Color organ with optical input.

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
So I’m looking to build a color organ, but I want complete isolation from my amp, without using mics.

Is this even possible?

I’m thinking toslink receiver then, a DAC, then from there I’m lost. (not the color organ itself, just the optical front end)

So the questions…

  1. Can this be done?

  2. Will I need to decode?

  3. How do I combine channels, looking for stereo but would like to have surround as well if possible.

  4. The part numbers if you already know what I will need.
A side note question…what kind of mic should I use during the development stage while I work out the bandpass filters and such? (mics won’t be used in final design)

And I’m sure this is only the start of my questions, so please bear with me, and the bumps.

Thanks in advance...
 

Reloadron

Joined Jan 15, 2015
7,501
About 10 years ago I did some work using what they call a Linear Opto Coupler (LOC) which may work for your application of isolating the amp from the color organ circuit. A Google of Linear Opto Coupler should get you some results like this or this. The Vishay data sheet may be of some help to you. Unfortunately I can't make a specific part number recommendation. Our concern when I did this was to monitor a battery voltage level and have isolation.

As to a microphone to use for test? I would just run the audio Line Out from your PC and generate tones using a program like Audacity as this way you can have exact tone frequencies. This assumes you do not have a function generator at your disposal.

Ron
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
Thanks for that, but I would need to power the chip with the Amp, and that is a no go.

Second suggestion sounds good but I will need to do tests in a place with no PC.
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
Here is what I have so far.
A_Color_Organ.jpg
Important questions:
  1. What is the diode I see in most designs at the output of the filter for?
  2. Do I need a virtual ground for the positive inputs of the amps as seen in most designs?
  3. What OP Amps do I use?
  4. Still need to know what microphone to use during the bread boarding process?
  5. Are the 2 caps combining the left and right channels better than using resistors?
  6. Why are the amps used in inverted mode?
Ok enough for now.
 

Reloadron

Joined Jan 15, 2015
7,501
OK, audio is not my forte and the member who is exceptionally good with audio is AudioGuru if he turns up in this thread but I can answer a few basics. First let's look at a very basic color organ circuit.

Color Organ Basic.png

Why are the amplifiers used in the inverted mode or why is the signal fed into the inverting input?
If we look at the drawing note that all of the non-inverting inputs are tied to A and at the top of the drawing we see how A is derived. A is actually 1/2 of Vcc or about 6.0 volts. The incoming audio is an AC signal going both above and below 0 volts. With a single supply we can't drive the op amp inputs below 0 volts. So we offset the non inverting input so with a 0 volt input to the inverting input the output will be 6 volts and that DC level becomes our baseline. This allows amplifying our audio signal. Following each op amp there is a cap to block the DC component level leaving us the amplified AC audio signal. Immediately following C7, C11 and C16 we have some diodes. D1, D4 and D7 will only allow unidirectional current flow and they will remove the AC signal below about 0.7 volts so we only have the positive peaks driving transistors Q1, Q2 and Q3. Since they are NPN transistors we can't drive their bases below 0 volts and the diodes have eliminated the audio signal negative peaks leaving only the positive peaks to drive the transistors.

As to the audio mixing at the input I would likely use both a resistor and a capacitor similar to Audio Mixer Circuits illustrations. A Google of audio mixer circuits will yield dozens more examples and the why. I would likely use pots for the resistors so you have control over each channel. As to the best suited operational amplifiers? Most of what I recall is ancient and I am not going to suggest a LM741 as there are endless newer better options which can drive rail to rail. Yes, you need a virtual ground if you use a single supply op amp for the reasons I mentioned. Microphone? I have no idea what would best suit your needs.

Hopefully an audio type will come along with some good dope and ideas.

Ron
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
Very good stuff, thank you, and diodes 2, 5, and 8 shunt the signal to actual ground during the negative cycle?


So I can create the virtual ground using an op amp in unity mode, with matching bridge resistors…right?
 

Reloadron

Joined Jan 15, 2015
7,501
Yes, if you look at D2, D5 and D8 they will clip the negative going signal. Sometimes you will see that diode configuration and sometimes just the single diode inline.

Yes, if you look at a few ways of doing it, creating a virtual ground, some just use a voltage divider as in the above example and others use an op amp. Frequently a quad op amp package is used so three of the four available op amps in the package are used for channels as Bass, Mid and Treble leaving one left over in the package so they use it to create a virtual ground. Bottom line is VCC/2 on the non inverting input of the actual op amp used for amplification.

Ron
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
So this is what I have now, I haven’t done anything with the mixer yet.

So I’m not using the transistor as the output driver, so I removed it and the cap & resistor that preceded it, and replaced it with the 3914. Are those changes correct, or do I need to change what I changed?

I removed the cap because I want the display to respond faster because it will be geared to movement rather than brightness, but I will replace it if I need to.

