Am79Q02/021/031 Quad Subscriber Line Audio-Processing Circuit (QSLAC™) Devices DISTINCTIVE CHARACTERISTICS Performs the functions of four codec/filters Software programmable: — SLIC input impedance — Transhybrid balance — Transmit and receive gains — Equalization (frequency response) — Digital I/O pins — Programmable debouncing on one input — Time slot assigner — Programmable clock slot and PCM transmit clock edge options Standard microprocessor interface A-law, µ-law, or linear coding Single or Dual PCM ports available — Up to 128 channels (PCLK at 8.192 MHz) per PCM port — Optional supervision on the PCM highway 1.536, 1.544, 2.048, 3.072, 3.088, 4.096, 6.144, 6.176, or 8.192 MHz master clock derived from MCLK or PCLK Built-in test modes with loopback, tone generation, and µP access to PCM data Low-power, 5.0 V CMOS technology 5.0 V only operation Mixed state (analog and digital) impedance scaling Performance characteristics guaranteed over a 12 dB gain range Real Time Data register with interrupt (open drain or TTL output) Supports multiplexed SLIC inputs Broadcast state 256 kHz or 293 kHz chopper clock for Legerity SLICs with switching regulator Maximum channel bandwidth for V.34 modems GENERAL DESCRIPTION The Am79Q02/021/031 Quad Subscriber Line AudioProcessing Circuit (QSLAC) devices integrate the key functions of analog linecards into high-performance, very-programmable, four-channel codec-filter devices. The QSLAC devices are based on the proven design of Legerity’s reliable SLAC™ device families. The advanced ar chitecture of the Q SLAC devic es implements four independent channels and employs digital filters to allow software control of transmission, thus providing a cost-effective solution for the audioprocessing function of programmable linecards. Advanced submicron CMOS technology makes the Am79Q02/021/031 QSLAC devices economical, with both the functionality and the low power consumption needed in linecard designs to maximize linecard density at minimum cost. When used with four Legerity SLICs, a QSLAC device provides a complete softwareconfigurable solution to the BORSCHT functions. Publication# 080147 Rev: H Amendment: /0 Issue Date: September 2001 TABLE OF CONTENTS Distinctive Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 General Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Ordering Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Connection Diagrams. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Pin Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Absolute Maximum Ratings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Operating Ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Electrical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Transmission Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Attenuation Distortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Group Delay Distortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Variation of Gain with Input Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Total Distortion, Including Quantizing Distortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Discrimination against Out-of-Band Input Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Discrimination against 12- and 16 kHz Metering Signals . . . . . . . . . . . . . . . . . . . . . . . . . 18 Spurious Out-of-Band Signals at the Analog Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Overload Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Switching Characteristics (PCM/MPI Mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Microprocessor Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 PCM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Master Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Auxiliary Output Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Switching Waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Input and Output Waveforms for AC Tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Master Clock Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Microprocessor Interface (Input Mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Microprocessor Interface (Output Mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 PCM Highway Timing for XE = 0 (Transmit on Negative PCLK Edge) . . . . . . . . . . . . . . . 23 PCM Highway Timing for XE = 1 (Transmit on Positive PCLK Edge) . . . . . . . . . . . . . . . . 24 Operating the QSLAC Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Power-Up Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Channel Enable Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 SLIC Control and Data Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Clock Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 E1 Multiplex Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Debounce Filters Operation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Real-Time Data Register Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Interrupt Mask Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Active State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Inactive State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Low Power State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Chopper Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Reset States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Signal Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Overview of Digital Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Two-Wire Impedance Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Frequency Response Correction and Equalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Transhybrid Balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Gain Adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2 Am79Q02/021/031 Data Sheet Transmit Signal Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transmit PCM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Receive Signal Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Receive PCM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analog Impedance Scaling Network (AISN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Speech Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signaling on the PCM Highway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robbed-Bit Signaling Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Default Filter Coefficients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Command Description and Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microprocessor Interface Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of MPI Commands* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MPI Command Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Programmable Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Description of CSD Coefficients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User Test States and Operating Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-Law and µ-Law Companding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Physical Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Revision Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 33 34 34 34 35 35 35 36 36 36 38 39 56 56 57 58 60 61 64 LIST OF FIGURES Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Attenuation Distortion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Group Delay Distortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-Law/µ-Law Gain Tracking with Tone Input (Both Paths) . . . . . . . . . . . . . . . . A-Law/µ-Law Total Distortion with Tone Input (Both Paths) . . . . . . . . . . . . . . . Discrimination Against Out-of-Band Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . Spurious Out-of-Band Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A/A Overload Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clock Mode Option. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SLIC I/O, E1 Multiplex, and Real-Time Data Register Operation . . . . . . . . . . . E1 Multiplex Internal Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MPI Real-Time Data Register or GCI Upstream SC Channel Data . . . . . . . . . . QSLAC Device Block Diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Robbed-Bit Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 14 15 16 17 18 19 26 28 29 30 32 36 LIST OF TABLES Table 1 Table 2 Table 3 dBm0 Voltage Definitions with Unity Gain in X, R, GX, GR, AX, and AR . . . . . 13 A-Law: Positive Input Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 µ-Law: Positive Input Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 SLAC Products 3 BLOCK DIAGRAM Quad SLAC Device Dual/Single PCM Highway Analog VIN1 VOUT1 VIN2 VOUT2 VIN3 VOUT3 VIN4 VOUT4 DXA Signal Processing Channel 1 (CH 1) DRA Signal Processing Channel 2 (CH 2) Time Slot Assigner (TSA) TSCA DXB DRB Signal Processing Channel 3 (CH 3) TSCB Signal Processing Channel 4 (CH 4) VREF SLIC Clock & Reference Circuits CD11 CD21 C31 C41 C51 FS PCLK MCLK/E1 CD12 CD22 C32 C42 C52 CD13 CD23 C33 C43 C53 CD14 CD24 C34 C44 C54 CHCLK SLIC Interface (SLI) Microprocessor Interface (MPI) INT CS RST DIO DCLK Microprocessor 19256A-001 4 Am79Q02/021/031 Data Sheet ORDERING INFORMATION Standard Products Legerity standard products are available in several packages and operating ranges. The order number (Valid Combination) is formed by a combination of the elements below. Am79Q02/021/031 J C TEMPERATURE RANGE * C = Commercial (0°C to 70°C; Relative Humidity = 15% to 95%) PACKAGE TYPE J = 44-Pin Plastic Leaded Chip Carrier (PL 044) —Am79Q02/021 Only 32-Pin Plastic Leaded Chip Carrier (PL 032) —Am79Q031 V = 44-Pin Thin Quad Flat Pack (PQT 044) —Am79Q021 Only DEVICE NUMBER/DESCRIPTION Am79Q02/021/031 Quad Subscriber Line Audio-Processing Circuit (QSLAC) Device Valid Combinations Am79Q02 JC Am79Q021 JC Am79Q031 JC Am79Q021 VC Valid Combinations Valid Combinations list configurations planned to be supported in volume for this device. Consult the local Legerity sales office to confirm availability of specific valid combinations, to check on newly released combinations, and to obtain additional data on Legerity’s standard military–grade products. Note: * Functionality of the device from 0°C to +70°C is guaranteed by production testing. Performance from –40°C to +85°C is guaranteed by characterization and periodic sampling of production units. SLAC Products 5 CHCLK 5 4 3 2 1 44 43 42 41 40 CS 6 C31 C41 CD22 C32 C42 CD21 CD12 CD11 44-Pin PLCC MCLK/E1 CONNECTION DIAGRAMS (PLCC PACKAGES) Top View VOUT1 7 39 DCLK VIN1 VOUT2 8 38 DIO 9 37 TSCA VIN2 VCCA VREF 10 36 TSCB 35 DGND 34 PCLK AGND 13 33 VIN3 VOUT3 14 32 VCCD DXA 15 31 VIN4 16 30 DXB FS VOUT4 17 29 RST 11 Am79Q02JC 12 DRA INT DRB C34 C44 CD14 CD24 C43 C33 CD23 CD13 18 19 20 21 22 23 24 25 26 27 28 19256A-003 DCLK DIO VIN2 10 36 TSCA VCCA 11 35 DGND 34 PCLK Am79Q021JC AGND VIN3 13 14 33 VCCD 32 DXA VOUT3 15 31 FS VIN4 16 17 30 RST 29 INT VOUT2 MCLK/E1 CS DCLK 6 28 DIO VIN2 7 27 TSCA VCCA 8 26 DGND VREF 9 C44 C54 DRA C34 CD24 C53 CD14 Am79Q031JC 25 PCLK AGND 10 24 VCCD VIN3 11 23 DXA VOUT3 12 22 FS VIN4 13 21 RST 18 19 20 21 22 23 24 25 26 27 28 CD23 C33 C43 CD21 29 19256A-005 14 15 16 17 18 19 20 Notes: 1. Pin 1 is marked for orientation. 2. RSVD = Reserved pin; should not be connected externally to any signal or supply. Am79Q02/021/031 Data Sheet INT 37 CD13 1 32 31 30 4 3 DRA 9 6 CD11 38 5 CD14 CD24 8 VOUT4 CD22 VIN1 VOUT2 VIN1 CD23 CS CD13 39 VOUT4 7 12 2 2 1 44 43 42 41 40 VOUT1 VREF CD12 VOUT1 MCLK/E1 C51 32-Pin PLCC CD21 C31 C41 CD11 5 4 3 C52 CD22 C32 6 C42 CD12 44-Pin PLCC 19256A-004 CONNECTION DIAGRAM (TQFP PACKAGE) Top View MCLK/E1 C51 CD21 C31 C41 CD11 C52 CD22 C32 C42 CD12 44-Pin TQFP 44 43 42 41 40 39 38 37 36 35 34 VOUT1 1 33 CS VIN1 VOUT2 2 32 3 31 DCLK DIO VIN2 4 30 TSCA VCCA 5 29 DGND VREF 6 28 PCLK AGND VIN3 7 8 27 VCCD 26 DXA VOUT3 9 25 FS 10 11 24 RST 23 INT VIN4 VOUT4 Am79Q021VC C44 C54 DRA C34 CD24 CD14 C53 CD23 C33 C43 CD13 12 13 14 15 16 17 18 19 20 21 22 19256A-023 Notes: 1. Pin 1 is marked for orientation. 2. RSVD = Reserved pin; should not be connected externally to any signal or supply. SLAC Products 7 PIN DESCRIPTIONS Pin Names Type Description CD11–CD14, CD21–CD24 Inputs/Outputs Control and Data. CD1 and CD2 are TTL compatible programmable Input or Output (I/O) ports. They can be used to monitor or control the state of SLIC or any other device associated with subscriber line interface. The direction, input or output, is programmed using MPI Command 22. As outputs, CD1 and CD2 can be used to control relays, illuminate LEDs, or perform any other function requiring a latched TTL compatible signal for control. The output state of CD1 and CD2 is written using MPI Command 20. As inputs, CD1 and CD2 can be processed by the QSLAC device (if programmed to do so). CD1 can be debounced before it is made available to the system. The debounce time is programmable from 0 to 15 ms in 1 ms increments using MPI Command 45. CD2 can be filtered using the up/down counter facility and programming the sampling interval using MPI Command 52. Additionally, CD1 can be demultiplexed into two separate inputs using the E1 demultiplexing function. The E1 demultiplexing function of the QSLAC device was designed to interface directly to Legerity SLICS supporting the ground key function. With the proper Legerity SLIC and the E1 function of the QSLAC enabled, the CD1 bit can be demultiplexed into an Off-Hook/ Ring Trip signal and Ground Key signal. In the demultiplex mode, the second bit, Ground Key, takes the place of the CD2 as an input. The demultiplexed bits can be debounced (CD1) or filtered (CD2) as explained previously. A more complete description of CD1, CD2, debouncing, and filtering functions is contained in the Operating the QSLAC Device section on page 25. Once the CD1 and CD2 inputs are processed (Debounced, Filtered and/or Demultiplexed) by the QSLAC device, the information can be accessed by the system in two ways: 1) on a per channel basis along with C3, C4, and C5 of the specific channel using MPI Command 21, or 2) by using MPI Commands 16 and 17, which obtain the CD1 and CD2 bits from all four channels simultaneously. This feature reduces the processor overhead and the time required to retrieve time-critical signals from the line circuits, such as off-hook and ring trip. With this feature, hookswitch status and ring trip information, for example, can be obtained from all four channels of a QSLAC device with one read command. C31–C34, C41–C44, C51–C54 Inputs/Outputs Control. C3, C4, and C5 are TTL-compatible programmable Input or Output (I/O) ports. They can be used to monitor or control the state of SLIC or any other device associated with subscriber line interface. The direction, input or output, is programmed using MPI Command 22. As outputs, C3, C4, and C5 can be used to control relays, illuminate LEDs, or perform any other function requiring a latched TTL compatible signal for control. The output state of C3, C4, and C5 is written using MPI Command 20. As inputs, C3, C4, and C5 can be accessed by the system by using MPI Command 21. The Am79Q021 QSLAC device contains a single PCM highway and five programmable I/Os per channel (CD1, CD2, C3, C4, and C5) in a 44-pin PLCC or TQFP package. In the Am79Q02 QSLAC device, the C51, C52, C53, and C54 I/Os are eliminated, enabling dual PCM highways and a chopper clock output in a 44-pin PLCC or TQFP package. In the Am79Q031 QSLAC device, the C31–C51, C32–C52, C33–C53, and C34–C54 I/Os are eliminated, enabling a single PCM highway and two control and data I/Os (CD1, CD2) per channel in a 32-pin PLCC package. 