IP PBX CHIP

C5509: 32 channels of conferencing, EC. DTMF Detect, Tone Gen

C6416: 496 channels of conferencing, tone detect, tone gen, voice record/playback, Caller ID.

The Adaptive Digital Technologies conference chip allows the ultimate flexibility. The number of conferences and number of members per conference is fully programmable. The conferencing chip adds the active conference input signals together to form a composite signal. The conference chip makes use of sophisticated voice activity detection, noise reduction, and dominant speaker selection algorithms in order to maximize voice quality, even in very large conferences. Automatic Gain Control (AGC) is used to compensate for channels with different attenuation characteristics. Before sending the composite signal back to each conference member, that member’s transmission is removed from the composite signal to avoid the perception of echo. The composite signal is also made available for recording purposes.

In addition to the conferencing feature, the conferencing chip provides tone detection and tone generation capabilities. Tone detection is typically used in a conference server to allow the user to enter a conference ID and password, or to perform other control functions. Tone generation is used to generate alerting signals.

A voice-playback feature allows the host to send pre-recorded speech messages to individual conference members and to broadcast messages to all conference members. The voice record feature allows the host to capture each individual conference member’s TDM input as well as capturing the composite conference output. Voice playback and voice record data transfers occur via the host’s packet interface.

The conference chip uses TDM serial ports, a packet interface, or both, to input and output PCM data to each channel. The TDM serial ports are fully programmable to allow connection to nearly any type of serial bus.

Conferencing Chip ANSI “C” API

The conference chip is controlled by using an ANSI “C” set of API functions that are provided to run on a host processor. The host processor communicates to the conference chip via a host port interface. The API functions configure and control the conference chip as well as return status information to the host application

FEATURES

512 channels on a single chip
Excellent voice quality maintained even in large conferences
Supports TDM or packet interface for input and output data
Provides tone detection and generation
Provides voice playback and voice record of conference members and composite signal

AVAILABILITY

ADT Conferencing Chip is available on the TMS320™ DSP Family

C5509: 32 channels of conferencing, EC. DTMF Detect, Tone Gen

The conference chip provides all the DSP functionality necessary for a 32 channel conferencing bridge. Up to 10 independent conferences can be set up, and each conference can bridge together 2 to 32 voice channels.

C6416: 496 channels of conferencing, tone detect, tone gen, voice record/playback, Caller ID.

The figure below is a block diagram of the conference chip. The conference chip is controlled via the DSP’s Host Port Interface (HPI) (or PCI interface on some 64XX devices). In order to simplify the use of the conferencing chip, the conferencing chip’s ANSI “C” API software resides on the host processor. This API provides an abstraction layer that hides the details of the control mechanisms from the host application.

Assigning A Channel To A Conference

Channels are assigned to a conference by the host dynamically. The host issues an API command with parameters indicating the channel ID and conference ID. Channels may not be assigned to a conference if they are already active members of another conference or, if the conference to which they are assigned already has the maximum permissible number of members.

Removing A Channel From A Conference

Channels are removed from a conference by the host dynamically. The host issues an API command with parameters indicating the channel ID and conference ID. The conference is deemed inactive after all conference members have been removed.

Enabling/Disabling A Channel’s Dtmf Tone Detector

The host can enable or disable a channel’s DTMF tone detector dynamically via an API command with parameters indicating the Channel ID and an enabled/disabled flag.

Serial Port Configuration

In order to interface to a wide variety of serial TDM busses, the serial port configuration is programmable. Table 1 below describes the serial port configuration parameters.

Serial Port Characteristics
Parameter	Valid Range	Default Value
Serial Port 0	Enabled/Disabled	Enabled
Serial Port 1	Enabled/Disabled	Enabled
Serial Port 2	Enabled/Disabled	Enabled
Number of Time Slots	0..255	128
Use Standard Mapping	True / False	True
Data Format	u-Law, A-Law, 8 bit Linear, or 16 bit Linear	u-Law
Transmit Sync Polarity	Active High or Active Low	Active High
Receive Sync Polarity	Active High or Active Low	Active High
Transmit Clock Polarity	Rising Edge or Falling Edge	Rising Edge
Receive Clock Polarity	Rising Edge or Falling Edge	Falling Edge
Transmit Data Delay	0 to 2	1
Receive Data Delay	0 to 2	1
DX Pin Delay	Enable or Disable	Disable

Table 1: Serial Port Characteristics

Enabling a Conference’s Voice-Record

Each conference’s composite output is always packetized and copied to the host’s packet interface each frame. There is no specific command to enable this action.

Disabling a Channel’s Voice-Record

The host can command a channel to disable it’s voice-recording output dynamically. The host issues an API command with parameters indicating Channel ID. No data is sent to the packet interface if a channel’s voice-record is disabled.

Sending Voice-Playback data to a Channel or Conference

The host can playback voice messages to individual channels, or can send voice messages to all members of a conference simultaneously. Voice playback data is input to each channel via the host’s packet interface. Each channel polls its packet buffer once per frame for new data from the host. If new data is found in the packet buffer, that data will overwrite the transmit TDM data the channel would have written onto the TDM stream.

The host can also playback voice messages to all members of a conference. In this case, the host inputs the voice message to a conference channel’s input packet buffer via the packet interface. Each conference has an associated “phantom” member that allows playback messages to be input to the conference. A “phantom” conference member is not allocated a TDM timeslot resource..

Commanding a Channel to Generate a Tone

A channel can be commanded to generate a tone via a host API command. An API tone generation command includes the following parameters:

Channel ID
Tone generation Mode – selections are: Continuous (always on), Burst (single burst of duration on time), and Cadence (follows on, off time)
Tone Level – level in dBm of each frequency (0…-20)
On Time – number of samples of

ON time

Off Time – number of samples of OFF time
Frequency 1 – frequency 1 in Hz
Frequency 2 – frequency 2 in Hz

When a channel is not generating a tone, a silence pattern is sourced onto the channel’s TDM timeslot.

Enabling a Channel’s Voice-Record

The host can command a channel to enable voice-recording dynamically. The host issues an API command with a parameter indicating Channel ID. When a channel’s voice-record is enabled, it packetizes its TDM input data and copies it to the host packet interface each frame.