Introduction: Musicator Jr - Mk 2
A simplified Sound-to-Light display using a few inexpensive parts.
Step 1: The Circuit - Overview
If you're familiar with the original Musicator-Jr, you'll know it uses a dual op-amp to convert sound from an electret microphone to display on LEDs. This new design uses a much-simplified circuit and is able to support up to 40 LEDs.
Like the USB-Musicator design, this circuit makes use of the LM431 "Programmable Zener" circuit. Although packaged like a transistor in a familiar TO92 form, there is a half-dozen transistors inside which forms a powerful op-amp and driver, which we will make use of here.
We have added a 2N4401 transistor as an output stage, which will allow us to drive LEDs up to 600mA. So it is not advisible to substitute any other device for it.
The basic design is for a 9-volt battery, but 12-volt operation is also possible with a few value changes.
A point-to-point diagram is supplied below, a larger / clearer version is also avilable here.
Step 2: The LED Display Arrays
To make the most of available voltage, we'll use the following formulae to work out the optimal LED composition for each chain:
Vs = (V-supply-1) * 0.9, then
Subtract 3 for each White, Blue, Pink or Violet LED
Subtract 2 for each Red, Yellow or Orange LED
repleat until you are as close to 0 as possible.
This design requires that at least two of the LEDs be Red, Yellow or Orange, but can be any 2.
In our sample with 2 reds and a blue and Vs = 7.2; so 7.2 - 2 - 2 - 3 = 0.2, which is very close to optimal.
Additional chains composed of similar LEDs (same number of 2 and 3 -type LEDs) can be attached to points (A) and (B) on the scematic, up to a limit of 400mA, or about 20 chains of 20mA LEDs.
Here's a video of it in action:
Step 3: Parts List
Q1 LM431 Programmable Zener
Q2 2N4401 NPN Driver Transistor
C-in 1uF to 5uF capacitor, + to Electret Mic side
R-bias 10K
R-drive 1.5K
R-feedback 150k to 560k. I used 2 x 470k in parallel to get 235k-ohm. Use 150k for the lowest sensitivity, ie LOUD environments. This is connected to the top of the 2 x RED LEDs.
An additional 0.1uF capacitor can be inserted from the top end of R-feedback to ground to extend the time LEDs stay on, but is entirely optional.
A series of images are provided to guide you in the wiring.
Step 4: Variations
For other voltages, use the formulae presented in step (2) and adjust R-bias to V*1000-ohms, R-drive to V*150-ohms.
Have fun! Visit my website for more ideas on lights and LEDs.
Here's a video of it working:

Participated in the
LED Contest
36 Comments
12 years ago on Introduction
7 years ago
im very interested your idea and i tried to apply it to my project but doesnt work well. im not engineer nor technician ,just lawyer. My project s sound activated laser. I humbly request you to leave email contact to udom_ta@hotmail.com . Thank you for your attention.
Reply 7 years ago
You have a great deal of work to try and modulate a laser like an LED. Lasers can only have an on-or-off state. Any modulation must be done with a combination of pulse-width or frequency.
Reply 7 years ago
i d like to share idea wt u. Could u pls send me email as abow.
8 years ago on Introduction
Hi qs, just want to ask if there's any other component that can replace the LM431 in your ckt. I found it hard to find that piece and I'm so eager to have that kind of masterpiece. Hoping for your reply. Thank you:)
Reply 8 years ago on Introduction
Do you think TL431 or KA431 could do?
9 years ago on Step 2
They took down your video for copyright reasons, that sucks. You can probably argue it since you aren't making money.
Reply 9 years ago on Step 2
It can also be argued that only Beethoven can do that, since he wrote the original!
10 years ago on Introduction
Hi how are you? I built your circuit on my breadboard. It works great with one exception. I can get the middle blue and white lights to light up beautifully. But the red ones hardly visable. I have each row in parallel like you have it. I have everything identical to your circuit. Hmm...
Reply 10 years ago on Introduction
Unfortunately I cannot see enough details in the photos, but since all the LEDs are connected in series, they should be lighting up evenly - it may be that you have some blue/white LEDs going directly to ground. That is incorrect: make sure that EACH white or blue LED is connected to TWO red LEDs before going to Ground.
Reply 10 years ago on Introduction
Thanks for the reply. I'll just try different Red LED's. Thanks again for the great instructable!!! I will be making your "Using AC with LEDs (Part 3) - The BIG light" once I find the BR. Take it easy!
10 years ago on Introduction
Very nice. I'll put this one on my to do list lol. Thanks again :).
Reply 10 years ago on Introduction
You are so very welcome. Happy building!
11 years ago on Introduction
Hello! Great instructable! Can I use a 500 trimmer instead of the 470k (based on the pint-to-point diagram)? Or I must use it in parallel with a 47k resistor?
Reply 11 years ago on Introduction
Hi, thanks for the comment.
If, by 500 you meant a 500K-ohm trimmer then that is fine, nothing else is needed, since you can always adjust the pot to a lower resistance value.
btw, if you had tried to parallel a 47K resistor to it, you would end up with a value around 43K. NOT what you'd have wanted.
11 years ago on Introduction
Interesting circuit. I'm trying to figure out how it works.
It seems the IC is operating as a comparator. It's comparing the IC's internal 2.5V reference against the audio signal's AC component + a DC bias generated from the forward voltage on the 2 lower LEDs.
The voltage of the LEDs should add up to be around 2.5V then. So the performance of this circuit depends on the Vf of the LEDs you use?
Also there's negative feedback that keeps the DC bias at 2.5V on average. So an LED with lower Vf will be driven at higher current even when there is no audio- to maintain 2.5V at the feedback pin. That means the LEDs will be on even when there is no audio...
I also noticed your R-feedback resistor value is specified as a range. Does the value of R-feedback control the frequency response or the amplitude or both?
I've built several sound to light circuits. I typically use an audio amplifier IC to drive a opto-isolator. But I'm interested in getting different responses. I'll have to play around some with this circuit.
Reply 11 years ago on Introduction
For this circuit, you have to think in linear terms.
The overall circuit IS an Op-amp and, with no input, it will try and hold the voltage at the top of the 2 LEDs at 2.5v. This gives each LED about 1.25v. With less than 1mA through them, they're barely lit, if at all. Remember that they are connected in series, so both LED must share the same current, regardless of any differences in Vf.
When a signal arrives from the mic, it changes the equilibrium and the output changes such that Vout = 2.5 - (Vin* ( 1 + (Rfeedback/Rin) )) where Rin is the impedence of the Electret and associative components, usually about 2k-ohm. The range of Rfeedback tailors the amplification to bring the 0.5uV signal of the mic to something visible on the LEDs.
11 years ago on Introduction
Seems like a wonderfully simple circuit, but... I tried building it with 9v, R-feedback=165K and two blue LEDs rated at 3.5V each. The top LED stays fully lit, and the bottom LED (replacing the two lower lights in your diagram) responds to the mic as expected. Extra two-LED chains are also always fully lit, but it seems like the voltage drop should be close enough... Any clues?
Reply 11 years ago on Introduction
The fact that the LEDs do not light up equally means there is a problem with the connections. The top LED is probably shorting somewhere.
Reply 11 years ago on Introduction
Thanks for the quick reply! It ended up being a faulty LED that lights up fine but I guess has a bad lead or draws way too much voltage...