8 CHCLK Output Chopper Clock. This output provides a 256 kHz or a 292.57 kHz, 50% duty cycle, TTL-compatible clock for use by up to four SLICs with built-in switching regulators. The CHCLK frequency is synchronous to MCLK, but the phase relationship to MCLK is random. The chopper clock is not available in all package types. CS Input Chip Select. The Chip Select input (active Low) enables the device so that control data can be written to or read from the part. The channels selected for the write or read operation are enabled by writing 1 s to the appropriate bits in the Channel Enable Register of the QSLAC device prior to the command. See EC1, EC2, EC3, and EC4 of the Command 14, page 42, for more information. If Chip Select is held Low for 16 rising edges of DCLK, a hardware reset is executed when Chip Select returns High. Am79Q02/021/031 Data Sheet Pin Names Type Description DCLK Input Data Clock. The Data Clock input shifts data into and out of the microprocessor interface of the QSLAC device. The maximum clock rate is 4.096 MHz. DIO Input/Output Data. Control data is serially written into and read out of the QSLAC device via the DIO pin, with the most significant bit first. The Data Clock determines the data rate. DIO is high impedance except when data is being transmitted from the QSLAC device. DRA/DRB Inputs PCM Data Receive A/B. The PCM data for channels 1, 2, 3, and 4 is serially received on either the DRA or DRB port during user-programmed time slots. Data is always received with the most significant bit first. For compressed signals, 1 byte of data for each channel is received every 125 µs at the PCLK rate. In the Linear state, two consecutive bytes of data for each channel are received every 125 µs at the PCLK rate. DRB is not available on all package types. DXA/DXB Outputs PCM Data Transmit. The transmit data from channels 1, 2, 3, and 4 is sent serially out on either the DXA or DXB port or both ports during user-programmed time slots. Data is always transmitted with the most significant bit first. The output is available every 125 µs and the data is shifted out in 8-bit (16-bit in Linear or PCM Signaling state) bursts at the PCLK rate. DXA and DXB are High impedance between time slots, while the device is in the Inactive state with no PCM signaling, or while the Cutoff Transmit Path bit (CTP) is on. DXB is not available on all package types. FS Input Frame Sync. The Frame Sync pulse is an 8 kHz signal that identifies Time Slot 0, Clock Slot 0 of a system’s PCM frame. The QSLAC device references individual time slots with respect to this input, which must be synchronized to PCLK. INT Output Interrupt. INT is an active Low output signal which is programmable as either TTL compatible or open drain. The INT output goes Low any time one of the input bits in the Real Time Data register changes state and is not masked. It also goes Low any time new transmit data appears if this interrupt is armed. INT remains Low until the appropriate register is read via the microprocessor interface, or the QSLAC device receives either a software or hardware reset. The individual CDxy bits in the Real Time Data register can be masked from causing an interrupt by using Command 26 of the MPI. The transmit data interrupt must be armed with a bit in the Operating Conditions register. MCLK/E1 Input/Output Master Clock (Input)/Enable CD1 Multiplex (Output). The Master Clock can be a 1.536 MHz, 1.544 MHz, or 2.048 MHz (times 1, 2, or 4) clock for use by the digital signal processor. If the internal clock is derived from the PCM Clock Input (PCLK), this pin can be used as an E1 output to control Legerity SLICs having multiplexed hookswitch and ground-key detector outputs. PCLK Input PCM Clock. The PCM clock determines the rate at which PCM data is serially shifted into or out of the PCM ports. PCLK is an integer multiple of the frame sync frequency. The maximum clock frequency is 8.192 MHz and the minimum clock frequency is 128 kHz for dual PCM highway versions and 256 kHz for single PCM highway versions. The minimum clock rate must be doubled if Linear state or PCM signaling is used. PCLK frequencies between 1.03 MHz and 1.53 MHz are not allowed. Optionally, the digital signal processor clock can be derived from PCLK rather than MCLK. RST Input Reset. A logic Low signal at this pin resets the QSLAC device to its default state. The RST pin may be tied to VCCD if it is not needed in the system. TSCA, TSCB Outputs Time Slot Control. The Time Slot Control outputs are open drain outputs (requiring pull-up resistors to VCCD) and are normally inactive (High impedance). TSCA or TSCB is active (Low) when PCM data is transmitted on the DXA or DXB pin respectively. VIN1–VIN4 Inputs Analog. The analog voice band signal is applied to the VIN input of the QSLAC device. The VIN input is biased at VREF by a large internal resistor. The audio signal is sampled, digitally processed and encoded, and then made available at the TTL-compatible PCM output (DXA or DXB). If the digitizer saturates in the positive or negative direction, VIN is pulled by a reduced resistance toward AGND or VCCD, respectively. VIN1 is the input for channel 1, VIN2 is the input for channel 2, VIN3 is the input for channel 3, and VIN4 is the input for channel 4. SLAC Products 9 Pin Names Type Description VOUT1– VOUT4 Outputs Analog. The received digital data at DRA or DRB is processed and converted to an analog signal at the VOUT pin. VOUT1 is the output from channel 1, VOUT2 is the output for channel 2, VOUT3 is the output from channel 3, and VOUT4 is the output for channel 4. The VOUT voltages are referenced to VREF. VREF Output Analog Voltage Reference. The VREF output is provided in order for an external 0.1 µF capacitor to be connected from VREF to ground, filtering noise present on the internal voltage reference. VREF is buffered before it is used by internal circuitry. The voltage on VREF is nominally 2.1 V, and the output resistance is 100 kΩ ±30%. The leakage current in the capacitor must be less than 20 nA. Power Supply AGND Analog ground DGND Digital ground VCCA +5.0 V analog power supply VCCD +5.0 V digital power supply bits are High, all channels enabled will receive the programming infor mation wr itten; therefore, a Broadcast state can be implemented by simply enabling all channels in the device to receive the information. The Channel Enable bits are contained in the Channel Enable register, which is written and read using Commands 14 and 15. The Broadcast state is useful in initializing QSLAC devices in a large system. Two separate power supply inputs are provided to allow for noise isolation and proper power supply decoupling techniques; however, the two pins have a low impedance connection inside the part. For best performance, all of the +5.0 power supply pins should be connected together at the connector of the printed circuit board, and all of the grounds should be connected together at the connector of the printed circuit board. The user-programmable filters set the receive and transmit gain, perform the transhybrid balancing function, permit adjustment of the two-wire termination impedance, and provide equalization of the receive and transmit paths. All programmable digital filter coefficients can be calculated using the AmSLAC4 or WinSLAC™ software. Data transmitted or received on the PCM highway can be 8-bit companded code (with an optional 8-bit signaling byte in the transmit direction) or 16-bit linear code. The 8-bit codes appear 1 byte per time slot, while the 16-bit code appears in two consecutive time slots. The compressed PCM codes can be either 8-bit companded A-law or µ-law. The PCM data is read from and written to the PCM highway in user-programmable time slots at rates of 128 kHz to 8.192 MHz. The transmit clock edge and clock slot can be selected for compatibility with other devices that can be connected to the PCM highway. FUNCTIONAL DESCRIPTION The QSLAC device performs the codec/filter and twoto four-wire conversion functions required of the s u b s c r i b e r l i n e i n t e r fa c e c i r c u i t r y in telecommunications equipment. These functions involve converting audio signals into digital PCM samples and converting digital PCM samples back into audio signals. During conversion, digital filters are used to band limit the voice signals. All of the digital filtering is performed in digital signal processors operating from a master clock, which can be derived either from PCLK or MCLK. Three configurations of the QSLAC device are offered with single or dual PCM highways. The Am79Q02 and Am79Q021 QSLAC devices with dual and single PCM highways respectively are available in the 44-pin packages. The Am79Q031JC QSLAC device is a single PCM highway version in a 32-pin PLCC package. Four independent channels allow the QSLAC device to function as four SLAC devices or two DSLAC ä devices. For programming information, each channel has its own enable bit (EC1, EC2, EC3, and EC4) to allow individual channel programming. If more than one Channel Enable bit is High or if all Channel Enable 10 PCM Highway Programmable I/O Chopper Clock Dual Four Yes 44 PLCC/TQFP Package Am79Q02 JC Single Five No 44 PLCC/TQFP Am79Q021 JC (or VC) Single Two No 32 PLCC Am79Q031 JC Am79Q02/021/031 Data Sheet Part Number ABSOLUTE MAXIMUM RATINGS OPERATING RANGES Storage Temperature . . . . . . . . –60°C < TA < +125°C VCCA, Analog Supply . . . . . . . . . . . . . +5.0 V ±0.25 V Ambient Operating Temperature –40°C < TA < +85°C VCCA, Analog Supply . . . . . . . . . . . . . . VCCD ±10 mV Ambient Relative Humidity . . . . . . . . . . . . 5% to 95% VCCD, Digital Supply . . . . . . . . . . . . . +5.0 V ±0.25 V (non-condensing) DGND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 V VCCA with respect to AGND . . . . . . . .–0.4 V to +7.0 V AGND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ±50 mV VCCA with respect to VCCD . . . . . . . . . . . . . . ±50 mV Ambient Temperature . . . . . . . . . . . 0°C < TA < +70°C VCCD with respect to DGND. . . . . . . .–0.4 V to +7.0 V Ambient Relative Humidity . . . . . . . . . . . 15% to 95% VIN with respect to AGND . . .–0.4 V to (VCCA +0.4 V) Operating Ranges define those limits between which functionality of the device is guaranteed by production testing. AGND with respect to DGND . . . . . . . . . . . . . . ±0.4 V Other pins with respect to DGND . . . . . –0.4 V to VCCD +0.4 V Total combined CD1–C5 current per device: Functionality of the device from 0°C to +70°C is guaranteed by production testing. Performance from –40°C to +85°C is guaranteed by characterization and periodic sampling of production units. Source from VCCD . . . . . . . . . . . . . . . . . . . . . 40 mA Sink into DGND . . . . . . . . . . . . . . . . . . . . . . . . 40 mA Latch-up immunity (any pin). . . . . . . . . . . . . ±100 mA Stresses above those listed under Absolute Maximum Ratings may cause permanent device failure. Functionality at or above these limits is not implied. Exposure to Absolute Maximum Ratings for extended periods may affect device reliability. SLAC Products 11 ELECTRICAL CHARACTERISTICS Typical values are for TA = 25 °C and nominal supply voltages. Minimum and maximum values are over the temperature and supply voltage ranges shown in Operating Ranges. Symbol VIL VIH IIL VOL VOH IOL VIR VIOS ZIN IIP IIN ZOUT IOUT ZREF VOR VOOS VOOSA LINAISN PD CI CO PSRR Parameter Descriptions Input Low voltage Input High voltage Input leakage current Output Low voltage CD1–C5 (IOL = 4 mA) CD1–C5 (IOL = 8 mA) TSCA, TSCB (IOL =14 mA) Other digital outputs (IOL = 2 mA) Output High voltage CD1–C5 (IOH = 4 mA) CD1–C5 (IOH = 8 mA) Other digital outputs (IOH = 400 µA) Output leakage current (HI = Z state) Analog input voltage range(AX = 0 dB) (Relative to VREF)(AX = 6.02 dB) Offset voltage allowed on VIN Analog input impedance to VREF300 to 3400 Hz Current into analog input for input voltages between 3.8 V and 5.0 V Current out of analog input for input voltages between 0 V and 0.5 V VOUT output impedance VOUT output current (F< 3400 Hz) VREF output impedance (F < 3400 Hz) VOUT voltage range(AR = 0 dB) (Relative to VREF)(AR = 6.02 dB) VOUT offset voltage (AISN off) VOUT offset voltage (AISN on) Linearity of AISN circuitry (input = 0 dBm0) Power dissipation All channels active 1 channel active All channels inactive, (in normal state) All channels inactive (in low power state) Input capacitance (Digital) Output capacitance (Digital) Power supply rejection ratio (1.02 kHz, 100 mVRMS, either path, GX = GR = 0 dB) Min Typ 2.0 –10 Max 0.8 Unit +10 µA 0.4 0.8 0.4 0.4 Note V 1 V VCCD – 0.4 V VCCD – 0.8 V 1 2.4 –10 10 ±1.584 ±0.792 µA Vpk –50 0.43 50 3.4 54 170 50 170 mV MΩ 2 µA 1 –4 70 10 4 130 ±1.584 ±0.792 –40 –80 –0.25 40 Ω mApk kΩ 3 Vpk 40 80 0.25 200 70 18 6 15 15 2 260 130 25 12 mV 4 LSB mW 5 pF dB Notes: 1. The CD1, CD2, C3–C5 outputs are resistive for less than a 0.8 V drop. Total current must not exceed absolute maximum ratings. 2. When the digitizer saturates, a resistor of 50 kΩ ±20 kΩ is connected either to DGND or to VCCD — (1 diode drop) as appropriate to discharge the coupling capacitor. 3. When the QSLAC device is in the Inactive state, the analog output will present either a VREF DC output level through a 15 kΩ resistor (VMODE = 0) or a high impedance (VMODE = 1). 4. If there is an external DC path from VOUT to VIN with a gain of GDC and the AISN has a gain of hAISN, then the output offset will be multiplied by 1/[1–(hAISN • GDC)]. 5. Power dissipation in the Inactive state is measured with all digital inputs at VIH = VCC and VIL = DGND and with no load connected to VOUT1, VOUT2, VOUT3, or VOUT4. 12 Am79Q02/021/031 Data Sheet Transmission Characteristics Table 1. 