Audio Amplifiers are designed to increase the amplitude of a small input signal while maintaining the linearity of the original signal. In other words, the detail and characteristics of the original signal will be maintained. The audio amplifier market has seen many changes, resulting in a wide selection of amplifier ICs available to design engineers.
Audio amplifier ICs are now available to address specific applications with a certain amplifier class and number of channels. Audio amplifiers are available in single channel (mono), two-channel (stereo), or multi-channel amplifiers that may support 6 or more channels of high quality audio. Various classes are used to differentiate between amplifier types for audio applications:
Class A amplifiers are simple analog gain stages that when operated short of clipping are highly linear, but are also highly inefficient.
Class B amplifiers amplify only half of the input waveform, but when implemented by matched transistors in a push-pull configuration, this arrangement is much more efficient than Class A; however watch out for crossover distortion when the input signal changes polarity.
Class AB amplifiers eliminate crossover distortion by not completely turning off the non-conducting transistor in a complementary pair, trading off efficiency for linearity. Additionally, Class A and B amplifiers provide current amplification which, depending on the specific application and design configuration, may require a blocking capacitor to prevent DC from reaching the load.
Class D amplifiers modulate sinusoidal input with a triangular wave, passing the resulting pulse width modulation (PWM) to high-power switching circuits, usually MOSFETs. The output from a Class D amplifier goes through a low-pass LC filter to remove switching transients and restore the waveform to its original form. Class D amplifiers are much more efficient and therefore smaller than Class AB amps, so they’re typically used in portable or space constrained designs. Other classes include Class G, Class DG, Class H and other less common classes.
Audio Amplifiers are available at Mouser Electronics from many industry leading manufacturers, including Texas Instruments, Maxim, STMicroelectronics, Analog Devices, NXP, THAT Corporation & more.
When transmitting audio or video signals over any distance, the connectors are potentially the most fragile part of the signal path. Paying attention to these seemingly mundane devices can pay major dividends in terms of signal integrity and product reliability. Many analog audio and video connections use standard RCA connectors, sometimes referred to as “phono connectors.” RCA connectors are commonly used in consumer audio applications such as connections for component hi-fi systems, but are also frequently used for IEC 60958 Type II (S/PDIF) digital audio signals.
Phone plugs are another popular connector and are available in three standard sizes, including the original ¼-inch (6.35 mm), 3.5 mm mini-jack and the 2.5 mm sub-miniature jack. Where more robust, professional audio connections are needed, XLR connectors provide the best choice. Advantages of XLR connections include the ability to lock the connector in place and provide a balanced connection that can potentially reduce unwanted noise. XLR connections can support both high quality analog and IEC 60958 Type I/AES3 digital audio signals.
The higher data rates required to move real time video may require fiber-optic FO5/TOSLINK (IEC 60958 Type II Optical) jacks and plugs. However, different data protocols often demand different connectors. With video you have a choice of HDMI, DVI, DVI-D, and Displayport connectors, although embedded designs with LCD screens may use onboard FFC or FPC connectors. While connectors are usually the last thing designed into a project, make sure they receive the attention they deserve.
If you want to do any digital processing with analog audio you’ll need to convert your analog signals into the digital domain using Audio Data Converters. As their names imply, Audio A/D Converters convert signals from the analog world into the digital domain and after processing Audio D/A converters will convert the digital signals back to the analog domain. There are different types of data converters, but the key specs to compare in audio applications are their resolution (number of bits) and the sampling rate. These specification will affect the dynamic range which should be pre-determined as part of your design. Compact discs provide high quality music (20 Hz to 20 kHz) are sampled at 44.1 kHz with 16-bit precision. High resolution audio more commonly uses 24-bit resolution with samples rates of 192 kHz or more. On the other end of the spectrum, telephone quality speech (200 Hz to 3.2 kHz) can be sampled at 8 kHz with 12 bit precision.
Evaluation boards let you quickly determine just how well the audio processor that looked good on the datasheet will perform in your pending design. Full development kits go even further and include a wide range of on-board peripherals, application programming interfaces (APIs), reference designs, firmware, software libraries, sample source code, user interfaces, and a serial or USB connection to connect the development kit hardware to a PC host. With a development board you can quickly prototype, test, and debug your design, eliminating the time and cost involved in developing a series of prototype boards.
Development boards typically come with either an onboard debugger or a separate debugger and an onboard debug interface. Check the software library for functions and filters that can be useful in audio applications—they can save a good deal of programming time and effort. Development kits pay for themselves (and more) in terms of shortened development time and avoidance of errors in the design process.
