The telephone CODEC usually has a Pulse Code Modulation (PCM) interface. Strictly speaking, the PCM concept contains most of the digital formats we are using today, including I2S; the original intention of PCM is to distinguish between digital encoding and analog techniques such as frequency modulation. However, in digital telephony, PCM generally refers to a specific tone data format that is not compatible with Hi-Fi stereo.
This article refers to the address: http://
The advent of computer audio has also spawned the emergence of another type of interface. Since quality requirements are similar to the existing consumer audio market, there is a need to play recorded audio files at different sampling rates (especially 8 kHz, 44.1 kHz, and 48 kHz). Although sampling rate conversion in software is possible, it is also very expensive. Therefore, the AC 97 standard, which is currently in common use, hands this task to the CODEC, and it can achieve very high efficiency through dedicated hardware. Currently, AC 97 has become the dominant industry standard in the field of computer audio.
The portable system initially maintained its original identity: the personal CD, mini disc and MP3 player use the I2S DAC, the mobile phone retains the PCM technology, and the audio-enhanced PDA generally uses the same AC 97 as the desktop computer. CODEC. Therefore, this is shocking. The first-generation combination system usually consists of two boards, a telephone and a PDA. The two are back-to-back arranged in one chassis. The PCM call CODEC is controlled by the communication processor, Hi-Fi stereo (AC 97). Or I2S) CODEC is controlled by the application processor. However, the CODEC is not designed for this type of application at all, and there is little or no consideration for the interconnection between the two audio subsystems. As a result, engineers typically insert discrete solidstate switches into the analog signal path, but this also introduces clicks, clicks, and harmonic distortion, and takes up PCB space.
Integration
Obviously, an integrated solution tailored to the above applications is very popular. Inspired by system-on-chip (SoC) design concepts, some vendors have integrated stereo DACs or CODECs into large-scale integrated circuits. However, this type of approach does not achieve the audio quality achieved by a dedicated audio chip. Combining power management ICs and audio ICs often compromises audio quality because power conditioners typically introduce noise into nearby audio signal paths.
Integrating audio into digital ICs is also tricky because true Hi-Fi components typically require a 0.35μm process that is best suited for mixed-signal applications, and digital logic circuits have evolved to 0.18μm or even below. For these two single-chip integration methods, it is not the performance of the compromised analog domain, that is, the size of the entire chip is increased to an unacceptable level (if the entire IC is designed according to larger geometric principles).
The horn amplifier generates a lot of heat and requires a suitable heat sink, which is especially difficult to integrate. Many combo chips lack this capability and therefore cannot be treated as a true "system-on-chip" solution, which typically requires an external speaker driver IC. Another common problem is the lack of analog input or output due to the desire to minimize the size of the IC. In square packages (such as QFN quad flat package, no leads), the pins are arranged around the IC. A few extra pins can be accommodated by extending the length of each side by 1 mm, resulting in a significant increase in IC pitch. For example, an increase from 5 x 5 mm to 6 x 6 mm requires an additional 11 mm2 PCB area; if starting from a 10 x 10 mm package, the additional area is 21 mm2.
Dedicated audio ICs avoid these problems. By integrating other mixed-signal functions such as Touch Screen digitization with call CODEC and Hi-Fi CODEC, the total number of chip pins can still be reduced. Here, the call CODEC is integrated into the phone chipset, and the Hi-Fi CODEC with additional analog input, output and internal mixing functions may be appropriate. On the other hand, a dual CODEC with a dedicated PCM interface for Bluetooth connectivity is also very beneficial.
There are many ways to implement audio integration. Sharing ADCs and DACs can reduce hardware costs, but not both audio streams can be played or recorded simultaneously. Configuring a dedicated converter for each function overcomes this problem and extends battery life because the power consumption of telephonygrade audio modules can be designed to be lower than the power consumption of Hi-Fi functions. However, such solutions add to the cost of the round. A common trade-off is to use some discrete DACs, but share these ADCs. This allows the audio to be played simultaneously (such as the ringing or music of the second incoming call), but the application processor does not record these during the call - this is an acceptable limitation because the user is like this In the use case, I don't want to see how much the second call has the value of the call. Turning off one channel and running the other at a low sample rate controls the power consumption of the DAC.