0 dBm0 Voltage Definitions with Unity Gain in X, R, GX, GR, AX, and AR Signal at Digital Interface Transmit Receive A-law digital mW or equivalent (0 dBm0) 0.7804 0.7804 µ-law digital mW or equivalent (0 dBm0) 0.7746 0.7746 ±22,827 peak linear coded sine wave 0.7804 0.7804 Unit Vrms When relative levels (dBm0) are used in any of the following transmission specifications, the specification holds for any setting of the GX gain from 0 dB to 12 dB and the GR loss from 0 dB to 12 dB. Description Gain accuracy, D/A or A/D Test Conditions 0 dBm0, 1014 Hz AX = AR = 0 dB 0 to 85°C –40°C AX = +6.02 dB and/or AR = –6.02 dB 0 to 85°C –40°C Gain accuracy digital-to-digital Gain accuracy analog-to-analog Attenuation distortion Single frequency distortion Idle channel noise Analog out 300 Hz to 3 kHz weighted unweighted Digital input = 0 A-law Digital input = 0 µ-law Analog VIN = 0 VAC A-law Analog VIN = 0 VAC µ-law 0 dBm0 300 Hz to 3400 Hz 0 dBm0 300 Hz to 3400 Hz 0 dBm0 1014 Hz, Average 1014 Hz, Average B = Z = 0; X = R = 1 Min Typ CrosstalkTX to RX same channelRX to TX Crosstalk between channels TX or RX to TX TX or RX to RX End-to-end group delay Unit Note –0.25 –0.30 +0.25 +0.30 –0.30 –0.40 –0.25 –0.25 –0.125 +0.30 +0.40 +0.25 +0.25 +0.125 –46 dB –68 –55 –78 12 –68 16 –75 –75 dBm0p dBm0 dBm0p dBrnc0 dBm0p dBrnc0 –76 –78 678 dBm0 4 µs 5 Digital looped back Digital out Max 0 0 1 2 3 3 3 3, 6 3 3, 6 dBm0 Notes: 1. Also see Figure 1 and Figure 2. 2. 0 dBm0 input signal, 300 Hz to 3400 Hz; measurement at any other frequency, 300 Hz to 3400 Hz. 3. No single frequency component in the range above 3800 Hz may exceed a level of –55 dBm0. 4. The weighted average of the crosstalk is defined by the following equation, where C(f) is the crosstalk in dB as a function of frequency, fN = 3300 Hz, f1 = 300 Hz, and the frequency points (fj , j = 2..N) are closely spaced: 1----• C ( fj – 1 ) 20 æ fj ö + 10 10 -÷ --------------------------------------------------------- • log ç -------è f j – 1ø 2 j --------------------------------------------------------------------------------------------------æf ö log ç ----N-÷ è f1 ø å Average = 20 • log 1----• C ( fj ) 20 5. The End-to-End Group Delay is the sum of the transmit and receive group delays (both measured using the same time and clock slot). 6. Typical values not tested in production. SLAC Products 13 Attenuation Distortion QSLAC Device Specification 2 Transmit curve 1.8 dB Attenuation (dB) Receive curve 1 dB 1 0.75 dB 0.125 0 Transmit only – 0.125 200 Frequency (Hz) 300 3000 3400 19256A-006 Figure 1. Attenuation Distortion Group Delay Distortion For either transmission path, the group delay distortion is within the limits shown in Figure 2. The minimum value of the group delay is taken as the reference. The signal level should be 0 dBm0. 420 QSLAC Device Specification (Either Path) Delay (µs) 150 90 0 500 600 1000 2600 Frequency (Hz) Figure 2. Group Delay Distortion 14 Am79Q02/021/031 Data Sheet 2800 19256A-007 Variation of Gain with Input Level The gain deviation relative to the gain at –10 dBm0 is within the limits shown in Figure 3 for either transmission path when the input is a sine wave signal of frequency 1014 Hz. QSLAC Device Specification 1.5 0.55 0.25 Gain dB 0 –55 –50 –40 –10 0 +3 Input Level dBm0 –0.25 –0.55 –1.5 19256A-008 a. A-law QSLAC Device Specification 1.4 0.45 0.25 Gain dB 0 –55 –50 –37 –10 0 +3 Input Level dBm0 –0.25 –0.45 –1.4 19256A-009 b. µ-law Figure 3. A-law/µ-law Gain Tracking with Tone Input (Both Paths) SLAC Products 15 Total Distortion, Including Quantizing Distortion The signal-to-total distortion will exceed the limits shown in Figure 4 for either transmission path when the input is a sine wave signal of frequency 1014 Hz. QSLAC Device Specification 35.5 Signal-to-Total Distortion (dB) 35.5 30 25 –45 –40 –30 0 +3 Input Level (dBm0) 19256A-010 a. A-law QSLAC Device Specification 35.5 Signal-to-Total Distortion (dB) 35.5 31 27 –45 –40 –30 0 +3 Input Level (dBm0) 19256A-011 b. µ-law Figure 4. A-law/µ-law Total Distortion with Tone Input (Both Paths) 16 Am79Q02/021/031 Data Sheet Discrimination against Out-of-Band Input Signals When an out-of-band sine wave signal with frequency and level A is applied to the analog input, there may be frequency components below 4 kHz at the digital output which are caused by the out-of-band signal. These components are at least the specified dB level below the level of a signal at the same output originating from a 1014 Hz sine wave signal with a level of A dBm0 also applied to the analog input. The minimum specifications are shown in the following table. Frequency of Out-of-Band Signal Amplitude of Out-of-Band Signal Level below A 16.6 Hz < f < 45 Hz –25 dBm0 < A ≤ 0 dBm0 18 dB 45 Hz < f < 65 Hz –25 dBm0 < A ≤ 0 dBm0 25 dB 65 Hz < f < 100 Hz –25 dBm0 < A ≤ 0 dBm0 10 dB 3400 Hz < f < 4600 Hz –25 dBm0 < A ≤ 0 dBm0 see Figure 5 4600 Hz < f < 100 kHz –25 dBm0 < A ≤ 0 dBm0 32 dB 0 QSLAC Device Specification –10 –20 Level (dB) –28 dBm –30 –32 dB, –25 dBm0 < input < 0 dBm0 –40 –50 3.4 4.0 4.6 19256A-012 Frequency (kHz) Note: The attenuation of the waveform below amplitude A between 3400 Hz and 4600 Hz is given by the formula: π ( 4000 – f ) Attenuation (db) = 14 – 14 sin -------------------------1200 Figure 5. Discrimination Against Out-of-Band Signals SLAC Products 17 Discrimination against 12- and 16-kHz Metering Signals If the QSLAC device is used in a metering application where 12-kHz or 16-kHz tone bursts are injected onto the telephone line toward the subscriber, a portion of those tones may also appear at the VIN terminal. These out-of-band signals may cause frequency components to appear below 4 kHz at the digital output. For a 12 kHz or 16 kHz tone, the frequency components below 4 kHz will be reduced from the input by at least 70 dB. The sum of the peak metering and signal voltages must be within the analog input voltage range. Spurious Out-of-Band Signals at the Analog Output With PCM code words representing a sine wave signal in the range of 300 Hz to 3400 Hz at a level of 0 dBm0 applied to the digital input, the level of the spurious out-of-band signals at the analog output is less than the limits shown below. Frequency Level 4.6 kHz to 40 kHz –32 dBm0 40 kHz to 240 kHz –46 dBm0 240 kHz to 1 MHz –36 dBm0 With code words representing any sine wave signal in the range 3.4 kHz to 4.0 kHz at a level of 0 dBm0 applied to the digital input, the level of the signals at the analog output are below the limits in Figure 6. The amplitude of the spurious out-of-band signals between 3400 Hz and 4600 Hz is given by the formula: π ( f – 4000 ) A = – 14 – 14 sin ---------------------------- dBm0 1200 0 QSLAC Device Specification –10 –20 Level (dBm0) –28 dB –30 –32 dB –40 –50 3.4 4.0 4.6 Frequency (kHz) 19256A-013 Figure 6. Spurious Out-of-Band Signals 18 Am79Q02/021/031 Data Sheet Overload Compression Figure 7 shows the acceptable region of operation for input signal levels above the reference input power (0 dBm0). The conditions for this figure are: (1) 1.2 dB < GX ≤ 12 dB; (2) –12 dB ≤ GR < –1.2 dB; (3) PCM output connected to PCM input; and (4) measurement analog-to-analog. 9 8 7 6 Fundamental Output Power 5 (dBm0) Acceptable Region 4 3 2.6 2 1 1 2 3 4 5 6 7 8 9 Fundamental Input Power (dBm0) 19256A-014 Figure 7. A/A Overload Compression SLAC Products 19 SWITCHING CHARACTERISTICS over operating range (unless otherwise noted) Min and max values are valid for all digital outputs with a 150 pF load, except CD1–C5 with a 30 pF load. Microprocessor Interface No. Symbol Parameter Min Typ Max 1 tDCY Data clock period 244 2 tDCH Data clock High pulse width 97 3 tDCL Data clock Low pulse width 97 4 tDCR Rise time of clock 25 5 tDCF Fall time of clock 25 6 tICSS Chip select setup time, Input state 70 7 tICSH Chip select hold time, Input state 0 8 tICSL Chip select pulse width, Input state 9 tICSO Chip select off time, Input state 2.5 10 tIDS Input data setup time 30 11 tIDH Input data hold time 30 12 tOLH SLIC output latch valid 13 tOCSS Chip select setup time, Output state 70 t DCY –10 14 tOCSH Chip select hold time, Output state 0 t DCH –20 15 tOCSL Chip select pulse width, Output state 16 tOCSO Chip select off time, Output state 17 tODD Output data turn on delay 18 tODH Output data hold time 19 tODOF Output data turn off delay 20 tODC Output data valid 0 21 tRST Reset pulse width 50 Units Note ns t DCY –10 t DCH –20 8t DCY µs 1000 1 ns 8t DCY µs 2.5 50 0 50 1 2 ns 50 µs PCM Interface PCLK not to exceed 8.192 MHz. Pull-up resistors of 360 Ω are attached to TSCA and TSCB. No. 20 Symbol Parameter Min Typ Max 22 tPCY PCM clock period 122 23 tPCH PCM clock High pulse width 48 24 tPCL PCM clock Low pulse width 48 25 tPCF Fall time of clock 26 tPCR Rise time of clock 27 tFSS FS setup time 25 28 tFSH FS hold time 50 30 tTSD Delay to TSC valid 5 80 31 tTSO Delay to TSC off 5 80 32 tDXD PCM data output delay 5 70 33 tDXH PCM data output hold time 5 70 34 tDXZ PCM data output delay to High-Z 5 70 35 tDRS PCM data input setup time 25 36 tDRH PCM data input hold time 5 Units 3 15 15 Am79Q02/021/031 Data Sheet t PCY –50 ns 4 4,5 6 Master Clock No. Symbol Parameter Min 37 AMCY Master clock accuracy 38 tMCR Rise time of clock 39 tMCF Fall time of clock 40 tMCH MCLK High pulse width 48 41 tMCL MCLK Low pulse width 48 Typ –100 Max Units +100 ppM Note 15 15 ns Auxiliary Output Clocks No. Symbol Parameter Chopper clock frequency Min CHP = 0 CHP = 1 Typ 256 292.57 42 fCHP 43 fE1 E1 output frequency (CMODE = EE1 = 1) 4.923 44 tE1 E1 pulse width (CMODE = EE1 = 1) 31.25 Max Units Note kHz µs Notes: 1. If CFAIL = 1 (Command 23), GX, GR, Z, B1, X, R, and B2 coefficients must not be written or read without first deactivating all channels or switching them to default coefficients; otherwise, a chip select off time of 25 µs is required. If the low power state (LPM = 1, Command 14) is selected and MCLK is also lost, this minimum chip select off time increases to 75 µs. 2. The first data bit is enabled on the falling edge of CS or on the falling edge of DCLK, whichever occurs last. 3. The PCM clock frequency must be an integer multiple of the frame sync frequency. The maximum allowable PCM clock frequency is 8.192 MHz. The actual PCM clock rate is dependent on the number of channels allocated within a frame. The minimum clock frequency is 128 kHz in Companded state and 256 kHz in Linear state, PCM Signaling state. The minimum PCM clock rates should be doubled for parts with only one PCM highway in order to allow simultaneous access to all four channels. 4. TSC is delayed from FS by a typical value of N • tPCY , where N is the value stored in the time/clock-slot register. 5. tTSO is defined as the time at which the output achieves the Open Circuit state. 6. There is a special conflict detection circuitry that will prevent high-power dissipation from occurring when the DXA or DXB pins of two QSLAC devices are tied together and one QSLAC device starts to transmit before the other has gone into a High-impedance state. SWITCHING WAVEFORMS Input and Output Waveforms for AC Tests 2.4 2.0 0.8 0.45 Test Points 2.0 0.8 19256A-015 Master Clock Timing 37 40 VIH VIL 41 39 SLAC Products 38 19256A-016 21 Microprocessor Interface (Input Mode) 1 2 5 VIH VIH DCLK VIL VIL 7 3 9 4 CS 6 8 11 10 Data Valid DI/O Data Valid Data Valid 12 Data Valid Outputs C5–C1 Data Valid 21108-019 Microprocessor Interface (Output Mode) VIH VIL DCLK 14 13 16 CS 15 20 18 17 DI/O Three-State VOH VOL Data Valid 19 Data Valid Data Valid Three-State 21108A-020 22 Am79Q02/021/031 Data Sheet PCM Highway Timing for XE = 0 (Transmit on Negative PCLK Edge) Time Slot Zero Clock Slot Zero 22 26 25 VIH PCLK VIL 23 24 27 28 FS 30 31 TSCA/ TSCB 32 33 34 VOH DXA/DXB First Bit VOL 35 DRA/DRB First Bit VIH Second Bit VIL SLAC Products 36 21108A-021 23 PCM Highway Timing for XE = 1 (Transmit on Positive PCLK Edge) Time Slot Zero Clock Slot Zero 22 26 25 VIH PCLK VIL 23 24 27 28 FS 30 31 TSCA/ TSCB 32 33 34 VOH DXA/DXB First Bit VOL 35 VIH DRA/DRB First Bit 36 Second Bit VIL 21108A-022 24 Am79Q02/021/031 Data Sheet OPERATING THE QSLAC DEVICE The following sections describe the operation of the four independent channels of the QSLAC device. The des cr i pti on is va li d fo r c hann el 1 , 2, 3, or 4; consequently, the channel subscripts have been dropped. For example, VOUT refers to either VOUT1, VOUT2, VOUT3, or VOUT4. Power-Up Sequence T h e r e c o m m e n d e d Q S L AC d ev i c e p ow e r - u p sequence is to apply: 1. VCC and ground multiple channel addressing is accomplished without increasing the number of I/O pins on the device. The Broadcast state can be further enhanced by providing the ability to select many chips at once; however, care must be taken never to enable more than one chip in the Read state. This can lead to an internal bus contention, in which excess power is dissipated. (Bus contention will not damage the device.) Most control commands defined for the DSLAC device are compatible with the QSLAC device, thereby minimizing the impact to existing system software. 2. Signal connections and Low on RST SLIC Control and Data Lines 3. High on RST The QSLAC device has up to five SLIC digital interface pins per channel (CD1–C5). Each of these pins can be programmed as either an input or an output using the I/O Direction register (Commands 22 and 23) (see Figure 9). The output latches can be written with Command 20; however, only those bits programmed as outputs will actually drive the pins. The inputs can be read with Command 21. If a pin is programmed as an output, the data read from it will be the contents of the output latch. It is recommended that any of the SLIC input/output data points, which are to be programmed as outputs, be written to their desired state via Command 21 before writing the data which configures them as outputs with the I/O direction register Command 22. This ensures that when the output is activated, it is already in the correct state, and will prevent unwanted data from being driven from the SLIC output pins. The software initialization should then include: 1. Wait 1 ms. 2. Select master clock frequency and source (Commands 12 and 13). This should turn off the CFAIL bit (Command 23) within 400 µs. While the CFAIL bit is on, normal programming can proceed, but no channels should be activated. 3. Program filter coefficients and other parameters as required. 