Design of Dual Channel Digital Voice Monitor Based on DSP

With the development of digital signal processing chips ( DSP ), voice codec technology is increasingly used in military, civilian and surveillance fields. As far as the voice monitor is concerned, it can be applied to many fields such as transportation, public security, fire protection, and telephone service quality supervision. In September 1999, the Ministry of Railways announced the "Technical Conditions for Digital Communication Recorders for Railway Transportation", which made the functions, technical requirements, measurement methods, inspection rules, etc. of the voice recording instruments used in the operation and dispatching of the railway transportation department. Detailed regulations. The system described in this paper complies with its various regulations and has passed the appraisal of the Ministry of Railways and is now being promoted and improved.

1 System function · FLASH semiconductor memory is used to make a pluggable card structure with variable storage capacity. The system is a dual-channel, dual-card architecture with a single card designed to store up to 14 hours of voice. The system can also work in a single card state.
· The user interface is the keyboard and LCD display, including the time display and status display of each function operation.
· The system is connected to the communication line, that is, it is in the monitor recording mode. When there is voice transmission on the line, the system starts the recording operation and simultaneously records the appearance time of the voice. The user can monitor the recorded voice to ensure that the voice has been recorded on the memory card. The recording process is repeated. If the card is not removed, the system will overwrite the oldest data with the latest voice data to ensure the latest hours of voice recording. Recording is a system background operation.
· The user can perform operations such as playback, loop playback, fast forward, fast reverse, and search by recording time. The length of each voice can also be known. These features are password protected and operate in the foreground of the system.
· The foreground operation can switch between two memory cards, while the background operation runs independently, regardless of the foreground operation, depending on whether there is voice on the line. In other words, playback, monitoring, retrieval and other operations do not affect the recording, you can record and play duplex work at the same time.

2 system overall module design As shown in Figure 1, the whole system can be divided into five functional units such as FLASH memory card, signal processing, signal input and output, system control and power management.

2.1 FLASH Memory Card South Korea's SAMSUNG Semiconductor's large-capacity memory has a high performance-price ratio. The system uses Samsung's 64-megabit FLASH memory chip KM29U64000 to store voice data. The chip operates at 2.7~3.6V, and its command, address, data line are multiplexed, and 8-bit width is transmitted. It can be erased a million times, and the data storage time is 10 years after power failure. The chip can store more than 3.5 hours of speech at a code rate of 5.3 kb/s, so that up to 14 hours of voice information can be recorded on a memory card using four of the chips.

2.2 Signal Processing Unit This unit includes a digital signal processor (DSP), codec (CODEC) and some control devices.
The DSP part adopts the 16-bit fixed-point digital signal processor ADSP2181 produced by ANALOG DEVICES. This is the core chip of the system for signal processing. It has the following characteristics:
· The operation speed is 33MIPS, which is suitable for the operation speed requirements of high quality medium and low rate speech coding algorithms;
· The chip contains 80K Byte RAM, which is divided into 48K Byte program RAM and 32K Byte data RAM;
· Two independent programmable full-duplex serial communication interfaces, support A law/μ law hardware decompression expansion, support automatic buffer operation. The two serial ports are used to transmit and receive voice data and communicate with the single chip microcomputer in the system;
· 4M Byte external addressing space;
· Support DMA operations between internal and external storage for storing DSP programs, algorithms and data;
· 13 programmable I/O ports for communication with microcontrollers and programmable logic devices.
The main role of DSP in the system is: (1) managing the data in the memory card. Since the recording time of the system is required to be long and the data is large, each storage page adopts the same data structure, and the voice data and the clock data are mixed and stored. The storage method is sequential and loop-covered. (2) Recording. Listen to the line at any time, perform voice coding, and store the compressed data in real time into the FLASH chip. The system adopts a speech coding algorithm improved according to ITU-T standard G.723, and the compression code rate reaches 5.3Kb/s, which has excellent speech quality. (3) Play. The speech decoding is performed according to the single chip command. (4) Voice search. This includes three functions: jump by voice segment, jump by 3 seconds, and time by user input.
DSP is the core processing chip of this system. Since the recording is a background operation, it is required to not interrupt the recording during playback, so the codec process is duplex; in addition, since the stored data is more, if the ordinary sequential search mode is adopted, The search process can take seconds, so the program uses a dichotomy to improve search efficiency.
The CODEC (PCM codec) used with the ADSP2181 is the MC145480 from MOTOROLA. The chip can realize A/D conversion, D/A conversion and A-law, μ-law companding of speech signals. The built-in RC filter can effectively filter out-band noise of input and output signals. There are a lot of CODECs that can be used with the ADSP2181, and the performance is not much different, so there is no special consideration in chip selection here.

2.3 Signal Input and Output Unit This unit is a full analog circuit that completes the mixing of input signals, gain control, and amplification of output signals. When the system is configured as a dual-card dual-channel model, each channel can access two signals with different gains; when configured as a single-card single-channel model, four signals can be accessed. In addition, there is a microphone input, mainly used for system testing. The system has automatic gain control for the access signals of both channels, and the control range can reach 15dB. Within this range, the circuit adjusts the input signal to the optimum amplitude for the CODEC operation, resulting in optimal DSP encoding. The output circuit is relatively simple, and the analog output signal of the two CODECs drives the speaker through the power amplifier.
In order to adapt to the signal channel with different gains in the railway sector, and to minimize the loss of voice quality of the DSP codec, an analog circuit with excellent performance is needed. This circuit unit is completed in the test according to the user's needs and the whole machine debugging result, and it is completed after many modifications. It cooperates with other units to make the DSP codec algorithm achieve its best performance.

Cabinet Light

Lighting is an integral part of building, and different lights have different functions, which can bring different spatial effects.Especial cabinet light, it become very important role in home lighting . It are mainly used for local lighting and decoration . There are two types of installation, one is surface mounting, it is used to on cabinets, wardrobe, reveal ark. Another type is recessed mounting. It is so suitable for wine cabinet . LEDER has a variety shapes of cabinet light, not only offer the effect of illumination, more let you have a beautiful feeling.

Cabinet Light

Under Counter Lighting,Cabinet Light Fixtures ,Under Counter Led Lights,Kitchen Under Cabinet Lighting

JIANGMEN LEDER LIGHTING CO., LTD , https://www.lederlight.com