Transistor Level Implementation of a Karaoke Machine - Essay Example

?Roberto del Rosario described Karaoke Sing-Along System (Patent No. UM-5269 d 2 June 1983, Patent No. UM-6237 d 14 November 1986), which he invented in 1975, as a handy multi-purpose compact machine that incorporates an amplifier speaker, one or two tape mechanisms, optional tuner or radio and microphone mixer with features to enhance one's voice; its name was derived from the Japanese expression for singing along a famous record with the vocals removed, "karaoke". (Bellis, n.d.) With Karaoke's continually growing popularity over the years?in all aspects of the community worldwide (i.e. home entertainment, business, etc.), the researchers came up with their sort of own version of Karaoke as a transistor application. The project, entitled "Transistor Level Implementation of Karaoke Machine with six-band Graphic Equalizer," aims to develop a transistor-based, Karaoke-type amplifier that is able to run from the mains power supply and consists of the following elements: 1. a 12-V regulated power supply, 2. two inputs: a microphone and a line input, 3. a Common-emitter mixer/preamplifier stage, 4. a six-band graphic equalizer stage, 5. a Common-emitter voltage amplifier stage, 6. a Common-collector power amplifier stage, and, 7. a loudspeaker output; as indicated in the schematic diagram below. Figure 1. Project Schematic Diagram Circuit Design and Operation Power Supply Power supplies, as defined by Howard (1998), are electronic circuits, basically composed of four sections: transformer, rectifier, filter, and regulator, designed such that an input ac signal is converted to dc, at any desired level. Shown below is a block diagram of a basic power supply. Figure 2. Basic Power Supply Block Diagram The input line voltage is either stepped up or stepped down by the transformer, depending on the application; in this case, a step-down transformer, T1 rated at 9.5Vac (10.5Vac on actual testing), was used, giving a peak voltage of 14.84V (Vp = 10.5 x ?2), and allowing the device to run from the mains power supply. In addition to that, the power supply is being isolated by this section from the power line. The rectifier section, specifically a full-wave bridge rectifier D1, then converts the resulting signal, still ac, to a pulsating dc, which is made purer by a simple capacitor (C18) filter section, giving a dc hold capacitor peak voltage of 13.44V (Vp – 2(0.7) = 14.84 – 1.4). This leaves enough voltage overhead for the final section, the 12V-regulator IC1 (LM7812) that maintains output at a constant level of 12V and about 1.3A continuous, regardless of changes in load current and/or input line voltages. This configuration has minimum power loss, and negates the need for a heat sink on IC1. The capacitor C19 removes any spikes from the regulator for a smoother output. Howard (1998) Mixer/Preamplifier When a combination of two or more audio signals is expected in a single output, simply connecting the inputs will result to the degradation of system efficiency and poor overall performance due to impedance mismatches of different signal sources and the amplifier input. Furthermore, the differing signal amplitudes of the sources, too, presents another problem since direct connection may result to higher-amplitude inputs obliterating the weaker inputs, and even worse, damage the sources. By isolating inputs and providing independently variable gains at each of these inputs, an audio mixer eliminates both dilemmas aforementioned, allowing input signals to be blend in the desired ratio. (Gibilisco, 2002) Shown below is a sample circuit of a simple transistor-based two-channel mixer/preamplifier. Figure 3. Transistor-based Two-channel Mixer/Preamplifier In this project, two signals, one from a microphone (J1) and another from a line input (J2) are to be mixed. Potentiometers (R1 and R4) were utilized as volume controls for each channel, adjusting the amount of signal passing from the inputs, from a maximum of the entire signal (Rmicin = R1||R3||RQ1in = 10k||10k||2.3k = 1.58kohm, Rlinein = R4||RQ1in = 10k||2.3k = 1.87kohm) to none (Rmicin = R1||R3 = 10k||10k = 5kohm, Rlinein = R4 = 10kohm), and