4. Activate (Command 5). If the power supply (VCCD) falls below approximately 1.0 V, the device is reset and will require complete reprogramming with the above sequence. A reset may be initiated by connection of a logic Low to the RST pin, or if chip select (CS) is held low for 16 rising edges of DCLK, a hardware reset is generated when CS returns high. The RST pin may be tied to VCCD if it is not used in the system. Channel Enable Register A channel enable register has been implemented in the QSLAC device in order to reduce the effor t required to address individual or multiple channels of the QSLAC device. The register is written using MPI Command 14. Each bit of the register is assigned to one unique channel, bit 0 for channel 1, bit 1 for channel 2, bit 2 for channel 3, and bit 3 for channel 4. The channel or channels are enabled when their corresponding enable bits are High. All enabled channels will receive the data written to the QSLAC device. This enables a Broadcast state (all channels enabled) to be implemented simply and efficiently, and Clock Mode Operation The QSLAC device operates with multiple clock signals. The master clock (MCLK) is used for internal timing including operation of the digital signal processing and may be derived from either the MCLK or PCLK source. The allowed frequencies are listed under Commands 12 and 13. The PCM clock (PCLK) is used for PCM timing and is an integer multiple of the frame sync frequency. The internal device clock (MCLK) can be optionally derived from the PCLK source by setting the CMODE bit (bit 4, Commands 12 and 13, 46/47h) to one. In this mode, the MCLK/E1 pin is free to be used as an E1 signal output. Clock mode options and E1 output functions are shown in Figure 8. SLAC Products 25 . MCLK/E1 PCLK (= 0) Time Slot Assigner (= 1) E1 (= 1) (= 0) CMODE (= 1) EE1 ÷N DSP Engine (= 0) CSEL E1 Pulses E1P Notes: 1. CMODE = Command 12, 13 Bit 4 2. CSEL = Command 12, 13 Bits 0–3 3. EE1 = Command 45, 46 Bit 7 4. E1P = Command 45, 46 Bit 6 Figure 8. Clock Mode Option E1 Multiplex Operation The QSLAC device can multiplex input data from the CD1 SLIC I/O pin into two separate status bits per channel (CD1 and CD1B bits in the SLIC Input/Output register, Commands 52/53h, and CDA and CDB bits in the Real Time Data register, Commands 4D/4Fh) using the E1 multiplex mode. This multiplex mode provides the means to accommodate dual detect states when connected to an Legerity SLIC device, which also supports ground-key detection in addition to loop detect. Legerity SLICs that support ground-key detect use their E1 pin as an input to switch the SLIC’s single detector (DET) output between internal loop detect or ground-key detect comparators. Using the E1 multiplex mode, a single QSLAC device can monitor both loop detect and ground-key detect states of all four connected SLICs without additional hardware. Although normally used for ground key detect, this multiplex function can also be used for monitoring other signal states. The E1 multiplex mode is selected by setting the EE1 bit (bit 7, Command C8/C9h) and CMODE bit (bit 4, Command 46/47h) in the QSLAC device. The CMODE bit must be selected (CMODE=1) for the master clock to be derived from PCLK so that the MCLK/E1 pin can be used as an output for the E1 signal. The multiplex mode is then turned on by setting the EE1 bit. With the 26 E1 multiplex mode enabled, the QSLAC device generates the E1 output signal. This signal is a 31.25 µ s (1/32 kHz) duration pulse occurring at a 4.923 kHz (64 kHz/13) rate. The polarity of this E1 output is selected by the E1P bit (bit 6, Command C8/ C9h) allowing this multiplex mode to accommodate all SLICs regardless of their E1 high/low logic definition. Figure 9 shows the SLIC Input/Output register, I/O pins, E1 multiplex hardware operation for one QSLAC device channel. It also shows the operation of the Real Time Register. The QSLAC device E1 output signal connects directly to the E1 inputs of all four connected SLICs and is used by those SLICs to select an internal comparator to route to the SLIC’s DET output. This E1 signal is also used internally by the QSLAC device for controlling the multiplex operation and timing. The CD1 and CD1B bits of the SLIC Input/Output register are isolated from the CD1 pin by transparent latches. When the E1 pulse is off, the CD1 pin data is routed directly to the CD1 bit of the SLIC I/O register and changes to the CD1B bit of that register are disabled by its own latch. When E1 pulses on, the CD1 latch holds the last CD1 state in its register. At the same time, the CD1B latch is enabled, which allows CD1 pin data to be routed directly to the CD1B bit. Am79Q02/021/031 Data Sheet Therefore, during this multiplexing, the CD1 bit always has loop-detect status and the CD1B bit always has ground-key detect status. Thi s multiplexin g state changes a lmost instantaneously within the QSLAC device but the SLIC device may require a slightly longer time period to respond to this detect state change before its DET output settles and becomes valid. To accommodate this delay difference, the internal signals within the QSLAC device are isolated by 15.625 µ s before allowing any change to the CD1 bit and CD1B bit latches. This operation is further described by the E1 multiplex timing diagram in Figure 9. In this timing diagram, the E1 signal represents the actual signal presented to the E1 output pin. The GK Enable pulse allows CD1 pin data to be routed through the CD1B latch. The LD Enable pulse allows CD1 pin data to be routed through the CD1 latch. The uncertain states of the SLIC’s DET output, and the masked times where that DET data is ignored are shown in this timing diagram. Using this isolation of masked times, the CD1 and CD1B registers are guaranteed to contain accurate representations of the SLIC detector output. SLAC Products 27 SLIC I/O Register MPI Command 20, 21 D — Q — CD1B C5 C4 C3 CD2 CD1 EN/HOLD * CD1 CD2 C3 C4 C5 D I/O Direction Register MPI Command 22 Q EN/HOLD * Output Latch 1 Ground Key Filter (time set via Commands 52, 53) GK Enable Debounce (time set via Commands 45, 46) (Channel 1 Shown) EE1 Bit E1 Source (Internal) { Delay Same for Channels 2, 3, 4 See Figure 10 for details Real Time Data Register (Command 16, 17) E1P INT MUX LD Enable SLIC Output Data Register MPI Command 20 MCLK/E1 0 CDB4 CDA4 CDB3 CDA3 CDB2 CDA2 CDB1 CDA1 ATI (CMD 28, 29, Bit 3) Interrupt Mask Register (Command 26, 27) MCDB4 MCDA4 MCDB3 MCDA3 MCDB2 MCDA2 MCDB1 MCDA1 Note: * Transparent latches: When enable input is high, Q output follows D input. When enable input goes low, Q output is latched at last state. Figure 9. SLIC I/O, E1 Multiplex and Real-Time Data Register Operation 28 Am79Q02/021/031 Data Sheet Pulse Period 203.125 µs 4.923 kHz (64 kHz/13) pulse rate 31.25 µs E1 15.625 µs 15.625 µs GK Enable LD Enable 15.625 µs DET Output from SLIC (CD1 Pin Input) CD1 Pin Input Data Contains Valid LD Status CD1 Register Operation Tracks DET State CD1B Register Operation Contains Valid GK Status CD1 Pin State Ignored CD1 Pin State Ignored Tracks DET State Hold Last State Hold Last State Tracks DET State Contains Valid LD Status Hold Last State Figure 10. E1 Multiplex Internal Timing Debounce Filters Operation Each channel is equipped with two debounce filter circuits to buffer the logic status of the CD1 and CD2/ CD1B bits of the SLIC I/O Data Register (Commands 20 and 21, 52/53h) before providing filtered bit’s outputs to the Real-Time Data Register (Commands 16 and 17, 4D/4Fh). One filter is used only for the CD1 bit. The other filter acts upon either the CD1B bit if E1 multiplexing is enabled, or on the CD2 bit if the multiplexing is not enabled. The CD1 bit normally contains SLIC loop detect status. The CD1 debouncing time is programmable with the Debounce Time Register (Commands 45 and 46, C8/C9h), and even though each channel has its own filter, the programmed value is common to all four channels. This debounce filter is initially clocked at the frame sync rate of 125 µ s, and any occurrence of changing data at this sample rate resets a programmable counter. This programmable counter is clocked at a 1 ms rate, and the programmed count value of 0 to 15 ms, as defined by the Debounce Time Register, must be reached before updating the CDA bit of the Real Time Data register with the CD1 state. Refer to Figure 11a for this filter’s operation. The ground-key filter (Figure 11b) provides a buffering of the signal, normally ground key detect, which appears in the CDB bit of the Real Time Data Register. Each channel has its own filter, and each filter’s time can be individually programmed. The input to the filter comes from either the CD2 bit of the SLIC I/O Data Register (Command 20 and 21, 52/53h), when E1 multiplexing is not enabled, or from the CD1B bit of that register when E1 multiplexing is enabled. The feature debounces ground-key signals before passing them to the Real Time Data Register, although signals other than ground-key status can be routed to the CD2 pin and then through the registers. The ground-key debounce filter operates as a dutycycle detector and consists of an up/down counter which can range in value between 0 and 6. This six-state counter is clocked by the GK timer at the sampling period of 1–15 ms, as programmed by the value of the four GK bits (GK3, GK2, GK1, GK0) of the Ground-Key Filter Data register (Commands 52 and 53, E8/E9h). This sampling period clocks the counter, which buffers the CD2/CD1B bit’s status before it is valid for presenting to the CDB bit of the Real Time Data Register. When the sampled value of the ground-key (or CD2) input is high, the counter is incremented by each clock pulse. When the sampled value is low, the counter is decremented. Once the counter increments to its maximum value of 6, it sets a latch whose output is routed to the corresponding CDB bit. If the counter decrements to its minimum value of 0, this latch is cleared and the output bit is set to zero. All other times, the latch (and the CDB status) remains in its previous state without change. It therefore takes at least six consecutive GK clocks with the debounce input remaining at the same state to effect an output change. If the GK bit value is set to zero, the buffering is bypassed and the input status is passed directly to CDB. SLAC Products 29 Q D CD1 Debounce Counter D Q D Q D Q CK EN/HOLD * DSH0–DSH3 Debounce Period (0–15 ms) 8 CDA Q RST FS (8 kHz) Notes: * Transparent latch: Output follows input when EN is high; output holds last state when EN is low Debounce Counter: Output goes high after counting to programmed (DSH) number of 1 ms clocks; Counter is reset for CD1 input changes at 125 µs sample period. DSH0–DSH3 programmed value is common for all 4 channels, but debounce counter is separate per channel a. Loop Detect Debounce Filter MUX CD2 or CD1B GK=0 GK0–GK3 Ground-Key Sampling Interval (1–15 ms) CDB UP/DN Q GK=0 GK 1 kHz RST Clock Divider (1–15 ms clock output) Six-State Up/Down Counter Notes: Programmed value of GK0–GK3 determines clock rate (1–15 ms) of six-state counter. If GK value = 0, counter is bypassed and no buffering occurs. Six-state up/down counter: Counts up when input is high; counts down when input is low. Output goes and stays high when maximum count is reached; output goes and stays low when counts down to zero. b. Ground-Key Filter Figure 11. MPI Real-Time Data Register or GCI Upstream SC Channel Data 30 Am79Q02/021/031 Data Sheet Real-Time Data Register Operation Active State To obtain time-critical data such as off/on-hook and ring trip information from the SLIC with a minimum of processor time and effort, the QSLAC device contains an 8-bit Real Time Data register. This register contains CDA and CDB bits from all four channels. The CDA bit for each channel is a debounced version of the CD1 input. The CDA bit is normally used for switchhook. The CDB bit for each channel normally contains the CD2 input bit; however, if the E1 multiplex operation is enabled, the CDB bit will contain the debounced value of the CD1B bit. CD1 and CD2 can be assigned to offhook, ring trip, ground key signals, or other signals. Frame sync is needed for the debounce and the ground key signals. If Frame sync is not provided, the real-time register will not work. The register is read using MPI Commands 16 and 17 (4D/4Fh), and may be read at any time regardless of the state of the Channel Enable Register. This allows off/on-hook, ring trip, or ground key information for all four channels to be obtained from the QSLAC device with one read operation versus one read per channel. If these data bits are not used for supervision information, they can be accessed on an individual channel basis in the same way as C3–C5; however, CD1 and CD1B will not be debounced. Each channel of the QSLAC device can operate in either the Active (operational) or Inactive (standby) state. In the Active state, individual channels of the QSLAC device can transmit and receive PCM or linear data and analog information. The Active state is required when a telephone call is in progress. The activate command (MPI Command 5), puts the selected channel(s) into this state (see channel enable register). Bringing a channel of the QSLAC device into the Active state is only possible through the MPI. Interrupt In addition to the Real Time Data register, an interrupt signal has been implemented in the QSLAC device. The interrupt signal is an active Low output signal which pulls Low whenever the unmasked CD bits change state (Low to High or High to Low); or whenever the transmit PCM data changes on a channel in which the Arm Transmit Interrupt (ATI) bit is on. The interrupt control is shown in Figure 9. The interrupt remains Low until the appropriate register is read. This output can be programmed as TTL or open drain. When an interrupt is generated, all of the unmasked bits in the Real Time Data register latch and remain latched until the interrupt is cleared. The interrupt is cleared by reading the register with Command 17, by writing to the interrupt mask register (Command 26), or by a reset. If any of the inputs to the unmasked bits in the Real Time Data register are different from the register bits when the interrupt is cleared, a new interrupt is immediately generated with the new data latched into the Real Time Data register. For this reason, the interrupt logic in the controller should be level-sensitive rather than edge-sensitive. Inactive State All channels of the QSLAC device are forced into the Inactive (standby) state by a power-up or hardware reset. Individual channels can be programmed into this state by the deactivate command (Command 1) or by the software reset command (Command 2). Power is disconnected from all nonessential circuitry while the MPI remains active to receive commands. The analog output is tied to VREF through a resistor whose value depends on the VMODE bit. All circuits that contain programmed information retain their data in the Inactive state. Low Power State If the Low Power state is turned on by setting LPM = 1 (Command 14), the internal clock speed substantially reduces when all four channels are deactivated. When this happens, the CFAIL bit is set to 1, and if MCLK also is lost, the microprocessor interface requires a minimum of 75 ms off time between commands. Chopper Clock On the Am79Q02JC there is a chopper clock output to drive the switching regulator on some Legerity SLICs. The clock frequency is selectable as 256 or 292.57 kHz by the CHP bit (Command 12). The chopper output must be turned on with the ECH bit (Command 45). Reset States The QSLAC device can be reset by application of power, by an active Low on the hardware Reset pin (RST), by a hardware reset command, or by CS Low for 16 or more rising edges of DCLK. This resets the QSLAC device to the following state: 1. A-law companding is selected. 