This bizarre mini amplifier is a small audio amplifier I built as self exercice with operational amplifiers. It was years since I didn't use this kind of integrated circuits and I wanted to practice with them to remind their characteristics and how do they work. It took me three weeks of my spare time to build it, and now I use it to listen to the music, podcasts or the radio I have on my Android mobile phone, so It is something like a dock station, that also includes a white noise generator.

I named it "mini" because it is not too big and has only 1W of power. And "bizarre" due to it's strange appearance, and because it also includes a noise generator, something unusual on amplifiers. But you may ask why did I put a noise generator on the amplifier. You will find the answer to that question in " the noise generator” section.

A video of the Bizarre Mini Amplifier

The device has these controls:
- a knob to change the output volume.
- a knob to modify the audio tone.
- a switch to turn on or off the noise generator.
- a switch to turn on or off the device ( a red LED brights when it is turned on )





The Bizarre Mini Amplifier
As you can see in the pictures the potentiometers and switches are tagged with embossed labels. I took that idea form a friend that is an artist, Javier Alvarez ( PoderesMentales ) , who uses his Dymo puncher to label the controls of his music instruments.
 
As you can see in the schematics, the circuit is composed of four stages:

- impedance adapt stage: this stage is composed of two OPAMPs ( TL082 ) in follower configuration. One of the operational amplifiers is for the right audio channel, and the other for the left audio channel. This gives to the mini amplifier a high input impedance to allow connecting different kind of audio sources without affecting to their output. The noise signal is also fed to the mixer through an operational amplifier in follower configuration to avoid affecting the noise generation circuit.

- mixer stage: this stage is placed after the impedance adapt stage and mixes the right, the left, and the noise audio signals into one single signal. This is achieved with an OPAMP ( TL082 ) in adder configuration. The “weight” on the output of each of the input signals is set with the input resistors values. The ratio between the value of these resistors and the output resistor sets the gain. The volume potentiometer works as the output resistor of this part of the circuit, and allows controlling the volume of the whole bizarre mini amplifier. This is the only OPAMP configuration in the circuit that inverts the resulting signal, but as all the input signals are affected in the same way this does not affect to the final sound.

- filter stage: this is a simple first order R-C low pass filter which allows removing part of the high frequencies present on the output signal. This part of the circuit is isolated from the rest of the circuit with another OPAMP ( TL082 ) in follower configuration. The resistor of the R-C network is the tone potentiometer and allows to the user to modify the filter cutoff frequency. So this is the tone potentiometer and changing it's value turns the amplifier sound brighter or more opaque.

- power stage: on the theory, ideal OPAMPs have 0 output impedance and can provide infinite current, but that is only on ideal world. Despite real OPAMPs have a very low output impedance and can provide a considerable amount of current for signal applications, they don't have enough power to move the 8 ohm speaker. So I had to use another specialized device in the power stage: a power amplifier integrated circuit ( LM386 ). The configuration of this device is the standard specified in the datasheet, and also added the bass boost improvement R-C network proposed in the same document. I had some noise problems with this IC which I solved placing a 20K resistor between the ‘-’ input and ground.


Schematics of the Bizarre Mini Amplifier

Inside view of the Bizarre Mini Amplifier
 

I found some difficulties during the circuit design.

The first one was that I had to use a single ( unipolar ) power supply and operational amplifiers are typically used with dual power supplies ( bipolar ) , but that doesn’t mean that they can not be used with single power supplies. Single power supplies have +Vcc and Ground terminals, while dual power supplies have +Vcc, Ground and -Vcc terminals. The way to simulate a dual power supply with a single power supply is dividing +Vcc by 2 and using that voltage as a virtual ground terminal. This can be done with a simple resistive voltage divider. Then the real +Vcc terminal is used as virtual +Vcc’ terminal and the real ground terminal as virtual -Vcc’ terminal, so the voltage would be from +Vcc’ (+Vcc ) to -Vcc’ (0), and the virtual ground would be placed at +Vcc/2 . But this approach has to be carefully taken, because that is not a real dual power supply and any current through the voltage divider resistors used to generate that virtual ground may cause fluctuations on that reference voltage and the whole circuit to work in a wrong way. So to avoid this, the right OPAMP configurations have to be chosen. Another option is to use a “rail splitter”, a device that keeps a voltage at the half of the real power supply voltage, and is able to source or sink current without voltage fluctuations.
 

Testing parts of the circuit before soldering.

The simple circuit composed of a decoupling capacitor and a resistive voltage divider that is placed before each of the input stages is used to center the input signal around +Vcc/2 , the virtual ground. There are other versions of this circuit that reduce noise, but the one I use is the simplest. The only disadvantage of that simple circuit is that it the capacitor may introduce a negative pulse into the input source when the device is switched off, or a positive pulse when switched on. The resistor to ground just before the input is used to discharge the capacitor when there is no power on the circuit.

 
The power supply voltage, without the capacitor and with the capacitor .

Another problem I had to face is removing the undesired 100Hz noise caused by the poor quality DC power supply I use with the mini amplifier (100Hz = 50Hz * 2 coming from the full wave rectifier ). During the development I used my laboratory power supply which provides a perfect flat and stable voltage. Everything sound good with it, but when I connected the small power supply I started to hear an awful background buzz. I opened the supply and realized that It was composed by the transformer, a full wave rectifier, and a little electrolitic capacitor that seemed to me too small to smooth the voltage. So I decided to place a bigger low ESR capacitor at the beginning of the circuit to stabilize the power and remove the undesired ripples that leak into the circuit and caused that horrible noise. That worked and that noise disappeared.

 

The Bizarre Mini Amplifier nearly finished
 
Years ago I discovered that some constant noises like the rain sound, or the noise of the air conditioners, relaxed me and helped me to sleep. Some summer nights I switched on the air conditioner ( or the heater on winter ) only to hear that sound and fall asleep faster. Also in some occasions I switched on a computer I had on my room to hear it’s fan noise and hide the disturbing “neighbours sounds” that didn’t let me sleep.

I supposed that the reason why these sounds helped me was because they were constant, and rich on different frequencies and this helped masking other sounds that awoke me while I was trying to sleep.

These sounds are very close to what is called white noise. White noise is a concept that can be extended to any kind of signal, and in acoustics it refers to a sound that contains the same intensity of all audible frequencies. White noise sounds like an untuned radio. Adding a random pattern of frequencies to a preexisting signal makes it unintelligible, and this is the reason why white noise is often used to hide sounds or information into signals. The white noise generated by the fans of my air conditioner, the heater, or the computer, masked other low volume sounds that made me difficult to sleep. I suppose that the evolution has "programed" our brains to be sensible and to be alert of sounds while we are sleeping, maybe because thousands of years ago these sounds could come from the steps of an approaching predator or from any other danger, but we don’t live in caves nowadays.

So I decided to add a white noise generator to the bizarre mini amplifier to check if it also helped me on my rest.


Some comercial noise generators.
After building it, I tested it and worked as I expected. I thought about sharing the idea in the Internet, and first I decided to search on Google to see if somebody had done something similiar before, and got surprised when I discovered that the usage of white noise generators to help people to sleep was really extended. I also got upset by the large variety of commercial noise generators focused on helping people to sleep that can be found on specialized and non specialized shops ( like amazon ) .
 

. . . Label, Program & Audiovisuals Tolaemon 2013. . .