Design of Current Mode Based Oscillator for Class D Audio Power Amplifiers : 6 Steps - healeywimen1958
Founding: Design of Current Fashion Based Oscillator for Class D Audio Power Amplifiers
In recent years, Class D audio power amplifiers have become the preferred result for portable audio systems such American Samoa MP3 and mobile phones attributable their high efficiency and low power economic consumption. The oscillator is an grave part of the class D audio amplifier. The oscillator has an serious regulate on the amplifier's sound quality, chip efficiency, electromagnetic interference and other indicators. To this end, this report designs a rife-controlled oscillator circle for class D power amplifiers. The module is based on the current mode and mainly implements deuce functions: unrivalled is to provide a triangular wave signalize whose amplitude is progressive to the power supply voltage; the other is to provide a square waving signal whose oftenness is almost main of the mightiness supply potential dro, and the duty ratio of the guileless wave signal is 50%.
Step 1: Current Mode Oscillator Principle
The working principle of the oscillator is to control the charging and discharging of the capacitor away the current source done the MOS switch tube to generate a triangular wave signalise. A block plot of a conventional current mode based oscillator is shown in Fancy 1.
Design of Rife Mode Based Oscillator for Class D Audio Power Amplifiers
In FIG. 1, R1, R2, R3, and R4 render threshold voltages VH, VL and a book of fact voltage Vref by dividing a voltage of a power supply electric potential. The reference voltage is then passed through an LDO structure of amplifiers OPA and MN1 to generate a book of fact topical Iref that is proportional to the supply voltage. So there are:
MP1, MP2, and MP3 therein system can bod a mirror current source to return charging current IB1. The mirror current source composed of MP1, MP2, MN2, and MN3 generates a exonerate on-line IB2. It is assumed that MP1, MP2, and MP3 have equal width to duration ratios, and MN2 and MN3 have isothermal breadth to length ratios. And then there are:
When the oscillator is working, during the charging phase t1, CLK=1, the MP3 tube charges the capacitor with a constant current IB1. Later on that, the potential dro at point A rises linearly. When the voltage at maneuver A is greater than VH, the electric potential at the output of cmp1 is turned to zero. The logic control module is mainly composed of RS riffle-flops. When the output of cmp1 is 0, the output terminal CLK is inverted to a low level, and CLK is a full level. The oscillator enters the discharge phase t2, at which point the capacitor C begins to discharge at a constant current IB2, causation the voltage at point A to drop. When the voltage drops below VL, the output potential difference of cmp2 becomes zero. The RS flip-bust flips, CLK goes high, and CLK goes low, completing a period of charge and discharge. Since IB1 and IB2 are equal, the charging and discharging times of the capacitor are equal. The rising edge slope of the A-point trilateral wave is up to the absolute value of the falling edge slope. Therefore, the CLK signaling is a square curl signal with a duty ratio of 50%.
The output frequency of this oscillator is independent of the supply voltage, and the amplitude of the triangular wave is graduated to the add voltage.
Step 2: Oscillator Circuit Implementation
The oscillator circuit design designed in this newspaper publisher is shown in Figure 2. The circuit is divided into trey parts: a threshold potential difference generating circuit, a charging and discharging on-going generating electric circuit and a logic control circuit.
Design of Live Way Based Oscillator for Class D Audio Force Amplifiers Figure 2 oscillator implementation circuit
2.1 Threshold voltage generation unit
The threshold voltage generating portion may be entrenched away MN1 and quadruplet electromotive force dividing resistors R1, R2, R3 and R4 having equal resistance values. The MOS transistor MN1 is here used as a switching transistor. When no sound signal is stimulation, the break away sets the CTRL concluding low, VH and VL are both 0V, and the oscillator stops on the job to reduce the static power consumption of the come off. When at that place is a signal stimulus, CTRL is inferior, VH=3Vdd/4, VL=Vdd/4. Due to the high frequency surgical process of the comparator, if point B and point C are directly connected to the comparator input, magnetic force noise May be generated to the threshold voltage through the leechlike capacitance of the MOS junction transistor. Therefore, this tour connects point in time B and point C to the buffer. Electrical circuit simulations show that the use of buffers tooshie effectively isolate electromagnetic interference and stabilize the room access voltage.
2.2 Generation of consign and liberation current
Up-to-date proportional to the supply voltage can be generated by OPA, MN2, and R5. Since the attain of the OPA is high, the voltage difference betwixt Vref and V5 is negligible. Due to the channel modulation effect, the currents of MP11 and MN10 are agonistic by the germ-drain potential. Therefore, the charge-discharge current of the capacitor is no longer linear with the supply potential difference. In this design, the current mirror uses cascode structure to stabilize the source-drain voltage of MP11 and MN10, and reduce the sensitivity to the power supply voltage. From an AC perspective, the cascode structure increases the output resistance of the current source (layer) and reduces the error in the output current. MN3, MN4, and MP5 are used to provide a bias potential for the MP12. MP8, MP10, MN6 can cater predetermine voltage for MN9.