2. Default B, X, R, and Z filter values are selected and the AISN is set to zero. 3. Default digital gain blocks (GX, GR) are selected. The analog gains, AX and AR, are set to 0 dB. Interrupt Mask Register The Real Time Data register data bits can be masked from causing an interrupt to the processor using the interrupt mask register. The mask register can be written or read via the MPI Commands 26 and 27. 4. SLIC I/Os (CD1–C5) are set to the Input state. 5. All of the test states in the Operating Conditions register are turned off (0’s). 6. All four channels are in the Inactive (standby) state. SLAC Products 31 7. Transmit time slots and receive time slots are set to 0, 1, 2, and 3 for channels 1, 2, 3, and 4, respectively. The clock slots are set to 0, with transmit on the negative edge. 16. The E1 Multiplex state is turned off and the polarity is set for high going pulses. 17. No signalling on the PCM highway. SIGNAL PROCESSING Overview of Digital Filters 8. DXA port is selected for all channels. 9. DRA port is selected for all channels. Several of the blocks in the signal processing section are user programmable. These allow the user to optimize the performance of the QSLAC device for the system. Figure 12 shows the QSLAC device signal processing and indicates the programmable blocks. 10. The master clock frequency selected is 8.192 MHz and is programmed to come from PCLK. 11. All four channels are selected in the Channel Enable register. 12. Any pending interrupts are cleared, all interrupts are masked, and the Interrupt Output state is set to open drain. The advantages of digital filters are: 13. The supervision debounce time is set to 8 ms. 14. The previously programmed B, Z, X, R, GX, and GR filters are unchanged. 15. The chopper clock frequency is set to 256 kHz but the chopper clock is turned off. High reliability No drift with time or temperature Unit-to-unit repeatability Superior transmission performance Flexibility Maximum possible bandwidth for V.34 modems High Pass Filter (HPF) * VIN AX Full Digital Loop back (FDL) + GX X * * Compressor LPF & HPF TSA Loopback (TLB) AISN Z * * B * Cutoff Receive Path (CRP) + VOUT Decimator Decimator ADC Cutoff Transmit Path (CTP) TSA Digital TX * AR Interpolator DAC + Interpolator VREF * Expander LPF GR R * * Receive Lower Gain (RG) 0 TSA Digital RX 1 kHz Tone (TON) programmable blocks 21108-027 Figure 12. QSLAC Device Block Diagram 32 Am79Q02/021/031 Data Sheet Two-Wire Impedance Matching Transmit Signal Processing Two feedback paths on the QSLAC device synthesize the two-wire input impedance of the SLIC by providing a programmable feedback path from VIN to VOUT. The Analog Impedance Scaling Network (AISN) is a programmable analog gain of –0.9375 to +0.9375 from VIN to VOUT . The Z filter is a programmable digital filter providing an additional path and programming flexibility over the AISN in modifying the transfer function from VIN to VOUT. Together, the AISN and the Z-Filter enable the user to synthesize virtually all required SLIC input impedances. In the transmit path (A/D), the analog input signal (VIN) is A/D converted, filtered, companded (for A-law or µ-law), and made available to the PCM highway in A-law, µ-law, or linear form. If linear form is selected, the 16-bit data will be transmitted in two consecutive time slots starting at the programmed time slot. The signal processor contains an ALU, RAM, ROM, and control logic to implement the filter sections. The B, X, and GX blocks are user-programmable digital filter sections with coefficients stored in the coefficient RAM, while AX is an analog amplifier that can be programmed for 0 dB or 6.02 dB gain. The B, X, and GX filters can also b e op e ra te d f r om an a l t er n a te s e t o f de fa u lt coefficients stored in ROM (Commands 24 and 25). Frequency Response Correction and Equalization The QSLAC device contains programmable filters in the receive (R) and transmit (X) directions that may be programmed for line equalization and to correct any attenuation distortion caused by the Z filter. Transhybrid Balancing The QSLAC device’s programmable B filter is used to adjust transhybrid balance. The filter has a single pole IIR section (BIIR) and an eight-tap FIR section (BFIR), both operating at 16 kHz. Gain Adjustment T h e Q S L AC d ev i c e ’s t r a n s m i t p a t h h a s t w o programmable gain blocks. Gain block AX is an analog gain of 0 dB or 6.02 dB (unity gain or gain of 2.0), located immediately before the A/D converter. GX is a digital gain block that is programmable from 0 dB to +12 dB, with a worst-case step size of 0.1 dB for gain settings below +10 dB, and a worst-case step size of 0.3 dB for gain settings above +10 dB. The filters provide a net gain in the range of 0 dB to 18 dB. The QSLAC device receive path has two programmable loss block s. GR is a digital los s bl ock that is programmable from 0 dB to 12 dB, with a worst-case step size of 0.1 dB. Loss block AR is an analog loss of 0 dB or 6.02 dB (unity gain or gain of 0.5), located immediately after the D/A converter. This provides a net loss in the range of 0 dB to 18 dB. An additional 6 dB attenuation is provided as part of GR, which can be inserted by setting the RG bit of Command 70/71h. This allows writing of a single bit to introduce 6 dB of attenuation into the receive path without having to reprogram GR. This 6 dB loss is implemented as part of GR and the total receive path attenuation must remain in the specified 0 to –12 dB range. If the RG bit is set, the programmed value of GR must not introduce more than an additional 6 dB attenuation. The decimator reduces the high input sampling rate to 16 kHz for input to the B, GX, and X filters. The X filter is a six-tap FIR section which is part of the frequency response correction network. The B filter operates on samples from the receive signal path in order to provide transhybrid balancing in the loop. The highpass filter rejects low frequencies such as 50 Hz or 60 Hz, and may be disabled. Transmit PCM Interface The transmit PCM interface transmits a 16-bit linear code (when programmed) or an 8-bit compressed code from the digital A-law/µ-law compressor. Transmit logic controls the transmission of data onto the PCM highway through output port selection and time/clock slot control circuitry. The linear data requires two consecutive time slots, while a single time slot is required for A-law/µ-law data. In the PCM Signaling state (SMODE = 1), the transmit time slot following the A-law or µ-law data is used for signaling information. The two time slots form a single 16-bit data block. The frame sync (FS) pulse identifies time slot 0 of the transmit frame and all channels (time slots) are referenced to it. The logic contains userprogrammable Transmit Time Slot and Transmit Clock Slot registers. The Time Slot register is 7 bits wide and allows up to 128 8-bit channels (using a PCLK of 8.192 MHz) in each frame. This feature allows any clock frequency between 128 kHz and 8.192 MHz (2 to 128 channels) in a system. The data is transmitted in bytes, with the most significant bit first. The Clock Slot register is 3 bits wide and may be programmed to offset the time slot assignment by 0 to 7 PCLK periods to eliminate any clock skew in the system. An exception occurs when division of the PCLK frequency by 64 kHz produces a nonzero remainder, R, and when the transmit clock slot is greater than R. In that case, the R-bit fractional time SLAC Products 33 slot after the last full time slot in the frame will contain random information and will have the TSC output turned on. For example, if the PCLK frequency is 1.544 MHz (R = 1) and the transmit clock slot is greater than 1, the 1-bit fractional time slot after the last full time slot in the frame will contain random information, and the TSC output will remain active during the fractional time slot. In such cases, problems can be avoided by not using the last time slot. The frame sync (FS) pulse identifies time slot 0 of the receive frame, and all channels (time slots) are referenced to it. The lo gic co nta ins userprogrammable Receive Time Slot and Receive Clock Slot registers. The Time Slot register is 7 bits wide and allows up to 128 8-bit channels (using a PCLK of 8.192 MHz) in each frame. This feature allows a n y c l o c k f r e q u e n c y b e t w e e n 128 kHz and 8.192 MHz (2 to 128 channels) in a system. The PCM data may be user programmed for output onto either the DXA or DXB por t or both por ts simultaneously. Correspondingly, either TSCA or TSCB or both are Low during transmission. The Clock Slot register is 3 bits wide and can be programmed to offset the time slot assignment by 0 to 7 PCLK periods to eliminate any clock skews in the system. An exception occurs when division of the PCLK frequency by 64 kHz produces a nonzero remainder (R), and when the receive clock slot is greater than R. In that case, the last full receive time slot in the frame is not usable. If the PCLK frequency is 1.544 MHz (R=1/8, or 1 clock slot within a time slot), the receive clock slot can be only 0 or 1 if the last time slot is to be used. The PCM data can be programmed for input from the DRA or DRB port. The DXA/DXB and TSCA/TSCB outputs can be programmed to change either on the negative or positive edge of PCLK. Tr a n s m i t d a t a c a n a l s o b e r e a d t h r o u g h t h e microprocessor interface using Command 47. Receive Signal Processing In the receive path (D/A), the digital signal is expanded (for A-law or µ-law), filtered, converted to analog, and passed to the VOUT pin. The signal processor contains an ALU, RAM, ROM, and Control logic to implement the filter sections. The Z, R, and GR blocks are user-programmable filter sections with their coefficients stored in the coefficient RAM, while AR is an analog amplifier which can be programmed for a 0 dB or 6.02 dB loss. The Z, R, and GR filters can also be operated from an alternate set of default coefficients stored in ROM (Commands 24 and 25). The low-pass filter band limits the signal. The R filter is composed of a six-tap FIR section operating at a 16 kHz sampling rate and a one-tap IIR section operating at 8 kHz. It is part of the frequency response correction network. The Analog Impedance Scaling Network (AISN) is a user-programmable gain block providing feedback from VIN to VOUT to emulate different SLIC input impedances from a single exter nal SLIC impedance. The Z filter provides feedback from the transmit signal path to the receive path and is used to modify the effective input impedance to the system. The interpolator increases the sampling rate prior to D/A conversion. Analog Impedance Scaling Network (AISN) The AISN is in the QSLAC device to scale the value of the external SLIC impedance. Scaling this external impedance with the AISN (along with the Z filter) allows matching of many different line conditions using a single impedance value. Linecards can meet many different specifications without any hardware changes. The AISN is a programmable transfer function connected from VIN to VOUT for each QSLAC device channel. The AISN transfer function alters the input impedance of the SLIC device to a new value (ZIN): Z IN = ZSL • ( 1 – G44 • h AISN ) ⁄ ( 1 – G 440 • h AISN ) where G440 is the SLIC echo gain into an open circuit, G44 is the SLIC echo gain into a short circuit, and ZSL is the SLIC input impedance without the QSLAC device. The gain can be varied from –0.9375 to +0.9375 in 31 steps of 0.0625. The AISN gain is determined by the following equation: 4 h AISN æ ö i = 0.0625 ç AISN i • 2 ÷ – 16 ç ÷ èi = 0 ø å Receive PCM Interface where AISNi = 0 or 1 The receive PCM interface logic controls the reception of data bytes from the PCM highway, transfers the data to the A-law/µ-law expansion logic for compressed signals, and then passes the data to the receive path of the signal processor. If the data received from the PCM highway is programmed for linear code, the A-law/µ-law expansion logic is bypassed and the data is presented to the receive path of the signal processor directly. The linear data requires two consecutive time slots, while the A-law or µ-law data requires a single time slot. There are two special cases to the formula for hAISN: 1) a value of AISN = 00000 will specify a gain of 0 (or cutoff), and 2) a value of AISN = 10000 is a special case where the AISN circuitry is disabled and VOUT is connected internally to VIN with a gain of 0 dB. This allows a Full Digital Loopback state where an input digital PCM signal is completely processed through the receive section, looped back, processed through the transmit section, and output as digital PCM data. 34 Am79Q02/021/031 Data Sheet During this test, the VIN input is ignored and the VOUT output is connected to VREF. Speech Coding The A/D and D/A conversion follows either the A-law or the µ-l aw stand ard as defin ed i n ITU-T Recommendation G.711. A-law or µ-law operation is programmed using MPI Commands 24 and 25. Alternate bit inversion is performed as part of the A-law coding. The QSLAC device provides linear code as an option on both the transmit and receive sides of the device. Linear code is selected using MPI Commands 24 and 25. Two successive time slots are required for linear code operation. The linear code is a 16-bit two’scomplement number which appears sign bit first on the PCM highway. Linear code occupies two time slots. Signaling on the PCM Highway If the SMODE bit is set in the Configuration register, each data point occupies two consecutive time slots. The first time slot contains A-law or µ-law data and the second time slot will have the following information: Bit 