New voice compression codec AMBE2000

This article refers to the address: http://

Abstract: AMBE2000 is a patented voice compression/decompression chip introduced by DVSI. The chip is capable of full-duplex compression and decompression to provide high quality speech quality at very high compression rates. The main features, pin functions, control interfaces and application circuits are described in the paper.

Keywords: speech compression; format word; FEC; VAD; AMBE2000

1 Introduction

AMBE2000 is a high-performance, low-power single-chip real-time voice compression decompression chip with a compression ratio ranging from 2kbps to 9.6kbps, with FEC (Forward Error Correction) and VAD (Voice Activation Detection). ) Function and DTMF signal detection function, so it can be widely used in satellite communication, digital mobile communication, secure communication, voice mail and so on.

2 pin description and characteristic parameters

The main features and parameters of AMBE2000 are as follows:

● High quality voice quality;

● The compressed data rate can be adjusted by software in the range of 2kbps to 9.6kbps;

● The error correction rate ranges from 50bps to 7.2kbps;

●With VAD (Voice Activation Detection) / CNI (Insert Comfort Noise) function;

● With DTMF signal detection and generation function;

●Power supply voltage: +3.3V;

● Operating temperature range: -40 ~ +100 ° C;

● It is packaged in a 100-pin TQFP.

Table 1 Host Interface Selection

Port type Select foot
CH_SEL1 CH_SEL0
Active format 0 0
Active unformatted 0 1
Passive format 1 0
Passive plain 1 1

Table 2 Voice data bit selection (no format)

Number of voice data bits in the word Select foot
BA_SEL1 BA_SEL0
1 0 0
2 0 1
3 1 0
4 1 1

The pin descriptions for the AMBE2000 are as follows:

CHANN_SEL0~1 (75,77) feet: host interface type selection terminal? The specific operation is listed in Table 1;

BAUD_SEL0~1 (80~81) feet: Unformatted voice data bit selection terminal? See Table 2 for details.

RATE_SEL0~4 (70~74) feet: speech compression rate and forward error correction rate selection end? The specific operations are listed in Table 3;

CODEC_SEL0~1 (84~85) feet: A/D and D/A interface select ports? The specific operations are listed in Table 4;

Table 3 AMBE2000 compressed data rate selection table

Pin 74 Pin 73 Pin 72 Pin 71 Pin 70 Speech rate (bps) FEC rate (bps) Total data rate (bps)
1 1 1 1 1 2000 0 2000
1 0 0 0 0 3600 0 3600
1 0 0 0 1 4000 0 4000
1 0 1 1 0 2400 1600
1 0 0 1 0 4800 0 4800
1 1 0 0 0 4000 800
1 0 1 1 1 3600 1200
1 1 0 0 1 2400 2400
1 0 0 1 1 6400 0 6400
1 1 0 1 0 4000 2400
1 1 0 1 1 4400 2800 7200
1 0 1 0 0 8000 0 8000
1 1 1 0 0 4000 4000
1 0 1 0 1 9600 0 9600
1 1 1 0 1 3600 6000
1 1 1 1 0 2400 7200

*Note: The compressed data rate can also be changed by the control word.

Table 4 A/DD/A hardware configuration

A/DD/A type CODEC_SEL1 CODEC_SEL0
Normal 16-bit linear (sampling rate 8 kHz) 0 0
AD company AD73311 (sampling rate 32kHz) 0 1
Ordinary μ law (sampling rate 8 kHz) 1 0
Normal A law (sampling rate 8 kHz) 1 1

VAD_EN (86) pin: Enable voice activation detection function (active high);

ECHOCAN_EN (78) pin: Enable echo cancellation function (active high);

SLEEP_EN (83) pin: Sleep mode control port (active high);

SLIP_EN (82) pin: Sliding mode control port (active high);

X2/CLKIN (68) pin: 16.384MHz clock input pin;

X1 (67) pin: internal oscillator output;

RESETN (69) pin: reset pin (low active);

EPR (20) foot: coding package ready flag;