Timing and interface
Although it is possible to share internal circuit blocks between the communication and application domains, this situation does not occur in interface applications. This is because each audio stream runs at a separate clock domain at its own clock frequency. Only in this case, the combined smart phone CODEC requires a PCM interface and a separate I2S or AC 97 connection.
In fixed systems, the audio clock is usually generated by a quartz crystal oscillator. For example, AC 97 specifies that a CODEC that conforms to this specification should have an on-chip oscillator connected to an external 24.576 MHz (512 x 48 kHz) quartz crystal, while I2S parts use multiple sample rates, the most common being 256. However, in smart phone design, driven by additional power consumption, PCB space, and clock quartz crystal cost, the designer had to derive a Hi-Fi clock from another clock already on the PCB. Although complex odd frequency ratios need to be implemented by phase-locked loops (PLLs), this approach still favors the use of external quartz crystals because low-power, low-noise PLLs can be integrated into mixed-signal ICs at a lower cost. in. This method is also suitable for some clock signals that other subsystems may require, such as the standard 27 MHz clock of an MPEG decoder in video enhanced telephony applications. In the I2S CODEC, different sampling rates require different clock frequencies. By simply multiplying the word clock LRCLK (which is the sampling rate) by 256 or any other fixed number, the PLL can provide the correct clock in each case. Therefore, component manufacturers generally prefer to integrate one or two PLLs in their smartphone CODEC.
microphone
In the smart phone design, many of the toughest problems are related to the microphone. There are usually at least two types of microphones we consider: built-in (internal) microphones and external microphones, which are part of the headset. In order to eliminate noise or achieve stereo recording, some auxiliary internal microphones may be required, and the car handsfree device may be connected to another external microphone. In addition to talking, these microphones can also record call logs under the control of the application processor, or even the sound track of a video clip.
To completely eliminate off-chip switching, the smartphone CODEC needs to provide enough microphone input, preferably with independently adjustable gain, and a flexible internal routing path to cover all applications. In addition to the recording function, a "side tone" function should also be provided. This adds an attenuated version of the analog output to the analog output so that the caller can hear their own voice. Insertion detection enables seamless switching between the internal microphone and the external microphone when the headset is plugged in or unplugged.
Noise is another common concern. The high frequency and digital components in the line generate interfering signals that are captured by the PCB trace of the loaded microphone signal and amplified by the on-chip preamplifier. Careful board routing plays a big role in avoiding this problem, and using differential microphone inputs is another effective method. However, the inputs have their own unique wiring requirements: the two PCB traces must be side by side and adjacent to each other so that any noise present on one line must also appear in the other line, thus making it Completely disappears from the mic preamp.
Eliminating noise is a separation problem that requires two microphones: one to pick up the speaker speech with background noise and the other to be responsible for background noise. In the analog world, simple subtraction operations can hardly achieve satisfactory results because the two noise signals differ in phase and amplitude, depending on the direction in which the noise comes in. Digital signal processing is also required here, but by digitizing the two microphone signals, the CODEC must make the task easier to accomplish.
Another type of noise that appears in outdoor applications is wind noise. This noise is mainly limited to frequencies below 200 Hz and can be greatly reduced by high-pass filters. The simplest solution is to use a smaller capacitance coupling capacitor at the input stage of the microphone. However, this also prevents the microphone from being used in indoor music recording - which may make the bass disappear. For dual-purpose microphones, filtering should therefore be an option. It is worth mentioning that most audio ADCs already have a built-in high-pass filter to remove the DC bias in the digital signal. Integrated circuit vendors have customized this feature for mobile phone applications: a few Hz for Hi-Fi and 100-200 Hz for calls with wind noise filtering enhancements. Of course, analog filtering and digital filtering can also be combined to create higher order filter characteristics.
Headphones and earplugs
A specific analog circuit is also required for mobile phone headsets. The first obvious task is to reroute the output signal from the earpiece or other speaker to the headset when the headset is plugged in. Although mechanical switches incorporating sockets can do this, they are bulky and costly.
Moreover, the signal level used for the speaker may not be suitable for headphones. Independent analog outputs for the handset, speakers, and earbuds with independent volume control solve this problem and allow for a simpler socket. Although a mechanical switch is still required, a single-hole, single-throw mechanical switch with one end connected to the ground pin is sufficient, so the socket requires only one external pin. However, in a multimedia phone, the activation of the switch does not necessarily represent the insertion of the headset; for a standard size socket, this may be equivalent to a headset without a microphone. Therefore, the presence or absence of a microphone should be detected independently.