7: Debounced CD1 bit (usually hookswitch) Bit 6: CD2 bit or CD1B bit Bits 5–3: Reserved Bit 2: CFAIL Bits 1–0: Reserved Bit 7 of the signaling byte will appear immediately after bit 0 of the data byte. A-law or µ-law Companded state must be specified in order to put signaling information on the PCM highway. The signaling time slot remains active, even when the channel is deactivated. Robbed-Bit Signaling Compatibility The QSLAC device supports robbed bit signaling compatibility. Robbed bit signaling allows periodic use of the least significant bit (LSB) of the receive path PCM data to be used to carry signaling information. In this scheme, separate circuitry within the line card or system intercepts this bit out of the PCM data stream and uses this bit to control signaling functions within the system. The QSLAC device does not perform any processing of any of the robbed bits during this operation; it simply allows for the robbed bit presence by performing the LSB substitution. If the RBE bit is set, then the robbed-bit signaling compatibility mode is enabled. Robbed-bit signaling is only available in the µ-law companding mode of the device. Also, only the receive (digital-to-analog) path is involved. There is no change of operation to the transmit path and PCM data coming out of the QSLAC device will always contain complete PCM byte data for each time slot, regardless of robbed-bit signaling selection. In the absence of actual PCM data for the affected time slots, there is an uncertainty of the legitimate value of this bit to accurately reconstruct the analog signal. This bit can always be assumed to be a 1 or 0; hence, the reconstructed signal is correct half the time. However, the other half of the time, there is an unacceptable reconstruction error of a significance equal to the value weighting of the LSB. To reduce this error and provide compatibility with the robbed bit signaling scheme, when in the robbed-bit signaling mode, the QSLAC device ignores the LSB of each received PCM byte and replace its value in the expander with a value of half the LSB’s weight. This then guarantees the reconstruction is in error by only half this LSB weight. In the expander, the eight bits of the companded PCM byte are expanded into linear PCM data of several more bits within the internal signal processing path of the device. Therefore, accuracy is not limited to the weight of the LSB, and a weight of half this value is realizable. When this robbed-bit mode is selected, not every frame contains bits for signaling, and therefore not every byte requires its LSB substituted with the halfLSB weight. This substitution only occurs for valid PCM time slots within frames for which this robbed bit has been designated. To determine which time slots are affected, the device monitors the frame sync (FS) pulse. The current frame is a robbed-bit frame and this half-LSB value is used only when this criteria is met: The RBE bit is set, and The device is in the µ-law companding mode, and The current frame sync pulse (FS) is two PCLK cycles long, and The previous frame sync pulse (FS) was not two PCLK cycles long. The frame sync pulse is sampled on the falling edge of PCLK. As shown in Figure 13, if the above criteria is met, and if FS is high for two consecutive falling edges of PCLK then low for the third falling edge, it is considered a robbed-bit frame. Otherwise, it is a normal frame. SLAC Products 35 PCLK Notice that these default coefficient values are retained in a read-only memory area within the QSLAC device, and those values cannot be read back using any data commands. When the device is selected to use default coefficients, it obtains those values directly from the read-only memory area, where the coefficient read operations access the programmable random access data memory only. If an attempt is made to read back any filter values without those values first being written with known programmed data, the values read back are totally random and do not represent the default or any other values. FS COMMAND DESCRIPTION AND FORMATS Microprocessor Interface Description PCLK FS Normal Frame (Not Robbed-Bit) A microprocessor can program and control the QSLAC device using the MPI. Data programmed previously can be read out for verification. Robbed-Bit Frame Figure 13. Robbed-Bit Frame Default Filter Coefficients The QSLAC device contains an internal set of default coefficients for the programmable filters. These coefficients were determined to allow reasonable system performance for initial power-up non-programmed situations, such as may exist before a system processor has opportunity to program any coefficients. The default filter coefficients are calculated assuming an Am7920 SLIC with 50 Ω protection resistors, a 178 kΩ transversal impedance (ZT), and a 90.5 kΩ receive impedance (ZRX). This SLIC has a transmit gain of 0.5 (GTX) and a current gain of 500 (K1). The transmit relative level is set to +0.28 dBr, and the receive relative level is set to –4.39 dBr. The equalization filters (X and R) are not optimized. The balance filter was designed to give acceptable balance into a variety of impedances. The nominal input impedance was set to 815 Ω. If the SLIC circuit differs significantly from this design, the default filters cannot be used and must be replaced by programmed coefficients. To obtain this above-system response, the default filter coefficients are set to produce these values: GX gain = +6 dB, GR gain = –8.984 dB AX gain = 0 dB, AR gain = 0 dB R filter: H(z) = 1, X filter: H(z) = 1 Z filter: H(z) = 0, B filter H(z) = 0 Commands assign values to the following channel parameters: Transmit time slot Receive time slot Transmit clock slot Receive clock slot Transmit gain Receive loss B-filter coefficients X-filter coefficients R-filter coefficients Z-filter coefficients AISN coefficient Read/Write SLIC Input/Output SLIC Input/Output Direction Select A-law, µ-law, or linear code Select Transmit PCM Port A or B or both Select Receive PCM Port A or B Programmed/Default B filter Programmed/Default Z filter Programmed/Default X filter Programmed/Default R filter Programmed/Default GX filter Programmed/Default GR filter Enable/disable AX amplifier Enable/disable AR amplifier Select test states Select Active or Inactive (standby) state Commands are provided to read values from the following channel monitors: SLIC status Transmit PCM data AISN = cutoff 36 Am79Q02/021/031 Data Sheet Commands are provided to assign values to the following global chip parameters: Transmit PCM Clock Edge Interrupt Output Drive state Chopper Clock Frequency Select Signaling on the PCM Highway Select Master Clock Frequency Channel Enable register Debounce Time for CD1 Enable E1 Output E1 Polarity Commands are provided to read values from the following global chip status monitors: Real Time Data register Power Interruption Bit Clock Failure Bit Interrupt Mask register Revision Code Number The following description of the MPI (Microprocessor Interface) is valid for channel 1–4. If desired, multiple channels may be programmed simultaneously with identical information by setting multiple Channel Enable bits. Channel enables are contained in the Channel Enable register and written or read using MPI Commands 14 and 15. If multiple Channel Enable bits are set for a read operation, only data from the first enabled channel will be read. The MPI physically consists of a serial data input/ output (DIO), a data clock (DCLK), and a chip select (CS). Individual Channel Enable bits EC1, EC2, EC3, and EC4 are stored internally in the Channel Enable register of the QSLAC device. The serial input consists of 8-bit commands which may be followed with additional bytes of input data, or may be followed by the QSLAC device sending out bytes of data. All data input and output is MSB (D7) first and LSB (D0) last. All data bytes are read or written one at a time, with CS going High for at least a minimum off period before the next byte is read or written. Only a single channel should be enabled during read commands. All commands that require additional input data to the device must have the input data as the next N words written into the device (for example, framed by the next N transitions of CS). Program all unused bits as 0 to ensure compatibility with future parts. All commands that are followed by output data will cause the device to output data for the next N transitions of CS going L ow. Th e Q S L AC d ev i c e w i l l n o t a c c e p t a ny commands until all the data has been shifted out. The output values of unused bits are not specified. An MPI cycle is defined by transitions of CS and DCLK. If the CS lines are held in the High state between accesses, the DCLK may run continuously with no change to the internal control data. Using this method, the same DCLK may be run to a number of QSLAC devices and the individual CS lines will select the appropriate device to access. Between command s equ enc e s, D CLK c an s tay i n th e Hi gh st ate indefinitely with no loss of internal control information regardless of any transitions on the CS lines. Between bytes of a multibyte read or write command sequence, DCLK can also stay in the High state indefinitely. DCLK can stay in the Low state indefinitely with no loss of internal control information, provided the CS lines remain at a High level. If a low period of CS contains less than 8 positive DCLK transitions, it will be ignored. If it contains 8–15 positive transitions, only the last 8 transitions matter. If it contains 16 or more positive transitions, it will cause a hardware reset in the part. If the chip is in the middle of a read sequence when CS goes Low, data will be present at the DIO pin even if DCLK has no activity. SLAC Products 37 SUMMARY OF MPI COMMANDS* Number Hex Description 1 00 Deactivate (Standby Mode) 2 02 Software Reset 3 04 Hardware Reset 4 06 No Operation Activate (Operational Mode) 5 0E 6,7 40/41 Write/Read Transmit Time Slot and PCM Highway Selection 8,9 42/43 Write/Read Receive Time Slot and PCM Highway Selection 10,11 44/45 Write/Read REC & TX Clock Slot and TX Edge 12,13 46/47 Write/Read Configuration Register 14,15 4A/4B Write/Read Channel Enable & Operating Mode Register 16 4D Read Real Time Data Register 17 4F Read Real Time Data Register and Clear Interrupt 18,19 50/51 Write/Read AISN and Analog Gains 20,21 52/53 Write/Read SLIC Input/Output Register 22,23 54,55 Write/Read SLIC Input/Output Direction and Status Bits 24,25 60/61 Write/Read Operating Functions 26,27 6C/6D Write/Read Interrupt Mask Register 28,29 70/71 Write/Read Operating Conditions 30 73 31,32 80/81 Read Revision Code Number (RCN) Write/Read GX Filter Coefficients 33,34 82/83 Write/Read GR Filter Coefficients 35,36 84/85 Write/Read Z Filter Coefficients (FIR and IIR) 37, 38 86/87 Write/Read B1 Filter Coefficients (FIR) 39, 40 88/89 Write/Read X Filter Coefficients 41, 42 8A/8B Write/Read R Filter Coefficients 43, 44 96/97 Write/Read B2 Filter Coefficients (IIR) 45, 46 C8/C9 Write/Read Debounce Time Register 47 CD Read Transmit PCM Data 48, 49 98/99 Write/Read Z Filter Coefficients (FIR only) 50, 51 9A/9B Write/Read Z Filter Coefficients (IIR only) 52,53 E8/E9h Write/Read Ground Key Filter Sampling Interval Note: *All codes not listed are reserved by Legerity and should not be used. 38 Am79Q02/021/031 Data Sheet MPI COMMAND STRUCTURE This section details each MPI command. Each command is shown along with the format of any additional data bytes that follow. For details of the filter coefficients of the form Cxymxy, refer to the General Description of CSD Coefficients section on page 56. Unused bits are indicated by “RSVD”; 0’s should be written to them, but 0’s are not guaranteed when they are read. *Default field values are marked by an asterisk. A hardware reset forces the default values. 1. Deactivate (Standby State) MPI Command (00h) Command D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0 0 0 0 0 In the Deactivated mode: All programmed information is retained. The Microprocessor Interface (MPI) remains active. The PCM inputs are disabled and the PCM outputs are high impedance unless signaling on the PCM highway is programmed (SMODE = 1). The analog output (VOUT) is disabled and biased at 2.1 V. The channel status (CS) bit in the SLIC I/O Direction and Channel Status Register is set to 0. 2. Software Reset MPI Command (02h) Command D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0 0 0 1 0 The action of this command is identical to that of the RST pin except that it only operates on the channels selected by the Channel Enable Register and it does not change clock slots, time slots, PCM highways, or global chip parameters. See the note under the hardware reset command that follows. 3. Hardware Reset MPI Command (04h) Command D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0 0 1 0 0 Hardware reset is equivalent to pulling the RST on the device Low. This command does not depend on the state of the Channel Enable Register. Note: The action of a hardware reset is described in Reset States on page 31 of the section Operating the QSLAC Device. SLAC Products 39 4. No Operation MPI Command (06h) D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0 0 1 1 0 Command 5. Activate Channel (Operational Mode) MPI Command (0Eh) Command D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0 1 1 1 0 This command places the device in the Active mode and sets CSTAT = 1. No valid PCM data is transmitted until after the second FS pulse is received following the execution of the Activate command. 6, 7. Write/Read Transmit Time Slot and PCM Highway Selection MPI Command (40/41h) R/W = 0: Write R/W = 1: Read D7 Command I/O Data D6 D5 D4 D3 D2 D1 D0 0 1 0 0 0 0 0 R/W TPCM TTS6 TTS5 TTS4 TTS3 TTS2 TTS1 TTS0 Transmit PCM Highway TPCM = 0* TPCM = 1 Transmit on Highway A (see TAB in Commands 10, 11) Transmit on Highway B (see TAB in Commands 10, 11) Transmit Time Slot TTS = 0–127 Time Slot Number (TTS0 is LSB, TTS6 is MSB) PCM Highway B is not available on the Am79Q021/031 QSLAC devices. * Power Up and Hardware Reset (RST) Value = 00h, 01h, 02h, 03h for Channels 1, 2, 3, and 4, respectively. 8, 9. Write/Read Receive Time Slot and PCM Highway Selection MPI Command (42/43h) R/W = 0: Write R/W = 1: Read D7 Command I/O Data D6 D5 D4 D3 D2 D1 D0 0 1 0 0 0 0 1 R/W RPCM RTS6 RTS5 RTS4 RTS3 RTS2 RTS1 RTS0 Receive PCM Highway RPCM = 0* RPCM = 1 Receive on Highway A Receive on Highway B Receive Time Slot RTS = 0–127 Time Slot Number (RTS0 is LSB, RTS6 is MSB) PCM Highway B is not available on the Am79Q021 and the Am79Q031 QSLAC devices. * Power Up and Hardware Reset (RST) Value = 00h, 01h, 02h, 03h for Channels 1, 2, 3, and 4, respectively. 