Deep Dive into the Theremin v.2 Audio Amplifier

See YouTube video https://www.youtube.com/watch?v=90T0ZkN-oos&t=3s for further details.

audio_amp_schematic

The Theremin v2 audio amplifier and preview audio amplifier share the same schematic.

The input of the amplifier is on pins P1 and P2, where P2 is the ground connection. The input is supposed to be DC decoupled with a small electrolytic capacitor, as suggested by the data sheet of the integrated circuit TDA2003, which is the main component of this amplifier. However, the capacitor was not put in this schematic because it will find its place directly on the output of the audio generator stage. No reason to have two capacitors in series: one at the output of the audio generator and one at the input of the audio amplifier.

The signal coming from P1 goes to a voltage divider consisting of the resistor R1 and the potentiometer R2. When the potentiometer cursor is all the way toward ground, there will be no signal going to pin 1 of the IC and, therefore, the amplifier will be silent. When the potentiometer cursor is all the way toward resistor R1, the signal coming from P1 is divided exactly in half, since the value of the potentiometer R2 and of resistor R1 are exactly the same. This is the position where the volume will be the maximum possible.

I chose the value of 100K for both R1 and R2 so to have a relatively high impedance toward the audio generator, so its output won’t be affected by the movement of the potentiometer. At the same time, I wanted a value of R2 low enough to keep the input of the amplifier solidly on the ground when there is no input signal.

R2 is a logarithmic potentiometer, which means that its value changes according to a logarithmic scale. This is done to compensate for how we perceive the sound volume: the potentiometer increases the power of the sound wave in the opposite way as our ear perceives it. This way, we feel that the volume increases and decreases linearly when we turn the potentiometer.

The output from pin 4 of the IC goes to the loudspeaker through capacitor C3, which is used to prevent that a DC current goes all the way through. In addition, capacitor C5 and R6 provide a high pass filter that shorts to ground all the high frequencies that are not supposed to go to the loudspeaker. Values of capacitors C3 and C5 and the resistor R6 reflect the suggestion on the data sheet.

data_sheet_schematic

The power supply comes in through pins P5 and P6 and is filtered by capacitors C1 and C2.

C2 takes care of shunting toward ground any unwanted high frequency signal coming from/to the power supply. Capacitor C1, which is a high value electrolytic capacitor, is mostly used to boost the input current when the amplifier requires sudden increases due to peaks of the volume. These capacitors have also the values suggested by the data sheet.

The last part of the amplifier is the negative feedback loop, connected between pins 4 and 2 of the TDA2003. The loop is composed of capacitors C4 and C6, and the resistors R3, R4, and R5.

The negative feedback is used to limit the output power of the amplifier outside the region of frequencies where it is requested, thus preventing auto oscillations of the amplifier, which would otherwise behave like an oscillator.

The values of these components are again those suggested by the data sheet. However, C4 and R5, which correspond to Cx and Rx in the data sheet, are calculated based on the data sheet formulas visible in the above picture, where B is the maximum frequency of the signal that we want to amplify, which I set to 10kHz. I don’t believe we want to hear from the Theremin sounds that are above that frequency.

For further information on this amplifier and for a demonstration on how it works, please refer to my corresponding YouTube video.

Theremin v.1 Pitch Reference Oscillator Analysis

This is my first post of this kind, which I plan to use to give insights on how an electronic circuit is designed. In the future, I plan to post a similar article for every new circuit I design and present on my YouTube channel.

This way, people who are interested in such details can go to this site to get them, thus sparing all the others that don’t care from watching them on the original video.

Today, to get started with the new series of posts, I will go into the insights of the design of the Pitch Reference Oscillator used in the Theremin version 1.

I decided to use the Pitch Reference Oscillator because all the other Theremin’s oscillators are based on the same principle, although for the variable ones, an antenna is added to the resonant circuit to make its capacitance increase based on the position of the hand on the player with respect to the antenna. In fact, the antenna adds up to 8pF to the resonating tank, based on the position of the hand of the player.

For reference, you can watch any of the YouTube videos on my channel related to oscillators in the Theremin Project playlist.

Here is the link to the playlist:

https://www.youtube.com/watch?v=moT9iAaZU-I&list=PL3SNC7XyKklYZOs48Qfh16nTO4t8edDPU

I began the design of the oscillator starting from the base concept of positive feedback amplifier, which can be represented as follows:

feedback_amp

We have an amplifier with a positive gain, which should be 1 when there is an oscillation, and we have a feedback impedance Zf that takes part of the output signal and brings it back to the input of the amplifier. Zf is supposed to be chosen to obtain a positive feedback at the oscillation frequency.

This basic configuration needs now to be modified to insert in it an LC resonating tank. This should be on the output of the amplifier, so its oscillations can be fed back to the input and be sustained with no decay.

oscillator_block_diagram

The amplifier can now be replaced with a real circuit made with a transistor. I choose in this case a common base configuration, which gives me the required gain of almost 1, which is enough to sustain the oscillation.

The Zf will need to provide a phase shift of 90° to compensate for the internal capacitance of the transistor, which will add another 90° for a total of 180°, thus providing a positive feedback, considering that the transistor has actually a negative value of amplification (-1).

Therefore, Zf can be replaced with a simple capacitor that will bring the output signal back to the input with the desired 90° phase shift.

And once all the polarization resistors are added, the circuit looks like the following:

oscillator

In this circuit, R2, R3, R4 and R5 are the polarization resistors, which are calculated based on the specs of the transistor itself.

L1, C4, C6, C7 constitute the resonating tank. The Two variable capacitors are added to allow for a fine tune of the frequency.

C5 is the feedback impedance, or Zf.

The resonating frequency is calculated based on the following formula:

resonating_frequency

where L = L1, and C = C4//C6//C7//CBE = C4 + C6 + C7 + CBE

and CBE is the capacitance between emitter and base.

Temporarily excluding CBE from the calculations, the frequency therefore can vary between 258.2 kHz and 312 kHz. The CBE can be obtained from the transistor data sheet, adjusting for the polarization between base and emitter.

Also, we can slightly affect the resonating frequency of the oscillator by changing the polarization point of the transistor. This can be achieved by a potentiometer added on the base of the transistor itself, as per the following schematic.

final schematic

All the other capacitors in the schematic are used to make the polarization point of the transistor independent from the oscillation frequency, by shunting the oscillation signal toward ground. Thus, their value is such that they behave like conductors with respect to the high frequency of the oscillator.

At the end, considering the increase in capacity given by CBE, the frequency ends up to be centered around 400kHz, which was the required frequency for the pitch reference oscillator.