giving an ac voltage difference of about +/- 40mV to +/- 100mV, peak-to-peak. RQ1in is the input resistance of transistor Q1, calculated from the equation: RQin = 26mV/IB, where IB is the base bias current, which in this case, supplied by R5 and calculated as, IB = (12V – VB)/R5 = (12V – 0.7V)/1Mohm = 11.3uA, giving RQin = 26mV/11.3uA = 2.3kohm. Coupling capacitors C1 and C20 then feed the inputs to the base of Q1, a BC548 transistor, for pre-amplification. A preamplifier is simply an amplifier used to prepare a signal, making it more suitable for the subsequent stages; typically, amplifying the signal, control its volume, maybe modify its input impedance characteristics, and if necessity calls it, determine its path through the stages to follow. (Boylestad and Nashelsky, 2002) Q1, R5 and R6 make up a Common-emitter (CE) amplifier, for the preamplifier stage. R5 biases Q1, which with R6 produces an ac voltage boost of around 200 times (gain of BC548 is listed as 100-300 on the specification sheet, the average value 200 has been used in the calculations), but inverted, since CE amplifier configuration introduces a 180-degree phase shift. Depending on the positions of R1 and R4, the signal is boosted to around 6Vpp. Using 200 as gain (?) of Q1, collector current, IC = ?IB =200 x 11.3uA = 2.26mA, giving a bias voltage across RC, VRc = 2.26mA x 2.2kohm = 4.97V. Output resistance of amplifier stage changes with the amount of current flowing in Q1 collector, having a worst case with Q1 not conducting, RQ1out = R6 = 2.2kohm. With collector at 6V, then resistance of Q1 must equal R6, then RQ1out = 2.2k||2.2k = 1.1kohm. Voltage gain of Q1, AV = -?Rc/ RQ1in = -(200 x 2.2kohm)/2.3kohm ? -191 (negative sign indicates phase reversal). The capacitor C2 couples the signal from the preamplifier into the 6-band graphic equalizer. Graphic Equalizer Figure 4. Universal N-band Graphic Equalizer A graphic equalizer is a device for adjusting the relative loudness of audio signals at various frequencies, which allows tailoring of the amplitude-versus-frequency output by using several independent “tone controls,” each affecting a different part of the audible spectrum. Gibilisco (2002) Shown above is a diagram for a universal n-band graphic equalizer. Input resistance varies with the settings of the potentiometers, and is difficult to work out, since each branch will short at different frequencies. At its highest (each band at minimum and feedback at zero), the effective input resistance is simply the parallel combination of the six potentiometers (R2, R8, R9, R10, R11 and R12), each having a value of 50kohm, giving Rin = 50kohm/6 = 8.3kohm, while the output resistance is six pairs of 33kohm and 22kohm resistors in series, connected in parallel, giving Rout = (33kohm + 22kohm)/6 = 9.2kohm. In this project, the six potentiometers serve as the band controls for a six-band graphic equalizer splitting the signal into six lines, each isolated from the others and have either a low-pass filter, a high-pass filter or a combination of both (band-pass filter), which can be independently varied, making it possible to change the effect that each path has on the signal that reaches the succeeding stage (Voltage Amplifier). Each line only reduces the signal; there is no boost in these filters. By selecting the appropriate values for components in each rung, the six paths, all together, conduct the entire audible spectrum. The project utilized RC filters for the graphic equalizer, where all resistors nominally chosen at 22kohm. Using the formula for RC filters' cutoff frequency fc, or the frequency at which the gain of the filter falls by 3dB, given by fc = 1/2?RC, capacitor values can be solved. (Dorf, 2006) Nominal values below 80 Hz and above 8 kHz were chosen for the lowest and highest filters (1st and 6th rung). In the topmost line, R13 and C10 make a low pass filter with the knee at 80Hz, so frequencies above this are attenuated. With R13 = 22kohm and fc = 80Hz, C10 is solved by rearranging the formula above into C = 1/2?Rfc, giving C10 = 1/(2? x 22kohm x 80Hz) = 0.09uF (0.1uF was chosen). Similarly, for the final filter point at 8kHz, C = 1/(2? x 22kohm x 8kHz) = 0.9nF (0.001uF was chosen). The rest of the span, between 80Hz and 8kHz, was divided into even multi-octave intervals. To achieve this, the step value between the 2nd and 3rd filter points must be double the step value from the 1st to 2nd filter points, and so on; all steps higher must be double the value next lower. Letting the first step value be x, the range must be x + 2x + 4x + 8x. Now solving for x across the span of 80Hz to 8kHz: x + 2x + 4x + 8x = 8kHz – 80Hz 15x = 7.92kHz x = 528Hz With the first filter point equal to 80Hz, the second filter point will be 608Hz (80Hz + 528Hz), the third at 1660Hz (608Hz + (2 x 528Hz) = 1664Hz ? 1660Hz), the fourth at 3770Hz (1660Hz + (4 x 528Hz) = 3772Hz ? 3770Hz), and finally, the fifth filter point preset at 8kHz, such that the first rung passes frequencies below 80Hz, the second between 80Hz and 608Hz, the third between 608Hz and 1660Hz, the fourth between 1660Hz, the fifth between 3770Hz and 8kHz, and lastly, the sixth rung passes those over 8kHz. With the other knee points known, corresponding capacitor values can be solved as follows: For 608Hz: C = 1/(2? x 22kohm x 608Hz) = 0.0112uF (0.01uF was chosen); For 1660Hz: C = 1/(2? x 22kohm x 1660Hz) = 0.0044uF (0.0039uF was chosen), and; For 3770Hz: C = 1/(2? x 22kohm x 3770Hz) = 0.0019uF (0.0022uF was chosen). The signal is recombined by resistors R14, R24, R25, R26, R27 and R38, each having a value of 33kohm, and then coupled by the capacitor C11 into the Common-emitter voltage amplifier stage. Voltage Amplifier Figure 5. Voltage Amplifier Block Diagram As voltage amplifier, as illustrated above and defined by Glading (1998), is simply an amplifier in which the output signal voltage is larger than the input signal voltage. In this project, a Common-emitter configuration was used as voltage amplifier because of its notable voltage gain, as well as, to revert the phase of the signal. The resistors around Q2 (2N3904) were chosen to supply enough current to the final stage, as well as turn on the transistor enough to keep the collector centered around 6V. By choosing an output resistor (R31) of 1kohm, the output resistance of the voltage amplifier is kept smaller than the input resistance of the power amplifier that follows. Bias current of R31 is given by, Ic = 6V/1kohm = 6mA. The base current required, assuming a transistor gain (?) of 200, is given by, IB = 6mA/200 = 30uA. The value of the base bias resistor R28, with base voltage (VB) ideally at 0.7V, can be found as, R28 = (6 – 0.7)V/30uA = 177kohm (180kohm was chosen). Furthermore, the transistor input resistance is found as, RQ2in = 26mV/30uA = 867ohm. Finally, the stage input resistance can be solved as, Ri = R28||RQ2in = 180kohm||867ohm = 863ohm. Since the signal was attenuated greatly in the preceding stage, a high voltage gain is necessary for this section. Voltage gain of Q2 is given by, AV = ?RC/ RQ2in = (200 x 1kohm)/867ohm = 231. However, a gain of around 230 can be too high, establishing the possibility of signal clipping. To prevent this, potentiometer R32 provides variable negative feedback for the voltage amplifier, and is used as the Master Volume control. At this point, the signal has been reverted and is coupled by capacitor C16 into the next stage, the power amplifier. Power Amplifier Figure 6. Power Amplifier Block Diagram As above illustrated, a power amplifier is an amplifier in which the output signal power is greater than the input signal power, usually utilized as the final amplifier which drives the output device (Glading, 1998), in this case, a loudspeaker. A Common-collector configuration, otherwise known as Emitter-follower, was used as power amplifier in this project, which, according to Dorf (2006), despite having less than unity voltage gain (roughly equal to 1), power efficiency can reach as much as 70% when the Q point is placed at ICQ equal to zero, using the majority of the power for the output signal. The resistors R33 and R34 should bias transistor Q4 (BD681) so that it is turned on enough to pull the emitter to around 6V, meaning that in steady state (no input), the transistor is drawing around 270 mA. To bias the output at approximately 6V, a minimal output