40 Am79Q02/021/031 Data Sheet 10, 11. Write/Read Transmit Clock Slot, Receive Clock Slot, and Transmit Clock Edge MPI Command (44/45h) R/W = 0: Write R/W = 1: Read D7 Command I/O Data D6 D5 D4 D3 D2 D1 D0 0 1 0 0 0 1 0 R/W TAB XE RCS2 RCS1 RCS0 TCS2 TCS1 TCS0 Transmit on A and B TAB = 0* TAB = 1 Transmit data on highway selected by TPCM (See Commands 6,7 on page 40). Transmit data on both highways A and B Transmit Edge XE = 0* XE = 1 Transmit changes on negative edge of PCLK Transmit changes on positive edge of PCLK Receive Clock Slot RCS = 0*–7 Receive Clock Slot number Transmit Clock Slot TCS = 0*–7 Transmit Clock Slot number The XE bit and the clock slots apply to all four channels; however, they cannot be written or read unless at least one channel is selected in the Channel Enable Register. * Power Up and Hardware Reset (RST) Value = 00h. 12, 13. Write/Read Configuration Register MPI Command (46/47h) R/W = 0: Write R/W = 1: Read Command I/O Data D7 D6 D5 D4 D3 D2 D1 D0 0 1 0 0 0 1 1 R/W INTM CHP SMODE CMODE CSEL3 CSEL2 CSEL1 CSEL0 Interrupt Mode INTM = 0 INTM = 1* TTL-compatible output Open drain output Chopper Clock Control CHP = 0* CHP = 1 Chopper Clock is 256 kHz (2048/8 kHz) Chopper Clock is 292.57 kHz (2048/7 kHz) PCM Signaling Mode SMODE = 0* SMODE = 1 No signaling on PCM highway Signaling on PCM highway Clock Source Mode CMODE = 0 CMODE = 1* MCLK used as master clock; no E1 multiplexing allowed PCLK used as master clock; E1 multiplexing allowed if enabled in commands 49, 50. The master clock frequency can be selected by CSEL. The master clock frequency selection affects all channels. SLAC Products 41 Master Clock Frequency CSEL = 0000 CSEL = 0001 CSEL = 0010 CSEL = 0011 CSEL = 01xx CSEL = 10xx CSEL = 11xx CSEL = 1010* 1.536 MHz 1.544 MHz 2.048 MHz Reserved Two times frequency specified above (2 x 1.536 MHz, 2 x 1.544 MHz, or 2 x 2.048 MHz) Four times frequency specified above (4 x 1.536 MHz, 4 x 1.544 MHz, or 4 x 2.048 MHz) Reserved 8.192 MHz is the default These commands do not depend on the state of the Channel Enable Register. * Power Up and Hardware Reset (RST) Value = 9Ah. 14, 15. Write/Read Channel Enable and Operating Mode Register MPI Command (4A/4B) R/W = 0: Write R/W = 1: Read D7 Command I/O Data RSVD D6 D5 D4 D3 D2 D1 D0 0 1 0 0 1 0 1 R/W RSVD RBE VMODE LPM EC4 EC3 EC2 EC1 Reserved for future use. Always write as 0, but 0 is not guaranteed when read. Robbed-bit Mode RBE = 0* RBE = 1 Robbed-bit Signaling mode is disabled. Robbed-bit Signaling mode is enabled on PCM receiver if µ-law is selected. VOUT Mode VMODE = 0* VMODE = 1 VOUT = VREF through a resistor when channel is deactivated VOUT high impedance when channel is deactivated. Low Power Mode LPM = 0* LPM = 1 Low Power mode off Low Power mode on while all channels are inactive Channel Enable 4 EC4 = 0 EC4 = 1* Disabled, Channel 4 cannot receive commands Enabled, Channel 4 can receive commands Channel Enable 3 EC3 = 0 EC3 = 1* Disabled, Channel 3 cannot receive commands Enabled, Channel 3 can receive commands Channel Enable 2 EC2 = 0 EC2 = 1* Disabled, Channel 2 cannot receive commands Enabled, Channel 2 can receive commands Channel Enable 1 EC1 = 0 EC1 = 1* Disabled, Channel 1 cannot receive commands Enabled, Channel 1 can receive commands * Power Up and Hardware Reset (RST) Value = 0Fh. 42 Am79Q02/021/031 Data Sheet 16, 17. Read Real-Time Data Register MPI Command (4D/4Fh) C = 0: Do not clear interrupt C = 1: Clear interrupt This register writes/reads real-time data with or without clearing the interrupt. D7 Command Output Data D6 D5 D4 D3 D2 D1 D0 0 1 0 0 1 1 C 1 CDB4 CDA4 CDB3 CDA3 CDB2 CDA2 CDB1 CDA1 Real Time Data CDA1 CDB1 CDA2 CDB2 CDA3 CDB3 CDA4 CDB4 Debounced data bit 1 on Channel 1 Data bit 2 or multiplexed data bit 1 on Channel 1 Debounced data bit 1 on Channel 2 Data bit 2 or multiplexed data bit 1 on Channel 2 Debounced data bit 1 on Channel 3 Data bit 2 or multiplexed data bit 1 on Channel 3 Debounced data bit 1 on Channel 4 Data bit 2 or multiplexed data bit 1 on Channel 4 This command does not depend on the state of the Channel Enable Register. 18, 19. Write/Read AISN and Analog Gains MPI Command (50/51h) R/W = 0: Write R/W = 1: Read Command I/O Data RSVD D7 D6 D5 D4 D3 D2 D1 D0 0 1 0 1 0 0 0 R/W RSVD AX AR AISN4 AISN3 AISN2 AISN1 AISN0 Reserved for future use. Always write as 0, but 0 is not guaranteed when read. Transmit Analog Gain AX = 0* AX = 1 0 dB gain 6.02 dB gain Receive Analog Loss AR = 0* AR = 1 0 dB loss 6.02 dB loss AISN coefficient AISN = 0* – 31 See below (Default value = 0) The Impedance Scaling Network (AISN) gain can be varied from –0.9375 to 0.9375 in multiples of 0.0625. The gain coefficient is decoded using the following equation: h AISN = 0.0625 [ ( 16 • AISN4 + 8 • AISN3 + 4 • AISN2 + 2 • AISN1 + AISN0 ) – 16 ] where hAISN is the gain of the AISN. A value of AISN = 10000 turns on the Full Digital Loopback mode and a value of AISN = 0000* indicates a gain of 0 (cutoff). * Power Up and Hardware Reset (RST) Value = 00h. SLAC Products 43 20, 21. Write/Read SLIC Input/Output Register MPI Command (52/53h) R/W = 0: Write R/W = 1: Read Command I/O Data RSVD D7 D6 D5 D4 D3 D2 D1 D0 0 1 0 1 0 0 1 R/W RSVD RSVD CD1B C5 C4 C3 CD2 CD1 Reserved for future use. Always write as 0, but 0 is not guaranteed when read. Pins CD1, CD2, and C3 through C5 are set to 1 or 0. The data appears latched on the CD1, CD2, and C3 through C5 SLIC I/O pins, provided they were set in the Output mode (see Command 22). The data sent to any of the pins set to the Input mode is latched, but does not appear at the pins. The CD1B bit is only valid if the E1 Multiplex mode is enabled (EE1 = 1). * Power Up and Hardware Reset (RST) Value = 00h 22, 23. Write/Read SLIC Input/Output Direction, Read Status Bits MPI Command (54/55h) D7 D6 D5 D4 D3 D2 D1 D0 Command 0 1 0 1 0 1 0 R/W Input Data RSVD CSTAT CFAIL IOD5 IOD4 IOD3 IOD2 IOD1 RSVD Reserved for future use. Always write as 0, but 0 is not guaranteed when read. Channel Status (Read status only, write as 0) CSTAT = 0 CSTAT = 1 Channel is inactive (Standby mode). Channel is active. Clock Fail (Read status only, write as 0) CFAIL* = 0 The internal clock is synchronized to frame synch. CFAIL = 1 The internal clock is not synchronized to frame synch. * The CFAIL bit is independent of the Channel Enable Register. I/O Direction (Read/Write) IOD5 = 0* IOD5 = 1 IOD4 = 0* IOD4 = 1 IOD3 = 0* IOD3 = 1 IOD2 = 0* IOD2 = 1 IOD1 = 0* IOD1 = 1 C5 is an input C5 is an output C4 is an input C4 is an output C3 is an input C3 is an output CD2 is an input CD2 is an output CD1 is an input CD1 is an output Pins CD1, CD2, and C3 through C5 are set to Input or Output modes individually. Pins C3–C5 are not available on the Am79Q031 QSLAC device, and C5 is available only on the Am79Q021 QSLAC device. * Power Up and Hardware Reset (RST) Value = 00h 44 Am79Q02/021/031 Data Sheet 24, 25. Write/Read Operating Functions MPI Command (60/61h) R/W = 0: Write R/W = 1: Read D7 Command I/O Data D6 D5 D4 D3 D2 D1 D0 0 1 1 0 0 0 0 R/W C/L A/µ EGR EGX EX ER EZ EB Linear Code C/L = 0* C/L = 1 Compressed coding Linear coding A-law or µ-law A/µ = 0* A/µ = 1 A-law coding µ-law coding EGR = 0* EGR = 1 Default GR filter enabled Programmed GR filter enabled EGX = 0* EGX = 1 Default GX filter enabled Programmed GX filter enabled EX = 0* EX = 1 Default X filter enabled Programmed X filter enabled ER = 0* ER = 1 Default R filter enabled Programmed R filter enabled EZ = 0* EZ = 1 Default Z filter enabled Programmed Z filter enabled EB = 0* EB = 1 Default B filter enabled Programmed B filter enabled GR Filter GX Filter X Filter R Filter Z Filter B Filter * Power Up and Hardware Reset (RST) Value = 00h. SLAC Products 45 26, 27. Write/Read Interrupt Mask Register MPI Command (6C/6Dh) R/W = 0: Write R/W = 1: Read D7 D6 D5 D4 D3 D2 D1 D0 0 1 1 0 1 1 0 R/W MCDB4 MCDA4 MCDB3 MCDA3 MCDB2 MCDA2 MCDB1 MCDA1 Command I/O Data Mask CD Interrupt MCDxy = 0 CDxy bit is NOT MASKED MCDxy = 1* CDxy bit is MASKED x Bit number (A or B) y Channel number (1 through 4) Masked: A change does not cause the Interrupt Pin to go Low. This command does not depend on the state of the Channel Enable Register. * Power Up and Hardware Reset (RST) Value = FFh. 28, 29. Write/Read Operating Conditions MPI Command (70/71h) R/W = 0: Write R/W = 1: Read Command I/O Data D7 D6 D5 D4 D3 D2 D1 D0 0 1 1 1 0 0 0 R/W CTP CRP HPF RG ATI ILB FDL TON Cutoff Transmit Path CTP = 0* CTP = 1 Transmit path connected Transmit path cut off Cutoff Receive Path CRP = 0* CRP = 1 Receive path connected Receive path cutoff (see note) High Pass Filter HPF = 0* HPF = 1 Transmit Highpass filter enabled Transmit Highpass filter disabled Lower Receive Gain RG = 0* RG = 1 6 dB loss not inserted 6 dB loss inserted Arm Transmit Interrupt ATI = 0* ATI = 1 Transmit Interrupt not Armed Transmit Interrupt Armed Interface Loopback ILB = 0* ILB = 1 TSA loopback disabled TSA loopback enabled Full Digital Loopback FDL = 0* FDL = 1 Full digital loopback disabled Full digital loopback enabled 1 kHz Receive Tone TON = 0* TON = 1 1 kHz receive tone off 1 kHz receive tone on * Power Up and Hardware Reset (RST) Value = 00h. The B Filter is disabled during receive cutoff. 46 Am79Q02/021/031 Data Sheet 30. Read Revision Code Number (RCN) MPI Command (73h) D7 Command I/O Data D6 D5 D4 D3 D2 D1 D0 0 1 1 1 0 0 1 1 RCN7 RCN6 RCN5 RCN4 RCN3 RCN2 RCN1 RCN0 This command returns an 8-bit number (RCN) describing the revision number of the QSLAC device. This command does not depend on the state of the Channel Enable Register. 31, 32. Write/Read GX Filter Coefficients MPI Command (80/81h) R/W = 0: Write R/W = 1: Read Command D7 D6 1 0 D5 D4 0 0 D3 D2 0 0 D1 D0 0 R/W I/O Data Byte 1 C40 m40 C30 m30 I/O Data Byte 2 C20 m20 C10 m10 The coefficient for the GX filter is defined as: H GX = 1 + ( C10 • 2 – m10 { 1 + C20 • 2 – m20 [ 1 + C30 • 2 – m30 ( 1 + C40 • 2 – m40 )]} ) Power Up and Hardware Reset (RST) Values = A9F0 (Hex) (HGX = 1.995 (6 dB)). Note: The default value is contained in a ROM register separate from the programmable coefficient RAM. There is a filter enable bit in Operating Functions Register to switch between the default and programmed values. 33, 34. Write/Read GR Filter Coefficients MPI Command (82/83h) R/W = 0: Write R/W = 1: Read Command: D7 D6 1 0 D5 D4 0 0 D3 D2 0 0 D1 D0 1 R/W I/O Data Byte 1 C40 m40 C30 m30 I/O Data Byte 2 C20 m20 C10 m10 The coefficient for the GR filter is defined as: H GR = C10 • 2 – m10 { 1 + C20 • 2 – m20 [ 1 + C30 • 2 – m30 ( 1 + C40 • 2 – m40 )]} Power Up and Hardware Reset (RST) Values = 23A1 (Hex) (HGR = 0.35547 (–8.984 dB)). See note under Commands 31 and 32. SLAC Products 47 35, 36. Write/Read Z Filter Coefficients (FIR and IIR) MPI Command (84/85h) R/W = 0: Write R/W = 1: Read This command writes and reads both the FIR and IIR filter sections simultaneously. Command D7 D6 1 0 D5 D4 0 0 D3 D2 0 1 D1 D0 0 R/W I/O Data Byte 1 C40 m40 C30 m30 I/O Data Byte 2 C20 m20 C10 m10 I/O Data Byte 3 C41 m41 C31 m31 I/O Data Byte 4 C21 m21 C11 m11 I/O Data Byte 5 C42 m42 C32 m32 I/O Data Byte 6 C22 m22 C12 m12 I/O Data Byte 7 C43 m43 C33 m33 I/O Data Byte 8 C23 m23 C13 m13 I/O Data Byte 9 C44 m44 C34 m34 I/O Data Byte 10 C24 m24 C14 m14 I/O Data Byte 11 C45 m45 C35 m35 I/O Data Byte 12 C25 m25 C15 m15 I/O Data Byte 13 C26 m26 C16 m16 I/O Data Byte 14 C47 m47 C37 m37 I/O Data Byte 15 C27 m27 C17 m17 The Z-transform equation for the Z filter is defined as: Hz ( z ) = z0 + z1 • z –1 + z2 • z –2 + z3 • z –3 + z4 • z –4 –1 z 5 • z6 • z7 • z + -------------------------------------–1 1 – z7 • z Sample rate = 32 kHz For i = 0 to 5 and 7 z i = C1i • 2 – m1i z 6 = C16 • 2 { 1 + C2i • 2 – m16 – m2i { 1 + C26 • 2 [ 1 + C3i • 2 – m26 – m3i ( 1 + C4i • 2 – m4i )]} } Power Up and Hardware Reset (RST) Values = 0190 0190 0190 0190 0190 0190 01 0190 (Hex) (Hz(z) = 0) See note under Commands 31 and 32. Note: Z6 is used for IIR filter scaling only. Its value is typically greater than zero but less than or equal to one. The input to the IIR filter section is first increased by a gain of 1/Z6, improving dynamic range and avoiding truncation limitations through processing within this filter. The IIR filter output is then multiplied by Z6 to normalize the overall gain. Z5 is the actual IIR filter gain value defined by the programmed coefficients, but it also includes the initial 1/Z6 gain. The theoretical effective IIR gain, without the Z6 gain and normalization, is actually Z5/Z6. 48 Am79Q02/021/031 Data Sheet 37, 38. Write/Read B1 Filter Coefficients MPI Command (86/87h) R/W = 0: Write R/W = 1: Read Command D7 D6 D5 D4 D3 D2 D1 D0 1 0 0 0 0 1 1 R/W I/O Input Data Byte 1 C32 m32 C22 m22 I/O Input Data Byte 2 C12 m12 C33 m33 I/O Input Data Byte 3 C23 m23 C13 m13 I/O Input Data Byte 4 C34 m34 C24 m24 I/O Input Data Byte 5 C14 m14 C35 m35 I/O Input Data Byte 6 C25 m25 C15 m15 I/O Input Data Byte 7 C36 m36 C26 m26 I/O Input Data Byte 8 C16 m16 C37 m37 I/O Input Data Byte 9 C27 m27 C17 m17 I/O Input Data Byte 10 C38 m38 C28 m28 I/O Input Data Byte 11 C18 m18 C39 m39 I/O Input Data Byte 12 C29 m29 C19 m19 I/O Input Data Byte 13 C310 m310 C210 m210 I/O Input Data Byte 14 C110 m110 RSVD RSVD The Z-transform equation for the B filter is defined as: –2 HB( z ) = B2 • z + … + B9 • z –9 – 10 B 10 • z + ----------------------------–1 1 – B 11 • z Sample rate = 16 kHz The coefficients for the FIR B section and the gain of the IIR B section are defined as: For i = 2 to 10, B i = C1i • 2 – mli [ 1 + C2i • 2 – m2i ( 1 + C3i • 2 – m3i )] The feedback coefficient of the IIR B section is defined as: B 11 = C111 • 2 – m111 { 1 + C211 • 2 – m211 [ 1 + C311 • 2 – m311 ( 1 + C411 • 2 – m411 )]} Refer to Commands 43, 44 for programming of the B11 coefficients. Power Up and Hardware Reset (RST) Values = 36 AB B8 22 93 AB 2B 6C 46 2C 63 B6 9F 60 (Hex) ( H B ( z ) = – 0.254 • z –2 – 0.891 • z + 0.014 • z –8 –3 – 0.656 • z + 0.013 • z –9 –4 – 0.090 • z –5 + 0.013 • z –6 + 0.017 • z –7 – 10 0.016 • z + ----------------------------------------) –1 1 – 0.97656 • z See note under Commands 31 and 32. RSVD Reserved for future use. Always write as 0, but 0 is not guaranteed when read. SLAC Products 49 39, 40. Write/Read X Filter Coefficients MPI Command (88/89h) R/W = 0: Write R/W = 1: Read Command D7 D6 D5 D4 D3 D2 D1 D0 1 0 0 0 1 0 0 R/W I/O Input Data Byte 1 C40 m40 C30 m30 I/O Input Data Byte 2 C20 m20 C10 m10 I/O Input Data Byte 3 C41 m41 C31 m31 I/O Input Data Byte 4 C21 m21 C11 m11 I/O Input Data Byte 5 C42 m42 C32 m32 I/O Input Data Byte 6 C22 m22 C12 m12 I/O Input Data Byte 7 C43 m43 C33 m33 I/O Input Data Byte 8 C23 m23 C13 m13 I/O Input Data Byte 9 C44 m44 C34 m34 I/O Input Data Byte 10 C24 m24 C14 m14 I/O Input Data Byte 11 C45 m45 C35 m35 I/O Input Data Byte 12 C25 m25 C15 m15 The Z-transform equation for the X filter is defined as: Hx ( z ) = x0 + x1 z –1 + x2 z –2 + x3 z –3 + x4 z –4 + x5 z –5 Sample rate = 16 kHz For i = 0 to 5, the coefficients for the X filter are defined as: Xi = C1i • 2 – m1i { 1 + C2i • 2 – m2i [ 1 + C3i • 2 – m3i ( 1 + C4i • 2 – m4i ) ]} Power Up and Hardware Reset (RST) Values = 0111 0190 0190 0190 0190 0190 (Hex) (Hx(z) = 1) See note under Commands 31 and 32. 