Giant audio controlled cylinders in a night club

Other Hardware Audio
1

I have some previous experience working with Arduino in my bar where i built sound activated columns with an MSGEQ7 chip. But this idea is a little out of my depth so im coming for advice.

Im going partners on this night club and it has these tacky large cages by the dance floor. Ill include a picture.

Im thinking about making it a cylinder with pieces of plexiglass and lighting up the plexiglass to go with the intensity of the music. Im picturing like 1 foot high pieces of plexiglass wrapped around the cages horizontally with each piece separated with some kind of divider and layering them like that to the ceiling. So it would look like a perfect cylinder with 1' high pieces of plexiglass layered on top of one another to the ceiling.

Now, the question becomes how would i use an arduino to light up a different level of the cylinder based on the intensity of the music? Is that possible?

This youtube video is a similar idea except his are all one color and are square. But its in the ballpark. LM3915 IC VU Meter Tower - YouTube

cage.jpg
cage.jpg1920×1200 191 KB

2

If you want to make this VU meter effect the LM3915 is probably the way to go. It's cheaper, easier, and there's no programming required!

If you want something more interesting and variable you might want to use an Arduino. I made a "Giant VU Meter" effect with the Arduino and it randomly reverses (goes down instead of up) and randomly inverts (the LEDs turn off with loudness instead of turning-on). And, since it's programmable it does some other flashing-sequencing effects. Programmability also means I can automatically calibrate the sensitivity for loud & quiet songs or volume control changes.

Like a lot of projects, there's input, output, and software...

On the input-side, since the Arduino can't accept negative voltages it can't directly accept the negative half of the AC audio waveform. The simplest solution is to use two equal value resistors to bias the input at 2.5V and a capacitor to block the DC bias from your audio circuit as shown [u]here[/u]. You can subtract-out the bias digitally in your sketch.

I use a peak detector circuit. The peak detector puts-out a changing DC voltage proportional to the peak signal level. It puts-out a positive voltage (it ignores the negative half of the audio waveform) so you don't have to worry about biasing the input, but the big advantage is that I can read the "volume" about 10 times per second instead of "sampling" the audio waveform thousands of times per second.

Either of these approaches will accept a line-level (or headphone-level) audio signal.

The output circuit will depend on the LEDs/lights you use. You can drive "regular" LEDs directly, but higher power LEDs/lights will need some kind of driver circuit. My "Giant VU Meter" uses regular "high brightness" LEDs, but I used serially addressed LED driver chips so I could drive/address 48 LEDs (24 per left & right channel).

The LM3915 would also require a driver for anything other than standard LEDs. But, I think it works at a higher voltage than the Arduino, so you may be able to drive several LEDs in series from each output.

The software for a VU meter effect can basically be a series of if-statements... If the ADC reading is greater than 100, turn on the 1st LED, if the signal is greater than 200, turn on the 2nd LED, etc. Except, if you are reading the waveform (not using a peak detector) you'll need to find the peak every 1/10th of a second or so. That's because you are as likely to sample a zero-crossing as positive or negative peak or anything in-between and any single sample tells you nothing about the volume.

3

I looked at the LM3915 and that would be fine if i just wanted it to do one color one thing but for how big these things are going to be, id like for it to change color according to the frequency of the audio at the moment as well. I think just a big blue VU meter might get old. Idk.

Yeah i think id rather go with the arduino where it has variables. That would be really cool.

As far as the input to the MSGEQ7, i think thats sort of what i did with my other project using that chip.

Im thinking driving SMD 5050 LEDs not individually addressable ones and just using different circuits for each level of the cylinder.

What im thinking i need to do is marry these 2 concepts together:

#1 Telling the strip lights to blink color according to frequency:
Video: LEDs and Music (Tutorial Included) - YouTube
Instructable: http://www.instructables.com/id/Blinking-LEDs-to-the-Frequency-of-Musi/?ALLSTEPS

#2 Telling individual strips to light up in those series of IF statements you said:
Video: Vu-meter LED-style. - YouTube
Instructable: http://www.instructables.com/id/Led-VU-Meter-2/?ALLSTEPS

4

You said you used the MSGEQ7 before. I've never used it, but it's a "powerful" chip that takes care of the audio input issues and it relieves your software from having to handle any frequency filtering.

I don't see any issues with combining the two concepts. If you understand what the software is doing your imagination is the only limitation of what you can do with 7-channels of frequency-intensity information.

If you want to find an example, look for spectrum analyzer examples. I was sort-of assuming you had made a spectrum analyzer when you used the MSGEQ7 before. A spectrum analyzer is basically several VU meters with each meter representing the level/intensity of a different frequency band.

