This article refers to the address: http://
Abstract: AC4830xC-C is a voice-specific chip produced by AudioCodes of the United States. It supports international standards for voice codec with multiple bit rates, and provides fax and data relay functions. The TCM38C17 is a voice PCM encoding chip from TI Corporation of the United States. The article introduces the basic performance and working principle of the two chips, and presents an implementation scheme of a four-way speech codec system based on the design of these two chips.Key words: AC4830xC-C; TCM38C17; speech codec
1 Introduction
The AC4830xC-C series chip is a voice-specific chip produced by AudioCodes of the United States. It can compress PCM voice stream into voice signals of multiple bitrates as needed. The AC4830xC-C has multi-channel voice transmission function and operation of each channel. Independent. The chip has a wide range of applications in voice communications, especially in IP telephony systems. The TCM38C17 provides two different PCM encoding functions, A and U, as well as four independent channels. This paper mainly introduces the basic performance and working principle of these two kinds of voice chips, and gives a system design scheme based on these two chips to realize four-way speech codec.
2 Functional principle of AC4830xC-C
AC4830xC-C series devices include AC48301C-C, AC48302C-C and AC48304C-C, which can provide 1, 2, 4 voice channels respectively, and each channel in the multi-channel chip can operate independently without interference. . The basic working principle of the AC4830xC-C is shown in Figure 1.
In the design of the voice system, the AC4830xC-C chip can be used to compress the PCM voice code stream into voice signals of various code rates as follows:
(1) G. 711 PCM code rate is 64kbps;
(2) G. 726 ADPCM code rate is 16kbps ~ 40kbps;
(3) G. 727 E-ADPCM code rate is 16 ~ 40kbps;
(4) G. 729 CS-ACELP code rate is 8.0kbps;
(5) G. 723.1 MP-MLQ/ACELP code rate is 6.3/5.3kbps;
(6) The NetCoder code rate is 6.4 kbps to 9.6 kbps.
After receiving the voice PCM stream or fax data, the AC4830xC-C first performs input gain control and echo cancellation, and then distinguishes whether the specific service type is voice data or fax data. If it is voice data, the voice is according to specific requirements. The signal is encoded, and the encoded data is sent through the host interface; on the other hand, the AC4830xC-C also reads data from the host interface and judges it. If it is judged to be voice data, it is decoded into a PCM stream and Sended through the PCM interface.
The AC4830xC-C chip must have peripheral support to function properly. It provides five external interfaces, namely host interface (HPI), PCM interface, clock interface, memory, and I/O interface and emulation test interface. The host interface of the AC4830xC-C is an 8-bit parallel interface. The DSP can be used to control the AC4830xC-C through the host interface and exchange data with it.
The clock control part of the AC4830xC-C includes an externally connected crystal oscillator and an internal clock circuit. The operating clock is generally determined by the oscillation frequency of the external crystal oscillator and the internal PLL value. At design time, the operating clock can be changed by changing the value of the PLLMUL register.
In addition, the AC4830xC-C also offers six operating clock modes to choose from, but requires that its operating frequency be controlled between 98.304 and 100 MHz.
3 How does the TCM38C17 work?
The TCM38C17 is a PCM speech coding chip produced by Texas Instruments Inc. (TI). It provides inter-conversion between analog voice signals and PCM streams, and integrates both A- and U-law standards. In addition, the chip itself provides functions such as gain control. The chip integrates four PCM chips together, which is equivalent to the function of four TCM29C13. When the device is in operation, it usually needs an external circuit to provide the clock signal PCMCLK, the frame synchronization signal PCMFS, and the specific frame synchronization signal of each voice. The working sequence of the four-way cooperative operation is shown in FIG. 2 .
4 system hardware design
Based on the functions provided by the above two chips, an implementation scheme of a four-way speech codec system is given below, and FIG. 3 is a system block diagram thereof.
In the system, when the analog voice signal sent by the source is PCM encoded by TCM38C17, the PCM stream will be compression-encoded by AC4830xC-C to form the required voice stream, and then controlled, frequency modulated, etc. by the DSP processor. Processing, and finally sent through the serial/parallel converter. On the other hand, the DSP writes the data read from the serial/parallel converter to the AC4830xC-C through the host interface, and the AC4830xC-C decodes the speech stream to generate the PCM stream, which is then restored to analog by the TCM38C17. voice signal. Since the AC4830xC-C only provides one frame sync signal, a counter is used in this system to generate the other three frame sync signals required by the TCM38C17 chip.
5 system software design
The system software is mainly used to control the AC4830xC-C. AudioCodes specifies the operation method of the AC4830xC-C and provides corresponding initialization kernel programs and working programs. When working, you should first reset the AC4830xC-C through the DSP. After the reset is successful, first write the kernel program, then write the working program, and finally write these programs to the AC4830xC-C through the host interface of the AC4830xC-C.
After downloading the working program, the chip can start working. At this time, the chip is in the initialization mode. The default encoding rate is 32kbps ADPCM encoding. The designer can set a separate mode for each channel, including which voices are enabled. Channel, adjust the amplitude of each voice signal, select the corresponding voice coding method according to specific technical requirements. The AC4830xC-C reads in one frame of voice data every 30ms, and each frame of voice data is compressed and encoded into a voice packet. The data packet contains information about the voice data and the encoding method, and when the data is transmitted, the encoded information must also be transmitted to the receiver at the same time.
6 Conclusion
The AC4830xC-C is a powerful voice signal processing chip that provides a variety of voice codecs and is easy to operate. This paper uses its cooperation with TCM38C17 to build a codec system for multi-channel voice communication. It is believed that in the field of IP telephony and multimedia communication, chips such as AC4830xC-C will be more and more widely used.
Vacuum cleaner Co., Ltd. , http://www.gzvacuumcleaner.com