When an event occurs that requires attention there are two ways to generate an alert: sight and sound. Usually an LED lights up and an audible alarm sounds. The audio alert can be anything from the buzzer on a cell phone to a resonant voice that announces, “I’m sorry Dave, I’m afraid I can’t do that.”
Audio alert devices are simple electromechanical transducers that span a tremendous range of technologies (buzzers, horns, strobes, sirens, tone generators), tones (chirps, chimes, beeps, whoops, and warbles), frequencies (6.3-9100 Hz), volumes (6-127 dB), and voltage ratings (0.1-220V). With so many audio indicators, there is bound to be an appropriate device available for any given application.
Product selection is very much application driven. How loud does it need to be? How immediate is the need for attention? A buzzer may be sufficient to alert the person in the next room that the dryer load is finished; but a smoke detector will need to issue a piercing alarm. When an event occurs that requires attention and it’s not safe to assume that someone will quickly see a visual alert, an audio alert or alarm should be a required part of the system design.
Amplifying the output from a microphone enough to fill a large auditorium requires a great deal of gain. In order to obtain the best possible signal-to-noise ratio (SNR) it’s imperative to amplify the microphone’s output as quickly and cleanly as possible. If you start with a noisy signal, no amount of downstream filtering will give you a satisfactory result.
Microphone preamplifier ICs are specialized amplifiers that provide high gain, low noise, low distortion, and—if embedded in the microphone—low power; their characteristics are dictated by the type of microphone.
Potentiometers (Pots) are variable resistance devices commonly used as the interface to control various features in audio products. They are simple but important devices that allow you to dial in the desired amount of resistance which may affect volume, tone or other aspects of the audio signal. Typically pots are analog devices but digital pots that mimic the behavior of their analog counterparts are also available. Potentiometers come in rotary or slide panel packages with various sizes, shaft configurations and tapers, and can be found in products designed for the pro audio, consumer, portable and automotive markets.
Audio signals almost invariably require processing, whether to filter out noise, remove echoes, reduce distortion, optimize frequency response, or compress them for transmission or storage. Audio signal processors offer a variety of tools to assist in optimizing sound for audibility, intelligibility, and fidelity. In commercial audio applications, audio signal processors are typically employed within or just after the mixing stage, but before amplification.
Audio signal processors generally fall into one of two categories: Audio SoCs—integrating multiple digital audio management features into a single, complex package—and audio DSPs, which range from hardware implementations of specific codecs to fully programmable DSPs. Audio DSPs are available with a wide range of features, so you can select the feature set to match design requirements.
Almost all MCUs can handle basic control applications; however processing high-speed audio signals in real-time is far more computationally intensive. Digital signal processing involves algorithms and techniques used to manipulate signals after they’ve been converted into digital form. Different algorithms are needed to enhance samples, recognize or generate speech, or compress data for transmission or storage. Implementing them entirely in software would be unacceptably slow, so some sort of hardware acceleration is needed. MCU manufacturers have added DSP functions to some MCUs to expand capability, creating a class of MCUs referred to as “digital signal controllers” (DSCs). DSCs combine the control capability of a CPU with the signal processing ability of a DSP.
Speakers are electromechanical transducers that convert an audio signal to audible sound, the opposite of microphones, with which they share many of the same characteristics. Like microphones, there different types of speakers.
The smallest speakers, if you can call them that, are the electret transducers used in hearing aids. Marvels of miniaturization, they include an electret microphone, programmable DSP processor, audio amplifier, and electret audio output device, all of which must work for days off a tiny zinc-air battery. The electret transducer makes that possible. Earbuds are the next step up in size, but they don’t face the same power constraints as hearing aids. They’re typically dynamic speakers, with a very small coil attached to a moving diaphragm. Larger headphones use the same design, though high-end ones may use electrostatic drivers that are essentially condenser microphones in reverse, with an electrically charged plate suspended between two perforated metal electrodes to which the audio signal is applied. Electrostatic speakers are more linear than moving coil designs and produce very crisp sound. The downside is they require a relatively high voltage power supply and they can’t move enough air to reproduce low frequencies well.
Dynamic speakers pass the signal through a voice coil that surrounds a permanent magnet; the signal causes the coil to move along with anything to which it’s attached. Dynamic speakers are simple, inexpensive, and can move a lot of air. At the low end of the sound spectrum are the tiny dynamic speakers in cell phones, which typically cover 400-20,000 HZ with up to five percent distortion, which is adequate for voice but not for music—hence earbuds.
