Where the circuits have L, C elements, so what is L? C? In theory, an LC circuit is also called a tank circuit, a tuned circuit or resonant circuit is an electric circuit built with a capacitor denoted by the letter ‘C’ and an inductor denoted by the letter ‘L’ connected together. That is, a combined circuit of capacitor and inductor. When get its meaning, you will have another question: where LC circuits used? You can get the answer from What is the Purpose of LC Circuits?

Summary: This text has explained the D audio amplifier technology and key element of the design in detail, and introduced the products characteristic of the class d amplifier of ADI Company. Keyword: Class d amplifier; Power of output stage; Export the transistor-resistor logic; EMI; LC filter

The tone quality is in the class d amplifier, the overall tone quality to be obtained must solve several problems. " the card is put up " Sound: Very disagreeable that the card given out when the amplifier is turned on or disconnected is put up sound. But unfortunately, they are easy to introduce in the class d amplifier, unless pay special attention to modulator state, time sequence and LC filter state of output stage as the noise quieting non disjunction noise quieting of the amplifier. Signal to noise ratio (SNR) : In order to avoid the Fizz that the background noise of the amplifier produces, as to the low power amplifier of walkie application, SNR should usually exceed 90 dB, designs SNR to exceed 100 dB as to the medium-sized power and should exceed 110 dB to design as to high-power. This can reach to various amplifiers, but the concrete noise source of necessary tracing, in order to guarantee to reach overall SNR that is satisfied with at time when the amplifier is designed. Distorted mechanism: The nonlinearity in it is paid attention to that distorted including the modulation technique or modulator is realized, and in order to solve the output stage of the impulse current problem. Usually encode the information including signal amplitude of the audio frequency in the D pulse width of modulator output. The Dead Time used for preventing the impulse current of the output stage from being additional will introduce the non-linear sequence error, the distortion which it produces in the loudspeaker, with directly proportional to the sequence error of ideal impulse widthing. Use for, avoid, bludgon most short Dead Time into to distorted to it minimizes to be most favorable frequently; Want to understand that please see and read materials 2 to optimize the distorted characteristic detailed design method of the output stage of the switch thoroughly. Other distorted sources include: Export the mismatch of pulse rise time and fall time, output the characteristic mismatch of time sequence of driving circuit of the grid of the transistor-resistor logic, and the nonlinearity of LC low-pass filter components and parts. The power is inhibited (PSR) : In the circuit illustrated in Fig. 2, the power supply noise nearly couples to and exports the loudspeaker directly, have very small inhibit function. This thing happen because the transistor-resistor logic of the output stage connects the power to the low-pass filter through a very low resistance. The electric-wave filter inhibits the high frequency noise, but all frequency of the audio frequency will be passed, including audio frequency noise. About please see the deep reading material 3 to the detalied specification of the single-end and circuit power noise effect of output stage of differentiating switching. If does not solve distorted problem and power problem, it is very difficult to reach PSR and is superior to 10 dB, or total harmonic distortion (THD) Superior to 0.1%. Even worse situation, the high-order that THD tends to the harmful tone quality is distorted. Fortunately, solve these problems in some good solutions. Use with high loop gain feedback (just as adopt in a lot of linear amplifier designs) Very useful. The feedback that the LC filter inputs will improve PSR and and decays all is not distorted sources of LC filters greatly. Non-linear LC filter can until loudspeaker that include go on, decay in feedback loop. In the loop-locked class d amplifier designed meticulously, can reach PSR > 60 dB and THD < 0.01% of the high-fidelity tone quality. But feedbacking makes the design of the amplifier complicated, because must meet stability (a kind of very complicated consideration to the high design of steps) of the cycle . The analog feedback of continuous time is essential too for catching the important information about sequence error of the pulse, so control the cycle must include the analogous circuit in order to deal with the feedback signal. In the ic amplifier is realized, this will increase the cost of die. In order to minimise IC cost, some manufacturers like not using or using the minimum artificial circuit part. Some products detect the change of power with a digital open loop modulator and an analog-digital converter, and compensate by adjusting the modulator, this can be seen and read materials 3 thoroughly. So can improve PSR, but will not solve any distorted problem. Other digital modulators attempt to carry on precompensation to the anticipated output stage sequence error, or correct the non- ideal modulator. So will deal with some distorted sources at least, but not all. Require loose application as to the tone quality, can deal with through the class d amplifier of these open loops, but to the best tone quality, some formal feedback seems to be essential.

The modulator of class d amplifier of modulation technique can have many kinds of methods to realize, a large number of relevant research and intellectual property right are supported. As space is limited, omitted here.

