If the LEDs are 3.6V then the current will be 12mA which is a little low.

Since you set the volume control so that the output is as high as possible without clipping then you can also set the input of your LED circuit with a trimpot so that it works properly. Then AGC is not needed.

A first-order filter is almost useless because its slope is so gradual. Use an active second or third order filter.

Is there any particular way to measure the LED voltages without using some sort of multiple resistor setup??

And by trimpot, is that just another word for potentiometer? Also, what's particularly odd about the schematic? I have 8 "pods" of light spread along the walls of the room, and I'm trying to design it so that if one LED goes out, only that particular "pod" fails, rather than the whole setup.

Other than that, is my general layout "acceptable"? Am i right to be using an NPN transistor to control the voltage to the LEDs, or is there an easier way?

The forward voltage of LEDs can be measured if they are powered from a constant current source that could be an LM317 regulator with a single resistor as shown in its datasheet.

A trimpot is a small potentiometer that is adjusted with a small screwdriver.

Your negative schematic is odd because a positive picture is normal.

If one LED in a series string of them burns out then all of the string will not light. It is extremely rare for an LED to burn out if the current is the max allowed or less. Oh, cheap Chinese LEDs bought on E-Bay frequently fail.

LEDs are fed current, not voltage. The amount of current controls the brightness. The voltage must be high enough for the LEDs to work.

Your transistor is an emitter-follower with no voltage gain (it has a voltage loss). It needs a very high signal level to brightly light the LEDs.
Maybe the transistor should be common-emitter that has almost no loss and has plenty of voltage gain.
You might need a rectifier so that the transistor's emitter-base junction does not have avalanche breakdown (at about 6V).