resistance, R35 was chosen as 22ohm, providing emitter bias current, IE = 6V/22ohm = 273mA, and base bias current given by, IB = IC/? (where IC ? IE , and ? for BD681 listed as 750) = 273mA/750 = 363uA. With base voltage known as VB = (6+0.7)V = 6.7V, resistor values in the voltage divider network (R33 and R34) can be chosen as: (a) low-side divider resistor R34 at 22kohm to be much greater than the previous stage output resistance, having current IL = 6.7V/22kohm = 304uA, which together with the base current, supplies current to the upper resistor IU = 304uA + 363uA = 667uA, and; (b) upper resistor calculated as, R33 = 5.3V/667uA = 7.95kohm (8.2kohm was chosen). Input resistance is given by, Rin = R33||R34||R32||?R35= 8.2kohm¦22kohm¦10kohm¦750 x 22ohm = 3.05kohm, while output resistance is simply Rout = R35 = 22ohm. The capacitor C17 then couples the output of Q4 to the speaker that centers on 0V. When the output of Q4 goes below the steady state point (negative part of input signal), C17 drives the speaker into reverse polarity through R35. This is the limiting drive current for the speaker, and therefore the limiting factor in the power output. Assuming a maximum peak-to-peak voltage of 10V, then the root-mean-square value is found as VRMS = (VPP/2)/?2 = (10/2)/?2 = 3.54V. With average current from Rout given by I = 6V/22ohm = 0.27A, maximum power at the output will be, P = IV = 0.27A x 3.54V = 0.97W. Review of Related Literature Machine de Karaoke Machine de Karaoke (Du, et al) is a karaoke recording machine mainly rooted for entertainment purposes, which is capable of removing the voice component of a music file (from an iPod) and storing the user’s singing voice with the background music to an external compact flash memory, in addition to working as a stand-alone voice recorder. The voice component of the music file is removed by getting the difference between the left and right channels, based on the assumption that the vocalist is in the middle (and most instruments are not) during stereo recording, resulting to equal voice components in both channels, hence the principle of cancellation can be applied. However, this setup is not always true for all music files, singer might not be in the center, and the bass instruments may also be placed in the middle. In addition to that, there is the presence of echoes, vocals cannot be completely removed. The Karaoke machine has six modes, explicitly: Mode 0: Song playback without any modifications, no recording Mode 1: Song playback without vocal in it, no recording Mode 2: Song playback with vocal and singer's voice, and record it into the external memory Mode 3: Song playback without vocal and record it into the external memory Mode 4: Song vocals removal and record user’s singing voice with the background music to an external compact flash memory Mode 5: Record user's voice into compact flash memory, no music Mode 6: Direct reading from memory Figure 7. System Block Diagram As illustrated in the system block diagram above, an external serial multi-channel ADC (a 4-channel MCP3204 chip with a 12-bit serial output resolution) was used to extract the music signal, which provides connection to a Mega32 microcontroller unit (MCU), the system's communications center. Accordingly, the MCU handles the necessary communication between system components (i.e. ADC–MCP3204, compact flash memory–ALFAT, DAC–DAC0832 and LCD–used for current mode display) following necessary protocols. 5-band Graphic Equaliser Ghos (2007) stated that equalizer circuits typically divide the audio spectrum into separate frequency bands having independent gain control, as he discussed the 5-band Graphic Equaliser. The equalizer implementation uses low-cost operational amplifiers (op amps), specifically the LM833, seeing that it meets the noise density, slew rate and gain-bandwidth product requirements. The multiple-feedback band-pass filter topology shown below was used in the spectrum partitioning, and by utilizing the same capacitor values, component values were basically found with the following relationships: Centre frequency, fo = 1/2?Cv(Ra||Rb)Rc Bandwidth, B = 1/?CRc Quality factor, Q = fo/B Gain, A = –Rc/2Ra Figure 