EMI high frequency commponent of dealing with D amplifier output worth considering conscientiously. If incorrect understood and dealt with, the weight will produce a large number of EMI and interfere with the work of other apparatuses. Two kinds of EMI need considering: Signal radiating the space and through the loudspeaker and signal that the power line conducted. The class d amplifier modulation scheme determines to conduct the group line spectra of EMI and radiation EMI weight. However, can use some board-level design methods to reduce the class d amplifier emissive EMI, and no matter how its line spectrum of group is. One useful principle to minimize cycle area to bear the weight of the high-frequency current, because the intensity, cycle area and cycle correlated to EMI relate to degree of closeness of other circuit. For example, the whole LC filter (including loudspeaker interstar connection) Overall arrangement should getting closer as much as possible,be and kept close to amplifier. Electric-current driving and the intersection of loop and track should concentrate on together in order to minimize the intersection of cycle and area . Another place that should notice will produce the large transient electric charge as the switch of grid capacitance of transistor-resistor logic of the output stage. Usually this electric charge comes from the storage capacitor, thus form one and include two capacitive current loops. Through minimize the intersection of cycle and area EMI of the attitude influence in the reducible cycle, mean the storage capacitor should charge it close to the transistor-resistor logic as much as possible. Sometimes, insert strategic point shed coil to be very much helpful with the intersection of amplifier and the intersection of power and cascade connected RF. Assign them correctly to limit the high-frequency transient current in the local loop of the amplifier, will not conduct at a long distance along the power line. If gate drive does not overlap time very long, know the parasitic diode in the crystal pipe end of the output stage of forward bias in induction current of the loudspeaker or LC filter. When not overlapping time, finishing, diode bias from turn reverse into forward. Before the complete separation of diode, will present the large reverse recovery current overshoot, thus produced EMI source of the trouble. Through keep, overlap very short time (propose distorted to minimize audio frequency still) Minimize EMI. If the reverse recovery scheme is still unacceptable, can use the Schottky (Schottky) The parasitic diode of diode and this transistor-resistor logic is connected in parallel, prevent the parasitic diode from being turned on all the time with the transfer current. This is very helpful, because the metal semiconductor junction of Schottky diode is free of backward recovery effect in essence. It is minimized that the LC filter with annular inductor magnetic core can influence the stray on-the-spot power transmission line that the electric current of the amplifier causes. A good mean method between cost and EMI characteristic is to reduce the radiation from low-cost drum-shaped magnetic core through shielding.

LC filter design, in order to save cost and PCB area, the LC filters of most class d amplifiers are adopted second order and low-passly and designed. Fig. 3 shows the second order LC filter of mathematical fraction of a difference. The loudspeaker is used for weakening the periodic reconance of the circuit. Though the intersection of loudspeaker and impedance similar to simple resistance sometimes, actual relatively more complicated impedance and include prominent idle component. Should win the best filter design result, the design engineer should always try to use the rigorous loudspeaker model. Common the intersection of electric-wave filter and design alternative for respond to, drop, minimize by winning the lowest bandwidth electric-wave filter under required highest the intersection of audio frequency and the intersection of frequency and terms in purpose. If for up to 20 kHz frequency, demand to drop less than 1 dB, requires the typical electric-wave filter to have 40 kHz Bart's fertile Sri Lankan (Butterworth)ing Respond to (in order to reach the maximally-flat band pass) . As to common loudspeaker impedance and canonial L value and C value, Table 1 provides nominal components and parts value and corresponding approximate Butterworth to respond to.

Table 1 if the design excluding loudspeaker is feedbacked, the loudspeaker THD will be to having sensitive linear scale of the components and parts of the LC filter. Design the consideration with inductor: Design or choose the important factor of the inductor to include rated current and shape of the magnetic core, and wind line resistance. Rated current: It should be greater than the highest electric current of the desired amplifier to choose the rated current of the magnetic core. The reason is that if the electric current exceeds the threshold value of the rated current and the current density is too high, the magnetic saturation will take place in a lot of inductor magnetic cores, lead to the fact the inductance is reduced sharply, this is that we are not desired. Form the inductance through turn around the magnetic core. If the number of turns is numerous, the resistance correlated to total turn length is very important. Because this resistance connects in series between semi-bridge and loudspeaker, will consume some output power. If resistance too high, should use choosing, requiring the intersection of number of turns and less other the intersection of metal and magnetic core of material around wired and thick, in order to offer the required inductance. Finally, don't forget the shape of inductors used will influence EMI too, just as mentioning above.