I'm not necessarily suggesting that you copy the example, but try to understand how it works and build something you want!

I think just a big blue VU meter might get old.

I agree... Just about any single effect can get boring after one song... That's why I built several options/variations and other effects into my "VU Meter". My VU meter has 7 effects with enough variations of each effect that it can take a long time before I see the exact same thing repeated... Each effect runs for between 30 & 90 seconds (duration is randomly set) until a different effect or different variation is randomly selected.

My VU effect is only one color (red LEDs) and I don't have any frequency detection. The ability to change colors and/or frequency-related effects can keep things even more interesting.

I have made some other lighting effects that use colored flood light, but none of my current effects have any frequency detection... The first lighting effect I built a million years ago was a 3-channel "color organ" (which has one color for each frequency band). When I got bored with that, I built something with more randomness.

P.S.

Im going partners on this night club

This is off topic and none of my business, but don't get too bogged-down with this! This kind of project can be very time consuming... Don't forget to pay attention to what's really important and don't forget to run the club! You can buy lots of lighting effects and you might want to start with some commercial effects while you get the club up and running.

Many commercial lighting effects can be sound activated or DMX controlled. You could start with the built-in sound activation and then as you have the time you can build (or buy) a sound activated DMX controller that you can program do whatever you want.

There might even be an advantage of making your cylinders DMX controlled so you can integrate them with commercial lighting effects.

5

Yeah ive used the MSGEQ7 chip before and it works great. I built these LED columns that are just really amazing looking.

Id love to be able to get it to invert what its doing. I think ill work on that after i can prove the concept can actually function properly.

Yeah i know not to get too bogged down in this. We didnt even sign the lease yet so this is why im starting this right now. Im anticipating a bunch of trial and error and time to plan the construction out as well. Theres a good bit of moving parts involved in getting this done but i have 4-5 months to get it done.

Ya know someone else suggested use DMX controls for the cylinders. They said to look into this: 48 channel DMX dimmer

But i couldnt find any good code for using a DMX protocol so i kinda scrapped the idea because i cant write code.

6

I think i might have a gameplan.

In my example number 1, the audio comes in to the MSGEQ7, out to the arduino, then from the arduino to 3 transistors for red, green, and blue colors.

In example number 2, he uses one transistor as just power and runs that live wire to just the red connection so they only do red. He could have just run it to blue or green instead. Doesnt matter. Then somehow he cuts down the voltage .6V using a 1n914 diode across each line of lights which gives it the equalizer look. I might have to try different diodes out to determine how much voltage i would want dropped over 10 levels.

But what im thinking is taking example number 1 and having 3 transistors for each line, so 30 transistors total. And then from example number 2, figuring out how he is dropping this voltage .6V to each line, tinkering with that until i get it perfect, and then just implementing that intentional voltage loss to each set of 3 transistors, for each line.

Im not sure if this is doable because the transistors hes using in #2 differ from the transistors in #1. Other than that, i dont really see why it wont work, but, wtf do i know lol.

What do you think? Im i out of my mind? Or should i start ordering stuff and giving this a shot?

7

OK i made a schematic to kind of show what im talking about.

You see the blue LED signal come in off the arduino to what would be LED row number one, then the diode drops the voltage by ~.7V and takes it to LED row number 2. Then another diode drops the voltage and takes it to LED row number 3. I would then repeat this for green and red.

Can someone please verify that im on the right track here?

8

I uploaded some videos of the LED columns powered by Arduino and an MSGEQ7 in action if anyone wants to take a look.

9

You see the blue LED signal come in off the arduino to what would be LED row number one, then the diode drops the voltage by ~.7V and takes it to LED row number 2. Then another diode drops the voltage and takes it to LED row number 3. I would then repeat this for green and red.

Can someone please verify that im on the right track here?

Those diodes are the wrong way round.
That is NOT a schematic it is a useless physical layout diagram.

I can't see how that circuit could possible give you any effect. The output from Pin 11 is always either 5V or 0V, it is not an analogue output it is a PWM output.

http://www.thebox.myzen.co.uk/Tutorial/PWM.html

10

Well i built the proof of concept today. It still needs some tinkering but its coming along. For some reason, the lights stay dim when theres no audio playing. Not sure why. Overall though, very excited. Now i just need to take this, tweak it, and adapt it to fit 4 foot wide cylinders.

11

Nice effect, high energy, will look bad a$$ in 4 foot wide cylinders. Makes me wanna dance :smiley:

12

Sweet thanks elac. Thats the plan! :slight_smile:

Im gonna change out some resistors and ill upload a better video with the 13 bars instead of 9.