The dynamic speakers used in high-fi systems are usually designed to work over a limited range of frequencies, with electronic crossovers dividing the audio signal between different speakers. Subwoofers require so much power that very low frequencies may require additional amplification.
While digital designers aren’t used to seeing transformers, they’re very useful in audio applications for isolation and impedance matching, both of which can be accomplished at the same time. Audio power amplifiers connect to 8-ohm loads, which is far lower than the output impedance of the amplifier. An output transformer matches the impedances and isolates the speaker from the amplifier’s collector or plate voltage.
When evaluating audio transformers it’s important to consider the impedance, power rating, isolation voltage, and frequency range. If the transformer will be operated near an RF or other EMI source, shielding may be necessary. Shielding shouldn’t be an issue with a low-impedance audio output transformer, but it definitely could be with a transformer feeding a high-impedance, high-gain input stage; in that case the transformer should be placed as close to the input jack as possible and far away from noise-emitting sources, for which its windings could act like an antenna.
Most audio transformers use laminated cores of high-permeability steel; these transformers are inexpensive to make and at audio frequencies display low eddy current and hysteresis loses. More expensive toroidal transformers have the primary and secondary windings wrapped concentrically around a donut-shaped permalloy or ferrite core (which are easily magnetized and de-magnetized.) Toroidal transformers are more efficient than those with laminated steel cores, plus they’re smaller, lighter, generate less hum, and are less susceptible to EMI. On the other hand, they display higher inrush current, are generally limited in power handling capability, harder to fabricate, and therefore more expensive. The type of audio transformer you choose really doesn’t matter as long as it meets your design requirements. Just be sure to take into account its external operating environment, too.
Analog audio can move readily from a microphone to an amplifier as long as the cable and connectors are shielded and in good repair. But audio that needs to travel to a remote site with some distance needs to be digitized, encoded, transmitted, decoded, and converted back to analog at the far end. The audio transmitter accepts audio, channel status, and user data, which is then multiplexed, encoded, and driven onto a cable; at the receiving end the audio receiver does the reverse. This is clearly a more complicated task than moving analog sound a few feet over a microphone cord.
Digital audio is rarely transmitted raw—it’s always compressed and later decompressed by a codec. The choice of codec depends on how much signal loss can be tolerated due to compression and how noisy the signal path is liable to be. Audio transmitters usually contain ADCs and hardware-based codecs; receivers are just the reverse; and transceivers contain the essential elements of both.
The choice of codec is usually dictated by the target application—e.g., mixing console, CD-R, telco, remote audio system, or set-top box. The transmission medium is also a factor; copper can’t handle the same transmission speeds as fiber optic cable. Finally, there are numerous interface standards for transmitting audio data, including AES/EBU, IEC 60958, S/PDIF, & AGIO CP-340/1201. Each of these standards specifies how to carry digital audio over various transmission media, all of which narrows down (or dictates) the choice of codecs, cables, connectors, and chips. Audio receivers must be able to recover the clock and synchronization signals and de-multiplex the audio and digital data, in addition to showing the channel status, user, and validity information. In comparing audio transmitters and receivers, compare baseline specs (frame rate, power) but also whether they can encode/decode audio data according to the applicable specification while staying within your power, distortion, and SNR design requirements.
Several types of test equipment can be used to capture and measure audio signals. The simplest test involves putting a single audio tone (usually 1 kHz) through the audio system and measuring the output with an oscilloscope and distortion analyzer, determining the acceptable input level before achieving an unacceptable level of total harmonic distortion (THD), which usually occurs just before clipping. Sweeping the input frequency will indicate the system’s bandwidth (+/- 3 dB).
Specialized audio signal generators can generate signals ranging from a short impulse, a burst, or a continuous signal. The signal can be a standard shape (pulse, ramp, sine wave, etc.), a sweep of frequencies, a controlled amount of noise, or an arbitrary waveform.
Spectrum analyzers measure the frequency content of the signal, which can be helpful in troubleshooting a non-linear system response, where small amounts of harmonic signal may be obscured by primary frequencies.
A digital oscilloscope (DSO) can measure the flatness of audio system response with greater precision than a spectrum analyzer by performing an FFT (Fast Fourier Transform) analysis on the signal from DC up to the Nyquist frequency. DSOs are a required test tool when designing or debugging audio components/systems that incorporate serial data methods to control the device or to encode the audio information.
Used in conjunction with function/arbitrary signal generators, DSOs can view the signal in both the time and frequency domains; some are able to sync on and help decode various serial data streams. These are two pieces of equipment that should be on every audio designer’s test bench.