SOFT_EN (79) pin: Enable software error judgment port (active high);

CHAN_RX_DATA (32) pin: used to receive host data to AMBE2000;

CHAN_TX_DATA (42) pin: used to send AMBE2000 data to the host;

CHAN_RX_LK (28) pin: The host receives the clock;

CHAN_TX_CLK (34) pin: The host sends a clock;

CHAN_TX_STRB (38): The host sends/receives a frame sync pulse;

CHAN_RX_STRB (30) pin: The host receives the frame sync pulse port;

CODEC_RX_STRB (29) pin: A/D frame sync pulse receiving end;

CODEC_TX_STRB (37) pin: D/A frame sync pulse transmission port;

CODEC_RX_DATA (31): A/D sends PCM data to AMBE2000;

CODEC_TX_DATA (41) pin: AMBE2000 sends PCM data to D/A;

CODEC_RX_CLK (27) pin: A/D clock;

CODEC_TX_CLK (33) pin: D/A clock;

CLOCK_MODE (51) pin: Clock mode select port, select the crystal oscillation mode when the port is connected to the high level; select the external clock when connected to the low level);

VDD (8,11,12,23,36,39,44,45,46,47,48,49,54,57,64,76,87,90 feet): positive power supply;

GND (1,9,10,25,26,35,40,50,52,53,56,63,65,88,89 feet): ground terminal;

The remaining pins: N. C, no need to hang.

3 host interface and data format

There are two types of interface data formats between AMBE2000 and the host: one is formatted; the other is unformatted.

3.1 formatted

In the case of format, AMBE2000 generates a compressed data packet every 20ms. This data packet consists of 24 words. After the data packet is generated, its EPR pin will output a short negative pulse indication, which will determine the data. The processing of the package. The specific composition of the data packet is as follows:

Format word 0: fixed to x13ECH, which is the header of the packet;

Format Word 1: High 8 bits for chip power control, 0x00 for normal operation mode, and 0x55 for low power mode. The lower 8 bits function (high effective) is listed in Table 5;

Table 5 format word 1 format

Types of 7: MSB 6 5 4 3 2 1 0: LSB
enter Make up lost frames NC NC NC NC NC Insert comfort noise NC
Output NC NC Decoding frame repetition Decoding silence indication NC NC Coded silence detection Coded DTMF detection

Table 6 format word 2-6 format

Word 2 Word 3 Word 4 Word 5 Word 6 Speech rate (bps) FEC rate (bps) Total data rate (bps)
0x0028 0x0000 0x0000 0x0000 0x6428 2000 0 2000
0x5048 0x0000 0x0000 0x0000 0x3948 3600 0 3600
0x5250 0x0000 0x0000 0x0000 0x4150 4000 0 4000
0x1030 0x0001 0x0000 0x34la 0x6750 2400 1600
0x5360 0x0000 0x0000 0x0000 0x6c60 4800 0 4800
0x5250 0x2010 0x0000 0x0000 0x7460 4000 800
0x5048 0x0001 0x0000 0x2412 0x6860 3600 1200
0x1030 0x0005 0x180c 0x3018 0x7360 2400 2400
0x6b80 0x0000 0x0000 0x0000 0x6c80 6400 0 6400
0x5250 0x0001 0x0000 0x542a 0x5280 4000 2400
0x5258 0x0009 0xle0c 0x4127 0x7390 4400 2800 7200
0x7fa0 0x0000 0x0000 0x0000 0x52a0 8000 0 8000
0x5250 0x0005 0x2010 0x6834 0x72a0 40000 4000
0x7fc0 0x0000 0x0000 0x0000 0x69c0 9600 0 9600
0x5048 0x000e 0x4010 0x6a2e 0x65c0 3600 6000
0x1030 0x000e 0x68la 0x511b 0x76c0 2400 7200

Format words 2 to 6: used to specify and represent the compressed data rate of the chip, as listed in Table 6;

Format word 7: This bit is not used in the input packet, so it should be set to 0x0000 at design time and the bit error rate in the output packet;

Format words 8, 9: not used, usually set to 0x0000;

Format Word 10: The upper 8 bits are used to control and represent the amplitude of the DTMF signal, ranging from 3 to -60 dBm0 (eg: 0x03=3, 0x00=0, 0xC4=-60); the lower 8 bits are used to control and detect the DTMF signal. The signals emitted and detected are standard North American calling tones. The DTMF codes are listed in Table 7.

