This article refers to the address: http:// Audio-specific special problems are easily encountered in portable product design. Because audio circuits seem simple, engineers typically do not spend too much time on relatively low-frequency audio circuits (20Hz to 20KHz). This article attempts to start with the most basic audio circuit design, to provide some design reference ideas and methods for engineering designers. Finally turn on the audio circuit This simple principle may be the most important, but it is often overlooked by system designers. Power amplifiers cannot distinguish between noise, hum and signal. If the amplifier is turned on prematurely, it will amplify all input signals without distinction. Portable product playback circuits typically include digital signal memories, digital-to-analog converters (DACs), amplifiers, speakers, or headphones (Figure 1). The digital signal in the memory is decoded and sent to the DAC for conversion. The analog output of the DAC is AC coupled to the input of the amplifier through a capacitor. The amplifier must be able to supply enough current to drive the low-impedance speaker. As mentioned above, when the amplifier is enabled, it will amplify any signal entering its input, including useful signals, noise, chirps or clicks. As shown in Figure 2, the speaker amplifier is connected between the 8Ω speaker and the audio DAC. An ac coupling capacitor between the DAC output and the amplifier is required to ensure proper input and output bias voltages for both devices. Most audio amplifiers have a bias voltage at the output, which is preset in order to reliably transmit the audio signal. A certain time interval must be set before turning on the power amplifier to establish the proper bias voltage. If the power amplifier is turned on prematurely, the bias voltage at which the DAC output is in the climb phase is equivalent to an attenuating pulse for the amplifier input. The signal is amplified by the -amplifier and enters the speaker, producing an audible click. Figure 2 assumes that the power amplifier is already on and has established an input bias before the DAC is turned on. When the DAC is enabled, the voltage at node A will climb to the DAC output bias voltage as shown. When the bias voltage of the DAC climbs, the high-pass filter consisting of the coupling capacitor and the input resistance of the amplifier produces a glitch at node B. The output signal after the amplifier is equal to the difference between the input signals [(IN+)-( IN-)] is multiplied by the gain of the amplifier. Low frequency response and input time constant The bias voltage used to isolate the DAC and the input capacitance of the amplifier input port, together with the input impedance of the amplifier, form a high-pass filter. Consider using a larger capacity capacitor to reduce low frequency attenuation, but due to the input bias voltage of the power amplifier, an increased input time constant may result in an output click. If the amplifier is turned on before the input is stable, it will cause a click. In the simplified model of the input of the power amplifier, the input impedance is represented by R IN , and the non-inverting terminal of the preamplifier is connected to the internal reference voltage. This input structure is a typical structure of a single-supply power amplifier. When the /SHDN of the amplifier is pulled high, the amplifier is activated after a fixed delay. This delay is called the turn-on time (t ON ) and is defined in the Electrical Characteristics section of the device manual. Figure 3 shows the waveforms at the input and output of the power amplifier when /SHDN is pulled high and the input capacitance is the recommended value. It can be seen that the input bias voltage of the power amplifier begins to climb after /SHDN is pulled high, but the output stage is still off. The time that the input bias voltage reaches the normal value is determined by the capacitance C IN and the input resistance (R IN ) of the amplifier. It is reasonable to set the turn-on time of the amplifier to establish a stable input bias voltage before the output stage turns on. For most power amplifiers, the turn-on time is fixed (in Figure 3, t ON = 24ms). When setting the turn-on time, the IC design engineer must consider the amplifier's input impedance as well as the input bias voltage and input bias capacitance. The input capacitor is chosen by the application engineer to provide a fast response time constant and to ensure that the low frequency response is as flat as possible. The test waveform in Figure 3 shows that after the /SHDN pin is pulled high, the input bias voltage climbs to a normal value, delays t ON and activates the output. If the activated output is turned on smoothly during this process, the speaker will not beep. Component selection Figure 4 shows the waveform when an oversized C IN is selected. The selected capacitor is 10 times the normal value. From the waveform, the C IN low frequency response is fairly flat, but the time constant is 10 times the original. The turn-on time of the amplifier is fixed at t ON , so when the output of the amplifier is already on, the input bias voltage is still rising! The power amplifier treats this voltage as a normal signal and amplifies it, resulting in a large output step in the speaker, resulting in an offensive hum. Note that the oscilloscope scale in the figure is 5V/div instead of 100mV/div. Explain this in an extreme case: we chose an input capacitor that is much larger than the recommended value. A certain margin is usually set when the input capacitor is selected so that the input bias voltage rises to the final value before t ON . In order to increase the C IN when necessary, leave a certain margin. In order to ultimately optimize the input capacitance, some experiments must be performed using the device manual to provide R IN and t ON . Knowing the low frequency response of the speaker is very helpful for the design. If the power amplifier drives a small size speaker that is difficult to recover low frequency signals, it is best to send all frequency components to the speaker. In this case, the best choice should be the standard C IN value. The speaker frequency response curve is usually available from the speaker manufacturer, data sheet, or from the manufacturer. Volume control design More and more audio ICs have volume control that can be programmed through the serial port or adjusted with the dc voltage of a DAC or digital potentiometer. The volume control circuit can help the end product manufacturer to optimize the turn-on time. If the actual application requires a special