For electret microphones, this can be done by monitoring the microphone's bias current: if no current flows, there is no microphone insertion. Conversely, an unusually large bias current is also meaningful: in order to avoid the addition of a single contact to the standard earphone/earphone socket, the button that answers the call from the earbud (also known as the hookswitch) usually shorts out the microphone. . Therefore, the bias current increases, indicating that the hookswitch is depressed. Adding a current sensor to the on-chip microphone bias circuit allows the smartphone CODEC to detect both conditions and automatically take the correct action for each situation.
speaker
Recently, the number of speakers and output power in mobile phones have increased. However, in the 1990s, a single earpiece was a basic configuration. The modern clamshell design features internal and external speakers that can be used to play their respective sounds when the phone is turned on or off. Stereo ringing requires two external speakers, and the popular hands-free feature may require a “large†(referred to as mobile phone standard) speaker in addition to a small handset. For microphones, providing a dedicated analog output for each speaker has many advantages over off-chip switching. Since speaker amplifiers can absorb large supply currents, it is critical to turn off their power supply in a non-energized state. Smartphone CODECs offer an increasing number of granular power management features that allow each output to be enabled or disabled, avoiding any unnecessary battery drain.
In addition, voltage regulators in existing power management schemes typically do not provide sufficient current to drive a speaker that operates at maximum volume. In response to this problem, CODEC manufacturers have designed some on-chip speaker amplifiers to run directly through the battery (the typical voltage of a lithium-ion battery is around 4.2V) instead of the adjusted supply voltage. Although this is usually not energy efficient (the speaker amplifier consumes extra power, the voltage regulator does the same), but does not require an additional voltage regulator.
Bell
Compared with the past few years, the complexity of the bell has been steadily increasing. It has evolved from the monotonous "哔哔" sequence tones into multi-tones, even WAV and MP3, literally speaking of any type of sound. Can be made into stereo. MIDI has become a multi-tone ring interface standard, and many manufacturers have developed low-cost MIDI chips for this new application. Integrating such an IC into the audio subsystem requires the CODEC to have an additional analog input.
These additional inputs are also useful for connecting FM radio ICs, adding another feature to multimedia devices. The MIDI ringtone generation circuit should of course be integrated into the CODEC; however, the general trend is to save the random ring as a sound file and play it through the existing Hi-Fi DAC, which limits the attraction of the idea to IC manufacturers. Force because their IP products don't yet contain MIDI.
future development
So, what about the future of smart phone audio? At present, some digital audio trends worthy of attention include: shifting from stereo to multi-channel surround sound format, and the recently launched "Azalia High Definition Audio" standard is likely to be used by most PCs and Used by laptops.
Although not long ago, those who once ridiculed the idea of ​​placing stereo speakers on mobile phones have been proven wrong by today's market reality; but in the foreseeable future, handheld devices cannot develop into multiple channels. Similarly, it is currently not proven that the cost and power consumption of the new Azalia features is higher than AC 97. The controversy between I2S and AC 97 is continuing, with some designers preferring less complex I2S interfaces, while others prefer AC 97 with fewer pin counts and various sample rates that are easier to handle. Since many low-power processors for mobile phones are now able to provide dual-standard audio interfaces that meet the above two camp hobbies, these two standards may continue to coexist. Conversely, designing a CODEC that supports two standards is quite difficult, because VRA (Variable Data Rate Audio), represented by AC 97, requires a completely different timing scheme than I2S, and a significant number of additional digital circuits.
The experience of successfully integrating an application processor and a communications processor into a single digital device using an audio clock will enable the combination of the call interface and the Hi-Fi audio interface, and may eventually lead the industry back to the low complexity CODEC. But for now, IC vendors are focusing on integrating other existing mixed-signal components into their audio CODECs, including touch-screen functions, voltage regulators, and power management.
Until now, the integration of imaging solutions has prevented the integration of audio and camera or video functions, but this is by no means a law set in stone. At the same time, audio features such as 3D enhancement, graphics compensation and dynamic compression are beginning to spread, and sound quality, power consumption and package size are becoming more sophisticated.
Touch Screen
Jumei Video(Shenzhen)Co.,Ltd , https://www.jmsxdisplay.com