2.3 Logic Control Section
The output CLK and CLK of the flip-flop are honest wave signals with inverse phases, which can embody used to control the opening and closing of MP13, MN11 and MP14, MN12. MP14 and MN11 act American Samoa switching transistors, which function as SW1 and SW2 in Figure 1. MN12 and MP13 act as auxiliary tubes, whose main function is to deoxidize the burrs of the thrill and discharge afoot and eliminate the sharp-shooting phenomenon of triangular waves. The sharp-pip phenomenon is mainly caused by the canalise electric charge injection consequence when the MOS transistor is in the state changeover.
Assuming that MN12 and MP13 are removed, when CLK transitions from 0 to 1, MP14 is aroused to the hit state, and the on-going source unperturbed of MP11 and MP12 is forced to enter the deep unsubdivided region from the saturation realm instantaneously, and MP11, MP12, MP13 are The channel charge is closed out in a rattling short time, which causes a large glitch current, causing a spike voltage at maneuver A. At the same time, MN11 jumps from the off state to the on state, and the prevailing layers composed of MN10 and MN9 go from the cryptical linear region to the saturation region. The channel electrical condenser of these three tubes is positively charged in a short time, which too causes a large Burr current and spike voltage. Similarly, if the auxiliary pipe MN12 is removed, the MN11, MN10, and MN9 too give a large bug current and a lace voltage when the CLK is hopped. Although MP13 and MP14 have the indistinguishable width-to-length ratio, the logic gate level is opposite, so MP13 and MP14 are alternately turned on. MP13 plays two main roles in eliminating the empale voltage. First, ensure that MP11 and MP12 work in the chroma region during the intact cycle to insure the continuity of the circulating and nullify the sharp-shooting voltage caused by the current mirror. Second, make MP13 and MP14 form a complementary tube. Thus, at the moment of the CLK voltage change, the channel capacitance of indefinite metro is charged, and the channel electrical condenser of the other thermionic vacuum tube is pink-slipped, and the positive and damaging charges cancel each other, thereby greatly reducing the glitch new. Similarly, the introduction of MN12 will act as the like function.
2.4 Application of repair engineering
The parameters of different batches of MOS tubes will vary between wafers. Under different process angles, the thickness of the oxide bed of the MOS tube leave also be divergent, and the proportionate Cox will also modify accordingly, causing the charge and discharge current to shift, causing the output relative frequency of the oscillator to change. In integrated racing circuit design, the trimming technology is mainly used to modify the resistance and resistor network (or capacitor network). Different resistor networks can be used to increase or decrease the electric resistance (operating room capacitance) to design different resistor networks (or capacitor networks). The charge and electric arc currents IB1 and IB2 are primarily determined by the current Iref. And Iref=Vdd/2R5. Consequently, this innovation chooses to prune the resistor R5. The trimming network is shown in Figure 3. In the figure, totally resistors are isometrical. In this design, the electric resistance of resistance R5 is 45kΩ. R5 is affiliated in series past cardinal small resistors with a resistivity of 4.5kΩ. Fusing the wire between the two points A and B can increase the resistance of R5 by 2.5%, and fusing the wire between B and C tush increase the resistivity away 1.25%, between A, B and B, C. The fuses are all breathless, which increases the resistance by 3.75%. The disadvantage of this trimming technique is that it potty only increase the resistance prise, but not the reduced one and only.
Estimate 3 resistance repair web social organization
Footstep 3: Simulation Results Analysis
This design lav atomic number 4 enforced on CSMC's 0.5μm CMOS process and ass be simulated with the Phantom tool.
3.1 Improvement of triangular beckon aside complemental switching tube
Figure 4 is a schematic plot showing the improvement of the triangular wave by the complementary switch vacuum tube. It can follow seen from Libyan Islamic Grou. 4 that the waveforms of MP13 and MN12 in that design have atomic number 102 provable peaks when the slope changes, and the wave shape sharpening phenomenon disappears after the auxiliary tube is added.
Anatomy 4 Improved wave form of the complementary shift tube to the triangular wave
3.2 Influence of power furnish voltage and temperature
It can be seen from Figure of speech 5 that the frequency of the oscillator changes to 1.86% when the power supply voltage changes from 3V to 5V. When the temperature changes from -40°C to 120°C, the oscillator frequency changes by 1.93%. It can be seen that when the temperature and the power supply potential difference vary widely, the outturn frequency of the oscillator can remain constant, so that the average surgical process of the chip can glucinium ensured.
Image 5 Effect of voltage and temperature along frequency
Step 4: Determination
This paper designs a current restricted oscillator for Class D audio power amplifiers. Typically, this oscillator can output solid and triangular wave signals with a frequency of 250 kilocycle per second. Moreover, the output absolute frequency of the oscillator can remain stable when the temperature and supply voltage vary widely. Additionally, the spike voltage can also be removed by adding complementary switching transistors. By introducing a resistor network trimming proficiency, an high-fidelity output frequency hindquarters be obtained in the presence of process variations. Currently, this oscillator has been used in a Class D audio amplifier.
Step 5:
Step 6:
Be the First to Share
Recommendations
-
Anything Goes Contest 2022
Source: https://www.instructables.com/Design-of-Current-Mode-Based-Oscillator-for-Class-/
Posted by: healeywimen1958.blogspot.com
0 Response to "Design of Current Mode Based Oscillator for Class D Audio Power Amplifiers : 6 Steps - healeywimen1958"
Post a Comment