I’m going to assume that I can’t place the grounds for the 3914 at VG so I will need an input setting from 6 volts to 9 volts, do I have the proper bridge setup to get this input range? (I’m only going to use the first 3 to 5 outputs from the 3914 hence the 6 to 9 range)

I guess I could just use the VG to set the lower voltage, and use a separate resistor to set the high, but I want to get it working this way first.

A_Color_Organ_2.jpg
 

Reloadron

Joined Jan 15, 2015
7,501
The 3914 would just use power ground. Everything it sees will be a positive value so you just use your power ground. Not quite sure what effects you want but the LM3914 will give a linear response and you can set the internal voltage reference anywhere between 1.2 and 12 volts if I remember correctly. The LM3915 is the logarithmic version of the LM3914 and commonly used in audio db meters.

You may want to increase your resistors to 2.5K, 2.5K and 5K. I don't see a need for a divider drawing 12 mA. What you have will work but your call.

Ron
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
Yea, the only reason I chose a divider with a total resistance of 1k was because of an example used by someone very well known on this very board.

I just assumed that was because you needed a lower resistance than the internal divider, but I will be sure to start with the values you suggest first when I finally bread board this project.
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
So I have changed the mixer and plugged in some of the values, but not the band pass caps and the front end op amp.

Please have a look and make any suggestions as to different values and a part number for the front end amp (must be thru hole 8 dip)

A_Color_Organ_3.jpg
 

AnalogKid

Joined Aug 1, 2013
10,987
Diodes 2, 5, and 8 work with capacitors C7, 11, and 15 to form voltage doublers. This gets a larger envelope voltage rant into the base resistors, so the three outputs have a wider dynamic range and respond to lower audio signal levels. Diodes 1, 4, and 7 complete the voltage doublers andAM-detect the audio so the voltage going to the output stages represents the total energy in the audio waveform, not the ups and downs of the individual cycles. This makes the lights follow the beat.
I want complete isolation from my amp
Thanks for that, but I would need to power the chip with the Amp, and that is a no go..
Why?

ak
 
Last edited:

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
Why?


Well for one I have no intention of connecting anything other than an optical cable to my amp, this thing has become a little cranky lately, and also because I wouldn’t know how to hook up the thing anyway. I’m sure it wouldn’t be that hard, but I’m on a learning curve anyway at this point.

Point is moot anyway because today I placed an order for the DAC.

As for the other information, thank you, that is valuable info, and I will keep it in mind when I set up the output section. I know that circuit is designed to flash LEDs but I am looking for motion instead of flashing and was getting the idea that the output might have too much lag time.

I’m pretty much looking for the 3914s to react to every little change in the music, then if it is too much reaction, I will need to tone it down a bit.

But when I breadboard this thing I will keep that info in mind if I am not getting the kick I need. Perhaps even start with those components in the circuit.
 

Sensacell

Joined Jun 19, 2012
3,432
What all your block diagrams miss is the "envelope detector" function- an essential idea for a color organ.

Google it.

Your network before the 3914 looks like it wants to be an envelope detector, but it's not arranged correctly.
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
So I need to add back the diode and the cap, and remove the resistor to get this “envelope”.

Most of what I have there is based on a different circuit then the one Reloadron posted.

I’m really confused now.:(
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
I think I’m starting to understand.

The 3914 may respond to the really low frequencies like I want, but not at the higher ones.

So the envelope creates a sine type wave by enclosing the sound pulses, creating a smoother slope than the much more rapid sound waves.

Is that close? I’m not really up on ac and audio.
 

AnalogKid

Joined Aug 1, 2013
10,987
Pretty much by definition, a color organ responds only to the low frequency content in the audio. This is because most people ability to see rapid flashing, or flicker, such as the peaks and valleys (ons and offs) of individual cycles of audio) rolls off quickly above about 20 Hz. This is why television works, retinal persistence.

Note I said low frequency content, not signals. Slower than the cycles in a bass note are the rise and fall times of the energy of a bass note. So to filter off a chunk of bandwidth for a color, like making the high frequencies trigger the red light, you bandpass filter the audio to separate out the highs, then envelope detect and lowpass filter that energy to detect the notes. It might be a high G on the violin but the human wrist can barely make 10 strokes per second, so the lowpass filter after the envelope detector often has the same cutoff freq for all bands.

ak
 

Thread Starter

ElectricSpidey

Joined Dec 2, 2017
2,758
Good explanation.

So I guess I will just have to keep swinging at this until I get it right, and I’m pretty sure I will have to make final changes when I test.

I won’t bug you guys about the values, I will try and work those out myself, but I would like somebody to point out any obvious mistakes, like impedance mismatches when I post the final (I hope) schematic.

And I still need advice on that first amp.

A_Color_Organ_4.jpg
 
Top