In the voice system of the inuse class d amplifier of the system cost, which its cost of totality of influence of important factors is there? How could we minimise the cost? The active device of the class d amplifier is output stage of the switch and modulator. The cost which forms this circuit is roughly the same as simulation linear amplifier. It is other components and parts of a system that the ones that really need to consider trade off. The low power consumption of the class d amplifier has saved the cost (and PCB area) of the heat dissipating double-fuselage ,For example, air-cooling fin or fan. The D ic amplifier can be adopted than the capsulation of little and low cost of imitating the linear amplifier size. When driving the digital audio source, simulation linear amplifier needs digital-analogue converter (DAC) Change the voice signal into the analog signal. Need to change so to the class d amplifier which deals with analog input too, but the integrated DAC function effectively of class d amplifier to digital input. On the other hand, the prime cost shortcoming of the class d amplifier is a LC filter. The components and parts of the LC filter, especially the inductor, take up PCB area and increase the cost. In the High Power Amplifier, the overall system cost of the class d amplifier is still competitive, because a large number of cost saved in the heat dissipating double-fuselage can offset the cost of LC filter. But in low cost, low power consumption are employed, the inductor is with very high costs. In extremely specific cases, for example, the low-cost amplifier used for cell phone, perhaps the cost of the amplifier IC is lower than the total cost of the LC filter. Even neglect the consideration of cost, PCB area that the LC filter takes up is a question in the small-scale application too. Consider, will totally cancel the LC filter sometimes, does not have the filter amplifier to design in order to adopt in order to satisfy these. So can save cost and PCB area, though has lost the advantage of the low-pass filter. Without the electric-wave filter, the increase of EMI and high-frequency power consumption will be unacceptable, unless the loudspeaker adopts the inductance type and nears the amplifier very much, the current loop is minimum in area, and the power level keeps very low. Though this kind of design is often adopted while employing portably, for example, cell phone, but is not fit for the high-power system, for example, the stereo of family. Another method is to count and reduce each of required LC filter components and parts of audio frequency passway to the minimum. This can be through using the output stage of single-end semi-bridge to realize, its required inductor and capacitor quantity are half of the whole-bridge circuit of differentiating. But if the semi-bridge output stage needs power of bipolar, it then will be too high with producing the relevant cost of negative power supply, unless the negative power supply has some other purposes, or the amplifier has abundant audio frequency passways, in order to share the cost of negative power supply. In addition, the semi-bridge can also supply power from the single power, but will reduce the output power and often need to use a big separating the direct current container in this way.

All design questions that the class d amplifier of ADI Company was discussed just now could sum up in quite rigorous project of a requirement. In order to save the design engineer's time, ADI Company offers various class d amplifiers IC, they contain programmable gain amplifier, modulator and output stage of power. In order to simplified and assess, ADI Company offered the demonstrating board for amplifier of each kind of type. PCB routing and material bill of these demonstrating boards can be regarded as the feasible reference design, thus needn't mainly design the problem to do to design to help the customer rapidly in order to solve the class d amplifier through proving, but the valid voice system of economy " The work of repeatability " . For example, may consider using AD19902, AD19923, AD19944 and AD199655 dual the intersection of amplifier and IC series product, they are suit requiring to be exported and up to 5 in every passway of two passways, 10,stereosound or monophonic application of the medium-sized power of 25 and 40 W. Thank ADI Mr. and Rajeev Morajkar, Art Kalb of Company, Mr. to herein to advise while being beneficial.

Read materials 1 thoroughly. International Rectifier, Application Note AN-978, " HV Floating MOS-Gate Driver ICs." 2. Nyboe, F., et al, " Time Domain Analysis of Open-Loop Distortion in Class D Amplifier Output Stages," presented at the AES 27th International Conference, Copenhagen, Denmark, September 2005.3. Zhang, L., et al, " Real-Time Power Supply Compensation for Noise-Shaped Class D Amplifier," Presented at the 117th AES Convention, San Francisco, CA, October 2004.4. Nielsen, K., " A Review and Comparison of Pulse-Width Modulation (PWM) Methods for Analog and Digital Input Switching Power Amplifiers," Presented at the 102nd AES Convention, Munich, Germany, March 1997.5. Putzeys, B., " Simple Self-Oscillating Class D Amplifier with Full Output Filter Control," Presented at the 118th AES Convention, Barcelona, Spain, May 2005.6. Gaalaas, E., et al, " Integrated Stereo Delta-Sigma Class D Amplifier," IEEE J. Solid-State Circuits, vol. 40,no. 12,December 2005, pp. 2388-2397. About the AD199x Modulator.7. Morrow, P., et al, " A 20-W Stereo Class D Audio Output Stage in 0.6 mm BCDMOS Technology," IEEE J. Solid-State Circuits, vol. 39,no. 11,November 2004, pp. 1948-1958. About the AD199x Switching Output Stage.8. PWM and Class-D Amplifiers with ADSP-BF535 Blackfin(r) Processors, Analog Devices Engineer-to-Engineer Note EE-242.ADI website: www.analog.com (Search) EE-242 (Go)