low-frequency response, it is inevitable to use a large input capacitor. In this case, the volume control circuit can be used to keep the output off for a certain period of time, and the input bias is completed. Set up. The simplified circuit in Figure 5 is a power amplifier with volume control. The volume of the IC is controlled by a separate pin (VOL). The VOL pin is connected to the input of the coarse ADC. The DC voltage applied to the VOL is encoded by the ADC. The code reflects a specific gain level. (VOL=V DD is the fully off state, VOL=GND is the maximum volume state.) The best way to ensure that there is no click is to keep the volume at the minimum output setting until /SHDN is pulled high and exceeds the t ON delay, then V VOL changes slowly (any wait time beyond t ON helps Input bias is stable). Volume control allows the use of large capacitors while providing acceptable click/pop suppression (Figure 6). It should be noted that increasing the input capacitance by 10 times is an extreme case, and is for illustrative purposes only. Output coupling capacitor Conventional single-supply amplifiers have a DC bias voltage at the output, typically half the supply voltage, which needs to be removed from the signal before feeding into the speaker (to avoid damaging the audio coil), usually requiring a larger Output capacitors for DC filtering. In order to avoid large attenuation of the low frequency components of the audio signal, a large capacitance is required. If the designer needs a particularly flat passband response and the passband is extended to a lower frequency (less than 100 Hz), then a large and expensive output capacitor is required. For example, a 100uF capacitor is used to achieve a frequency response as low as 50Hz under 32Ω load conditions. When the amplifier is turned on, such a large capacitance also causes a click sound during the turn-on process. The DC blocking capacitor and the load of the speaker together form a high-pass filter. When the DC bias is applied as a step voltage to the DC-blocking capacitor output, the load terminal of the capacitor rises at the same time and decays according to the size of the capacitor and the time constant determined by the load. This pulse signal produces an audible noise through the speaker. In order to eliminate the hum, the most popular way is to use a "capacitorless amplifier". Typically, such amplifiers use another amplifier to bias the speaker or be configured as a differential output (BTL) amplifier. The best capacitorless amplifiers can be connected directly to the speaker (Maxim calls DirectDrive) and do not require a bias amplifier or differential output. The DirectDrive amplifier contains an internal inverting charge pump that is used by the charge pump to generate a negative voltage for the output stage. The output stage is driven by positive and negative supplies, and since the output signal is biased at ground potential, the amplifier no longer needs to bias the speaker. Designers can replace two large output coupling capacitors with a pair of small charge pump capacitors. DirectDrive amplifiers have twice the dynamic range of conventional amplifiers or bias amplifiers. Figures 7A-7D show three single-supply amplifiers. Figure A shows a conventional stereo audio amplifier with a DC blocking capacitor at the output; Figure B shows a “no-capacitance†amplifier that uses a third amplifier to generate a bias voltage; A DirectDrive amplifier that requires any capacitance on the signal path; the differential output amplifier is shown in Figure D. Direct connection to the speaker can greatly reduce the click/pop when powering on and off. In this case, the click is only related to the output offset of the amplifier. The typical output offset voltage of the DirectDrive amplifier is ±1mV to ±5mV. The small offset voltage step at startup produces only minimal turn-on transient response and is noticeable by hearing-sensitive people. Design Class D Amplifier Class D amplifiers produce switching outputs, and audio information is stored in the pulse width modulated signal of the output signal, which is very efficient compared to Class AB amplifiers, but high efficiency comes at the cost of cost. In order to achieve high efficiency, the output stage of the amplifier must be switched quickly so that the output transistor quickly passes through the linear region. This high speed switching produces large transient currents in the speaker coil, resulting in strong electromagnetic interference (EMI). In order to reduce EMI, it is necessary to shorten the connection between the speaker and the Class D amplifier as much as possible. It is best to place the amplifier near the speaker to shorten the lead length of the speaker, which is capable of transmitting EMI to the surrounding circuitry. It is often difficult to place the power amplifier near two speakers because the speakers must be separated by a distance to achieve an effective stereo effect. In order to achieve stereo effects while reducing EMI, it is best to replace the stereo amplifier with two mono Class D amplifiers. If a mono amplifier cannot be selected due to cost constraints, the use of ferrite beads with long cables can reduce EMI well. Use an inexpensive ferrite bead and a small 1nF capacitor on the output pins of each Class D amplifier to reduce EMI (assuming a Class D amplifier requires a filtered modulation architecture, which means that the load voltage is zero at zero input) zero). Figure 8 shows a Class D speaker amplifier with an output containing a ferrite filter. The output spectrum comparison is also provided with and without a ferrite filter. Box-type transformer substation Combined Transformer,Power Transformer,Distribution Transformer Bao Ding Tian Wei Group Transformer Co., Ltd. , http://www.jspowertransformer.com
Figure 1: Typical audio subsystem.
Figure 2: The large size coupling capacitor and the input and output bias voltages together cause the speaker subsystem to hum.
Figure 3: Input and output waveforms for the circuit of Figure 2 when the appropriate input coupling capacitor is selected.
Figure 4: Input and output waveforms of the circuit of Figure 2 when the capacitance is increased by a factor of 10.
Figure 5: This Class AB audio power amplifier includes volume control.
Figure 6: Using the volume control function of the audio IC to compensate for large input coupling capacitors.
Figure 7: Traditional single-supply audio amplifier and a new “capacitorless†audio amplifier.
Figure 8: Class D amplifiers contain a ferrite EMI bead at each output, and the lower curve gives an output spectrum comparison with and without a filter.