Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Homemade music lighting system with controller board(s)

Status
Not open for further replies.

keukenrol

Newbie level 4
Newbie level 4
Joined
Feb 26, 2019
Messages
5
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
67
Hi

the past few years I have been experimenting with different LED setups to make them reactive to music.
In order to complete my project I would like to implement the following things:
- A 16 x 32 LED matrix spectrum analyser (5V)
- A 2m WS2812B LED strip (5V)
- Stereo 2 x 8 LED VU meter
- Different R, G, B and W LED's across a wooden panel
- A RGBW LED strip (12V)
- A white LED strip (24V)

The goal would be to make a wooden panel of about 1.5m in the shape of a half circle. Mounting the white LED's on the outside to be used as strobes.
I have a few questions:
- Best power supply to use for the different systems working on different voltages
- Single controller option or every system with its dedicated controller, connected to one audio input
-> For example, my spectrum analyser is currently running a STM32F746G microcontroller. This one is calculating the DFT and is multiplexing the LED's -> HIGH Speed is required, unfortunately the ADC broke down due to a loose power supply cable hitting the ADC input (MAX 3.3V, power supply was 5V).
- Best ADC for stereo operations, I am currently using a DC offset signal because most controllers dont like negative voltages.
- Controllers to use? I have experience with nRF52 and STM32.

Idea's are welcome, the goal would be to have a system which can run on its own on music, or to be manually controlled (control panel should be made too).

Best regards

Tim
 

Oof, that's a lot of LEDs, including the white ones too! Off topic question, but have you considered controlling the brightness of the LED using PWM via your microcontroller? Perhaps with the lower brightness, your power draw would be a lot more lesser than a LED at full brightness. With that many LEDs, I believe there would be a significant difference in low brightness vs full brightness in terms of the circuit load.

Although I think that a 30V/5A (or 7A, 10A depending on the power draw) metal-case AC to DC switching power supply (with individual voltage regulators for each system that requires its own voltage level) would suit well for your scenario, I would wait for a second opinion from a much experienced user since I can't really guesstimate how much power your entire circuit will be drawing due to not knowing which LEDs will be constantly on, which will be on/off semi-frequently and so on (along with your multiplexing method, as in you're going to multiplex each and every single LED, or multiplex one LED "strip" and go to the next, do the same thing and repeat, or?).

Is it possible if you show us all the systems that will be utilized in your circuitry for the microcontroller part? Depending on the kind of systems you're implementing, they may or may not work with a "central" microcontroller.

Sorry, I can't answer the ADC question.

Controllers to use for what purpose, exactly?
 

Without knowing how much current each device draws, it’s impossible to answer your question.
 

Without knowing how much current each device draws, it’s impossible to answer your question.

It will be a raw estimate as I will be making a full block schematic tomorrow. I would like to have enough reserve to swap some things out if they dont fit or add more LED's.
At the moment we are looking at:
2M 60LED's/m WS2812B -> They draw about 50mA each with RGB all on. 60 * 2 * 50mA = 6A (all white, will not happen, I have white's for that). I am currently using a 3A 5V power supply and it works great, perfect brightness.
Then the spectrum analyser, measured about 2A max.
Then, lets say the RGBWW ledstrip and the whites, I should somewhat reach a total of 15A.
My estimation would be that I need between 100 and 150 Watt power supply.
 

Allright, this is the current list:
2M WS2812B LED strip
IRFZ44N MOSFET's, to switch the 12V for the 2m RGBWW LED strip. (Strobes and monochromatic)
5V/12V Dual power supply (5V for the controllers, ADC and WS2812B, 12V for the other ledstrip).
PCM1801U as ADC, 16 bit, 48 kSPS, communication happens over I2S.

The individual LED's are quite easy to add, a transistor driven by a controller should do the trick.

1 Controller is responsible for multiplexing the LED's of my 16 x 32 matrix spectrum analyser and showing the audio spectrum.
1 Controller is dedicated to audio analysis, direct input from the ADC to do spectrum analysis. This controller needs to be able to communicate over I2S. I would like to use the data not only for the spectrum analyser, but also for bass detection. For example when the low frequencies pass a certain value, they make certain LED's go on to light up on the rythm.
1 Controller for the input panel and showing manual effects on the RGBWW LED strip. This controller also drives the WS2812B LED strip.

I have split the complete task up in 3 controllers mainly due to speed. I have used a STM32F7 before with DFT and noticed that 20hz refresh rate for the spectrum analyser combined with multiplexing the display was barely possible. Since I would like to make this with the least amount of money possible, I think I have to switch to lower speed controllers and divide the tasks.
 

Stereo 2 x 8 LED VU meter

You may be familiar with the 3914 IC (dot/bar display driver). Ten output pins activate ten led's (or ten transistors which each control a high current led).
3915 has logarithmic response, to indicate audio decibels.
3916 is designed for VU meter response. Suitable option is to drive red-yellow-green lights.

I made a volume meter simply by feeding unaltered audio to the input pin. Got entire ten-level range on a couple volts amplitude. Negative voltage did not seem to hurt.
 
Allright, I was unaware of that chip series! I will have to re-solder them as they are currently in a matrix, but this eleminates the need for a controller for sure!
 

A bargraph can also be made from op amps arranged as a window comparator. The more led's in your display, the more impact it has.

For the fun of experimentation this simulation has 12 led's, driven by 12 buffer gates.

bargraph 12 led 12 inv-gates 5V signal grn-yel-red VU.png

Outputs change state as input crosses supply/2. The resistor stack needs proper volt levels at each end, carefully adjusted so the bargraph responds to an incoming audio signal. Inputs respond to a narrow voltage range. This is apparent along the entire resistor stack, as seen by waveforms Out1 and Out2 at the extremes of the stack.

Outputs can also drive mosfets or transistors, which turn on high current lamps.

Stacks can be cascaded. Logic gates are cheap compared to 3914 IC's.
 
A bargraph can also be made from op amps arranged as a window comparator. The more led's in your display, the more impact it has.

For the fun of experimentation this simulation has 12 led's, driven by 12 buffer gates.

View attachment 155379

Outputs change state as input crosses supply/2. The resistor stack needs proper volt levels at each end, carefully adjusted so the bargraph responds to an incoming audio signal. Inputs respond to a narrow voltage range. This is apparent along the entire resistor stack, as seen by waveforms Out1 and Out2 at the extremes of the stack.

Outputs can also drive mosfets or transistors, which turn on high current lamps.

Stacks can be cascaded. Logic gates are cheap compared to 3914 IC's.

Cool! The advantage of using this is that is is directly coupled too, no sampling delay or anything.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top