Table 7 DTMF code

DTMF code DTMF number DTMF code DTMF number
0x80 1 0x84 2
0x88 3 0x81 4
0x85 5 0x89 6
0x82 7 0x86 8
0x8A 9 0x87 0
0x83 * 0x8B #
0x8C A 0x8D B
0x8E C 0x8F D
0xFF DTMF is invalid

Table 8 Format Word 11 Format

Types of 7: MSB 6 5 4 3 2 1 0: LSB
enter NC NC Voice activation (1) NC Sleep (1) Echo cancellation (1) Encoding/decoding selection
Output NC NC NC NC Sleep (1) Echo cancellation (1) Encoding/decoding selection

Format Word 11: The upper 8 bits are used for the decoder volume control, where 0x80 represents the original volume, 0xFF is twice the original volume, and less than 0x80 is the volume reduction; the lower 8 bits are listed in Table 8. The encoding/decoding is selected as follows: 00 is encoding + decoding; 01 is encoding; 10 is decoding; 11 is not used;

Format words 12 to 23: Compress data information. The number of data bits varies depending on the data rate. The relationship between the number of data bits and the compressed data rate is as follows:

Number of bits = compressed data rate (bps) / 50 (bps)

For example, at a minimum compression ratio of 2 kbps, the number of data bits is:

2000/50=40.

3.2 Unformatted

In the case of no format, only the data bits need to be transmitted, but the data is reduced by 50 bits per second (compared to the formatted case), thus degrading the voice quality. See Table 2 for the number of bits of data contained in each word. The relationship between the number of words per frame and the compressed data rate is as follows:

Word count = [compressed data rate (bps) / 50 (bps)] / number of digits

In the formula, the selected number of bits must be divisible by the selected compressed data rate. The AMBE2000 serial clock rate is up to 2.048MHz.

Figure 1 Practical circuit for speech sampling, compression and decompression composed of AMBE2000 and AD73311

4 application circuit

Figure 1 shows a practical sampling circuit for compressing, compressing and decompressing the AMBE2000 and AD73311. The AD73311 is a vocoder chip of A-DI, which can sample data at 32 kHz and 16 bits. It has good sound quality and can be used with AMBE2000. It has achieved good results in practical applications. In the figure, the pin connection selected by AMBE2000 defines the compression ratio as 2kbps. At the same time, the circuit is set in the external input clock mode, and the echo cancellation, voice activation, and sliding compensation functions are valid, and the EPR pin is used to output the packet ready indication signal. After the circuit is powered up, the AMBE2000 can also be reset by software. Therefore, the AMBE2000 voice compression codec is very flexible to apply.

5 Conclusion

Voice communication plays an important role in digital communication. The data compression method formed by AMBE2000 and AD73311 described in this paper can not only save storage resources, but also reduce the need for communication bandwidth. Therefore, in some special application fields, AMBE2000 voice compression codec chip has a very broad application prospect.

Water cooling computer kit is a package that launched by the manufacturer, one kind is all water cooling accessories, players purchased and according to individual case to install, other kind is that manufacturer provide the full suit which including computer hardware and water cooling accessories. manufacturer will provide many varies of solutions for buyers to choose, in particular case, buyers can customized produce according to the personal preference. In a word, water-cooling kit simplifies the process of players using a water cooling computer,they can enjoy water cooling computer faster and more convenient.

Computer Water Cooling Kits

Computer Water Cooling Kits,Water Cooling Computer Kits,Diy Water Cooling Computer Kits,Colorful Cooled Computer Kits

Dongyuan Syscooling Technology Co., Ltd. , http://www.syscooling.com