USB audio is a commonly used interface in most devices, unless it is the oldest personal computer hardware and operating system. With its robust connection and data transmission rate, one might think that it is simple to transmit high-quality audio over this interface. However, today's successful USB-based audio products are all doing a lot of chip and system work, and need to solve difficult problems such as clock recovery.
The essence of the problem is that the final output device sends audio to speakers, headphones, or line output sockets, which requires a "master clock" to adjust the audio conversion speed. This master clock needs to have two independent properties: 1) It must be an integer multiple of the audio sampling rate, which must be very accurate (so you do n’t need to discard or copy audio samples when the timing is wrong); 2) its jitter (Or it can be said that the phase noise) must be low enough so that the digital-to-analog conversion process will not be affected. The challenge here is that we have to meet both requirements.
Part of the difficulty comes from the fact that the receiving end of the data stream through the USB cable does not know the exact sampling rate. In fact, it can only infer the theoretical sampling rate. More importantly, the data from the USB cable does not have any form of clock. This is an obvious shortcoming compared to most other serial interfaces. Other serial interfaces either have a transmit clock or build data, so that when running, you can always find a clock from the connection.
The only clock information that can be obtained from the USB interface is that every millisecond of a specific type of data packet will send out the initial frame, and this event can be detected by the receiving hardware. According to the known method, this millisecond value can be derived from the system clock at the transmission end, and the original audio sampling rate is the same (we will briefly discuss an exception later).
A simple solution may be that we can put the 1 kHz clock into a PLL-based multiplier and multiply the frequency as needed to establish the audio master clock, and all the sub-clocks are based on this. However, in systems that process CD audio, the sampling frequency is 44.1 kHz, and the typical traditional audio digital-to-analog converter requires a master clock of 256 times, or 11.2896 MHz. The fact is that multiplying the input frequency by such a large multiple on a single PLL will certainly not be very good. This is hitting the crux of the multiplier: loop bandwidth, reference excitation rejection, and jitter of the voltage-controlled oscillator. More importantly, in this case, we need to multiply 1kHz by a number that is not an integer. It is even more difficult to complete this task.
The stacking of two fairly complex multiplier loops will lead to work with phase noise and false rejection. However, this method often results in high power consumption, which requires high-end chips and clever analog design. Or in this way, I would rather slow down accordingly to change the clock frequency requirements. The nominal sampling rate of the USB audio link may change rapidly between lines. Waiting for nearly a second to stabilize will result in unreliable performance. This method was originally used for digital audio connections in fixed-frequency studios, where cost and size are not important.
In the past few years, there have been different ways to create the required audio master clock, no longer need to be troubled by the PLL frequency multiplication problem, they have been integrated into many dedicated chipsets, such as USB speakers, headphones, external sound cards. These devices do what they need without having to spend extra chip area or pin count on "if and what" capabilities. This of course can reduce costs so that everyone is happy.
However, if your next-generation USB interface requirements cannot be met on a special function chip, what should you do? Mobile devices (such as media players and the latest writing pads) are built on new platforms and run new operating systems Yes, this requires increasingly standardized USB standards as a wide range of accessories and new features for wired connection options. Some of these systems have integrated USB audio chips, but cannot meet the demand, which has caused a "hit" for the device to provide basic functions. USB audio is one of the more and more functions required by these small mobile devices.
There are several advantages to extracting audio in digital form from a mobile device. Analog audio interfaces are no longer limited by system sound quality factors. This allows manufacturers of audio system or player accessories to achieve a higher level of sound performance through their own circuit design. It is also important that the digital audio link improves the impedance to the TDMA interface (the impedance of the cluster mobile device cellular modem coupled to the analog circuit of the audio playback portion of the system).
There are many microcontrollers on the market that integrate USB peripherals, but none of them are designed with the necessary clock generation and recovery circuits, and these are used to transmit high-quality audio data (this is the current demand). Sometimes this problem can be solved, you can use an external "clock restart" chip or a more complex audio converter (integrated PLL or sample rate converter) to compensate for the gap between the accuracy and quality of the main clock. However, this brings the system back to these problems: high cost, high power consumption, a large number of components, or all of them. In addition, the audio "down-conversion technology" makes a very long memory buffer can not be used in any system, video images (even slides) need to adjust the time for audio.
Honeycomb Activated Carbon Filter
Honeycomb Carbon Filter,Activated Carbon Air Filter,Honeycomb Mask With Carbon Filter,Honeycomb Activated Carbon Filter
Dongguan V1 Environmental Technology Co., Ltd. , https://www.v1airpurifier.com