8. Multiple-feedback Band-pass Filter Topology To prevent overlapping of adjacent bands appropriate values of Q should be selected. The output of each band is mixed and then fed to an external audio power amplifier. A Q = 1.7 and A = 4 was used in this equalizer as recommended by the National Semiconductor. The circuit uses a non-inverting amplifier with A = 2 as a buffer stage, but before that, the input signal was halved by a resistive network to produce a unity net gain. Metal-film type resistors and polyester type capacitors should be used because of the filters' sensitivity. To prevent the dc component from getting amplified and propagated, each stage of the op amp was coupled to the succeeding stage using 10µF, 16V capacitors for a good low-frequency response. To bypass noise, ground the Vcc of each op amp with a 0.1µF ceramic disk capacitor. The circuit is powered by a 12V DC regulated supply and it was recommended to use a well-regulated supply utilizing 7812. Condenser Mic Audio Amplifier Figure 9. Condenser Mic Audio Amplifier Schematic Diagram Illustrated above is a compact and low-cost implementation of a Condenser Mic Audio Amplifier that utilizes a BEL1895 IC in providing a good-quality audio output of 0.5W at 4.5V, suitable for use in intercoms, walkie-talkies, packet radio receivers, and low-power transmitters, specifically low-power HAM radio transmitters, since it can provide for the necessary audio power for modulation. Transistors T1 and T2, as indicated in the schematic diagram, form the preamplifier of the microphone. The necessary bias for the condenser microphone is provided for by the resistor R1, while its gain is varied and controlled by the preset VR1 functions. The preamplifier stage, however, produces a low-level audio output and to increase the audio power, it is coupled using the capacitor C7 to the audio power amplifier centered about a BEL1895 IC. BEL1895 is a monolithic audio power amplifier IC specifically designed to deliver 1W into 4ohm load at 6V power supply voltage, and used mainly for sensitive AM radio applications. Its low noise and distortion characteristics and voltage requirement at around 3-9V makes it ideal for battery operation. The Condenser Mic Audio Amplifier is also equipped with a turn-on pop reduction circuit to prevent thuds during power supply turn-on, a ripple-rejection filter (C9) and a damping circuit for output oscillations (R15-C13 network) on top of the boot strapping function provided by the capacitor C12. The amplifier's low-frequency response is determined by capacitor C7. References: Bellis, M. (n.d.). Roberto del Rosario. Retrieved May 20, 2011, from http://inventors.about.com/od/filipinoscientists/p/Karaoke.htm Boylestad, R., & Nashelsky, L. (2002). Electronic Devices and Circuit Theory (8th ed.). New Jersey, USA: Pearson Education, Inc. Dorf, R. (2006). The Electrical Engineering Handbook: Electronics, Power Electronics, Optoelectronics, Microwaves, Electromagnetics, and Radar (3rd ed.). USA: Taylor & Francis Group, LLC. Du, S., Lee, J., Cho, J. Y., & Qi, S. (2006). Machine de Karaoke. Retrieved May 20, 2011, from http://instruct1.cit.cornell.edu/courses/ee476/FinalProjects/s2006/sq24/sq24_jl435_jc287_sd285/index.htm Ghos, S. (2007). 5-band Graphic Equaliser. Retrieved May 20, 2011, from http://www.electronicsforu.com/electronicsforu/lab/ad.asp?url=/EFYLinux/circuit/May2007/CI-01_May07.pdf&title=5-Band%20Graphic%20Equaliser Gibilisco, S. (2002). Teach Yourself Electricity and Electronics (3rd ed.). USA: McGraw-Hill, Inc. Glading, K. (1998). Introduction to Amplifiers. In Navy Electricity and Electronics Training Series (Mod. 8). USA: Naval Education And Training Professional Development And Technology Center Howard, R. S. (1998). Introduction to Solid-State Devices and Power Supplies. In Navy Electricity and Electronics Training Series (Mod. 7). USA: Naval Education And Training Professional Development And Technology Center Prabakara, D. (2001). Condenser Mic Audio Amplifier. Retrieved May 20, 2011, from http://www.electronicsforu.com/electronicsforu/lab/ad.asp?url=/efylinux/circuit/feb2003/Condenser%20Mic%20Audio%20Amplifier.pdf&title=Condenser%20Mic%20Audio%20Amplifier Read More