50 Am79Q02/021/031 Data Sheet 41, 42. Write/Read R Filter Coefficients MPI Command (8A/8Bh) R/W = 0: Write R/W = 1: Read Command D7 D6 D5 D4 D3 D2 D1 D0 1 0 0 0 1 0 1 R/W I/O Input Data Byte 1 C46 m46 C36 m36 I/O Input Data Byte 2 C26 m26 C16 m16 I/O Input Data Byte 3 C40 m40 C30 m30 I/O Input Data Byte 4 C20 m20 C10 m10 I/O Input Data Byte 5 C41 m41 C31 m31 I/O Input Data Byte 6 C21 m21 C11 m11 I/O Input Data Byte 7 C42 m42 C32 m32 I/O Input Data Byte 8 C22 m22 C12 m12 I/O Input Data Byte 9 C43 m43 C33 m33 I/O Input Data Byte 10 C23 m23 C13 m13 I/O Input Data Byte 11 C44 m44 C34 m34 I/O Input Data Byte 12 C24 m24 C14 m14 I/O Input Data Byte 13 C45 m45 C35 m35 I/O Input Data Byte 14 C25 m25 C15 m15 HR = H IIR • H FIR The Z-transform equation for the IIR filter is defined as: –1 1–z H IIR = --------------------------------–1 1 – ( R6 • z ) Sample rate = 8 kHz The coefficient for the IIR filter is defined as: R 6 = C16 • 2 – ml6 { 1 + C26 • 2 – m26 [ 1 + C36 • 2 – m36 ( 1 + C46 • 2 – m46 )]} The Z-transform equation for the FIR filter is defined as: H FIR ( z ) = R 0 + R 1 z –1 + R2 z –2 + R3 z –3 + R4 z –4 + R5 z –5 Sample rate = 16 kHz For i = 0 to 5, the coefficients for the R2 filter are defined as: R i = C1i • 2 – m1i { 1 + C2i • 2 – m2i [ 1 + C3i • 2 – m3i ( 1 + C4i • 2 – m4i )]} Power Up and Hardware Reset (RST) Values = 2E01 0111 0190 0190 0190 0190 0190 (Hex) (HFIR (z) = 1, R6 = 0.9902) See note under Commands 31 and 32. SLAC Products 51 43, 44. Write/Read B2 Filter Coefficients (IIR) MPI Command (96/97h) R/W = 0: Write R/W = 1: Read Command D7 D6 D5 D4 D3 D2 D1 D0 1 0 0 1 0 1 1 R/W I/O Data Byte 1 C411 m411 C311 m311 I/O Data Byte 2 C211 m211 C111 m111 This function is described in Write/Read B1 Filter Coefficients (FIR) on page 49. Power Up and Hardware Reset (RST) Values = AC01 (Hex) (B11 = 0.97656) See note under Commands 31 and 32. 45, 46. Write/Read Debounce Time Register** MPI Command (C8/C9h) R/W = 0: Write R/W = 1: Read Command I/O Data D7 D6 D5 D4 D3 D2 D1 D0 1 1 0 0 1 0 0 R/W EE1 E1P DSH3 DSH2 DSH1 DSH0 RSVD ECH Enable E1 EE1 = 0* EE1 = 1 E1 multiplexing turned off E1 multiplexing turned on E1 Polarity E1P = 0* E1 is a high-going pulse E1P = 1 E1 is a low-going pulse There is no E1 output unless CMODE = 1. Debounce for Switchhook DSH = 0–15 Debounce period in ms DSH contains the debouncing time (in ms) of the CD1 data (usually switchhook) entering the Real Time Data register described earlier. The input data must remain stable for the debouncing time in order to change the appropriate real time bit. RSVD Reserved for future use. Always write as 0, but 0 is not guaranteed when read. Enable Chopper ECH = 0* ECH = 1 Chopper output (CHCLK) turned off Chopper output (CHCLK) turned on * Power Up and Hardware Reset (RST) Value = 20h. ** This command applies to all channels and does not depend on the state of the Channel Enable Register. 52 Am79Q02/021/031 Data Sheet 47. Read Transmit PCM Data MPI Command (CDh) D7 Command D6 D5 D4 D3 D2 D1 D0 1 1 0 0 1 1 0 1 Output Data Byte 1 XDAT7 XDAT6 XDAT5 XDAT4 XDAT3 XDAT2 XDAT1 XDAT0 Output Data Byte 2 RSVD RSVD RSVD RSVD RSVD RSVD RSVD RSVD RSVD Reserved for future use. Always write as 0, but 0 is not guaranteed when read. Upper Transmit Data XDAT contains A-law or µ-law transmit data in Companded mode. XDAT contains upper data byte in Linear mode with sign in XDAT7. 48, 49. Write/Read FIR Z Filter Coefficients (FIR only) MPI Command (98/99h) R/W = 0: Write R/W = 1: Read This command writes and reads only the FIR filter section without affecting the IIR. Command D7 D6 1 0 D5 D4 0 1 D3 D2 1 0 D1 D0 0 R/W I/O Data Byte 1 C40 m40 C30 m30 I/O Data Byte 2 C20 m20 C10 m10 I/O Data Byte 3 C41 m41 C31 m31 I/O Data Byte 4 C21 m21 C11 m11 I/O Data Byte 5 C42 m42 C32 m32 I/O Data Byte 6 C22 m22 C12 m12 I/O Data Byte 7 C43 m43 C33 m33 I/O Data Byte 8 C23 m23 C13 m13 I/O Data Byte 9 C44 m44 C34 m34 I/O Data Byte 10 C24 m24 C14 m14 The Z-transform equation for the Z filter is defined as: Hz ( z ) = z0 + z1 • z –1 + z2 • z –2 + z3 • z –3 + z4 • z –4 –1 z 5 • z6 • z7 • z + -------------------------------------–1 1 – z7 • z Sample rate = 32 kHz For i = 0 to 5 and 7 z i = C1i • 2 z 6 = C16 • 2 – m1i – m16 { 1 + C2i • 2 – m2i { 1 + C26 • 2 [ 1 + C3i • 2 – m26 – m3i ( 1 + C4i • 2 – m4i )]} } Power Up and Hardware Reset (RST) Values = 0190 0190 0190 0190 0190 0190 01 0190 (Hex) (Hz(z) = 0) See note under Commands 31 and 32. Note: Z6 is used for IIR filter scaling only. Its value is typically greater than zero but less than or equal to one. The input to the IIR filter section is first increased by a gain of 1/Z6, improving dynamic range and avoiding truncation limitations through processing within this filter. The IIR filter output is then multiplied by Z6 to normalize the overall gain. Z5 is the actual IIR filter gain value defined by the programmed coefficients, but it also includes the initial 1/Z6 gain. The theoretical effective IIR gain, without the Z6 gain and normalization, is actually Z5/Z6. SLAC Products 53 50, 51. Write/Read IIR Z Filter Coefficients (IIR only) MPI Command (9A/9Bh) R/W = 0: Write R/W = 1: Read This command writes/reads the IIR filter section only, without affecting the FIR. Command D7 D6 1 0 D5 D4 0 1 D3 D2 1 0 D1 D0 1 R/W I/O Data Byte 11 C45 m45 C35 m35 I/O Data Byte 12 C25 m25 C15 m15 I/O Data Byte 13 C26 m26 C16 m16 I/O Data Byte 14 C47 m47 C37 m37 I/O Data Byte 15 C27 m27 C17 m17 The Z-transform equation for the Z filter is defined as: Hz ( z ) = z0 + z1 • z –1 + z2 • z –2 + z3 • z –3 + z4 • z –4 –1 z 5 • z6 • z7 • z + -------------------------------------–1 1 – z7 • z Sample rate = 32 kHz For i = 0 to 5 and 7 z i = C1i • 2 – m1i z 6 = C16 • 2 { 1 + C2i • 2 – m16 – m2i { 1 + C26 • 2 [ 1 + C3i • 2 – m26 – m3i ( 1 + C4i • 2 – m4i )]} } Power Up and Hardware Reset (RST) Values = 0190 0190 0190 0190 0190 0190 01 0190 (Hex) (Hz(z) = 0) See note under Commands 31 and 32. Note: Z6 is used for IIR filter scaling only. Its value is typically greater than zero but less than or equal to one. The input to the IIR filter section is first increased by a gain of 1/Z6, improving dynamic range and avoiding truncation limitations through processing within this filter. The IIR filter output is then multiplied by Z6 to normalize the overall gain. Z5 is the actual IIR filter gain value defined by the programmed coefficients, but it also includes the initial 1/Z6 gain. The theoretical effective IIR gain, without the Z6 gain and normalization, is actually Z5/Z6. 54 Am79Q02/021/031 Data Sheet 52, 53. Write/Read Ground Key Filter MPI Command (E8/E9h) R/W = 0: Write R/W = 1: Read Command I/O Data D7 D6 D5 D4 D3 D2 D1 D0 1 1 1 0 1 0 0 R/W RSVD RSVD RSVD RSVD GK3 GK2 GK1 GK0 Filter Ground Key GK = 0–15 Filter sampling period in 1 ms GK contains the filter sampling time (in ms) of the CD1B data (usually Ground Key) or CD2 entering the Real Time Data register described earlier. A value of 0 disables the Ground Key filter for that particular channel. Power Up and Hardware Reset (RST) Value = 00h. RSVD Reserved for future use. Always write as 0, but 0 is not guaranteed when read. SLAC Products 55 PROGRAMMABLE FILTERS General Description of CSD Coefficients Mi = the number of shifts = Mi ≤ Mi + 1 Bi = sign = ±1 N = number of CSD coefficients. The filter functions are performed by a series of multiplications and accumulations. A multiplication occurs by repeatedly shifting the multiplicand and summing the result with the previous value at that summation node. The method used in the QSLAC device is known as Canonic Signed Digit (CSD) multiplication and splits each coefficient into a series of CSD coefficients. hi in Equation 4 represents a decimal number, broken down into a sum of successive values of: Each programmable FIR filter section has the following general transfer function: HF ( z ) = h 0 + h 1 z –1 + h2 z –2 + … + hn z –n The transfer function for the IIR part of Z and B filters: Equation 2 The transfer function of the IIR part of the R filter is: –1 – M1 + B2 2 – M2 + … + BN 2 – MN ±1.0 multiplied by 1, or 1/2, or 1/4 … 1/128 … When M1 is 0, the value is a binary 1 in front of the decimal point, that is, no shift. If M2 is also 0, the result is another binary 1 in front of the decimal point, giving a total value of binary 10 in front of the decimal point (i.e., a decimal value of 2.0). The value of N, therefore, determines the range of values the coefficient hi can take (e.g., if N = 3 the maximum and minimum values are ±3, and if N = 4 the values are between ±4). Equation 3 The values of the user-defined coefficients (h i) are assigned via the MPI. Each of the coefficients (hi) is defined in the following general equation: h i = B1 2 2) The coefficient hi in Equation 4 is a value made up of N binary 1s in a binary register where the left part represents whole numbers, the right part decimal fractions, and a decimal point separates them. The first binary 1 is shifted M1 bits to the right of the decimal point; the second binary 1 is shifted M2 bits to the right of the decimal point; the third binary 1 is shifted M3 bits to the right of the decimal point, and so on. where the number of taps in the filter = n + 1. 1–z HI ( z ) = ------------------------------–1 1 – h( n + 1) z ±1.0 multiplied by 2–0, or 2–1, or 2–2 … 2–7 … The limit on the negative powers of 2 is determined by the length of the registers in the ALU. Equation 1 1 HI ( z ) = ------------------------------–1 1 – h( n + 1) z 1) Equation 4 where: Detailed Description of QSLAC Device Coefficients The CSD coding scheme in the QSLAC device uses a value called mi, where m1 represents the distance shifted right of the decimal point for the first binary 1. m2 represents the distance shifted to the right of the previous binary 1, and m3 represents the number of shifts to the right of the second binary 1. Note that the range of values determined by N is unchanged. Equation 4 is now modified (in the case of N = 4) to: hi = B1 2 – m1 h i = C1 • 2 + B2 2 – m1 – m2 + B3 2 – m3 + C1 • C 2 • 2 + B4 2 – ( m1 + m2 ) Equation 5 – m4 + C1 • C 2 • C 3 • 2 – ( m1 + m2 + m3 ) + C1 • C2 • C3 • C4 • 2 – ( m1 + m2 + m3 + m4 ) Equation 6 h i = C1 • 2 – m1 { 1 + C2 • 2 – m2 [ 1 + C3 • 2 – m3 ( 1 + C4 • 2 – m4 )] } where: M1 = m1 M2 = m1 + m2 M3 = m1 + m2 + m3 M4 = m1 + m2 + m3 + m4 56 B1 = C1 B2 = C1 • C2 B3 = C1 • C2 • C3 B4 = C1 • C2 • C3 • C4 Am79Q02/021/031 Data Sheet Equation 7 In the QSLAC device, a coefficient, hi, consists of N CSD coefficients, each being made up of 4 bits and formatted as Cxy mxy, where Cxy is 1 bit (MSB) and mxy is 3 bits. Each CSD coefficient is broken down as follows: User Test States and Operating Conditions The QSLAC device supports testing by providing test states and special operating conditions as shown in Figure 9 (see Operating Conditions register). Cxy is the sign bit (0 = positive, 1 = negative). mxy is the 3-bit shift code. It is encoded as a binary number as follows: 000: 0 shifts 001: 1 shifts 010: 2 shifts 011: 3 shifts 100: 4 shifts 101: 5 shifts 110: 6 shifts 111: 7 shifts Lower Receive Gain (LRG): When LRG = 1, an extra 6.02 dB of loss is inserted into the receive path. y is the coefficient number (the i in hi). Arm Transmit Interrupt (ATI) and Read Transmit PCM Data: The read transmit PCM data command, Command 47, can be used to read transmit PCM data through the microprocessor interface. If the ATI bit is set, an interrupt will be generated whenever new transmit data appears in the channel and will be cleared when the data is read. When combined with Tone Generation and Loopback states, this allows the microprocessor to test channel integrity. Cutoff Transmit Path (CTP): When CTP = 1, DX and TSC are high impedance and the transmit time slot does not exist. This state takes precedence over the TSA Loopback (TLB) and Full Digital Loopback (FDL) states. Cutoff Receive Path (CRP): When CRP = 1, the receive signal is forced to 0 just ahead of the low pass filter (LPF) block. This state also blocks Full Digital Loopback (FDL), the 1 kHz receive tone, and the Bfilter path. High Pass Filter Disable (HPF): When HPF = 1, all of the high pass and notch filters in the transmit path are disabled. x is the position of this CSD coefficient within the hi coefficient. The most significant binary 1 is represented by x = 1. The next most significant binar y 1 is represented by x = 2, and so on. Thus, C13 m13 represents the sign and the relative shift position for the first (most significant) binary 1 in the 4th (h3) coefficient. The number of CSD coefficients, N, is limited to 4 in the GR, GX, R, X, and Z filters; 4 in the IIR part of the B filter; 3 in the FIR part of the B filter; and 2 in the post-gain factor of the Z-IIR filter. The GX filter coefficient equation is slightly different from the other filters. h iGX = 1 + h i Equation 8 Please refer to the section detailing the commands for complete details on programming the coefficients. TSA Loopback (TLB): When TLB = 1, data from the TSA receive path is looped back to the TSA transmit path. Any other data in the transmit path is overwritten. Full Digital Loopback (FDL): When FDL = 1, the VOUT output is turned off and the analog output voltage is routed to the input of the receive path, replacing the voltage from VIN. The AISN path is temporarily turned off. This test state can also be entered by writing the code 10000 into the AISN register. 1 kHz Receive Tone (TON): When TON = 1, a 1 kHz digital milliwatt is injected into the receive path, replacing any receive signal from the TSA. SLAC Products 57 A-Law and µ-Law Companding Table 2 and Table 3 show the companding definitions used for A-law and µ-law PCM encoding. Table 2. A-Law: Positive Input Values 1 Segment Number 2 3 4 # Intervals Value at x Interval Segment Size End Points Decision Value Number n 5 6 7 Character Signal pre Quantized Decision Inversion of Value (at Even Bits Value xn Decoder (See Note 1) Output) yn Bit No. 8 Decoder Output Value No. 12345678 4096 7 (128) (4096) 127 3968 113 2176 112 2048 16 x 128 11111111 4032 128 2112 113 1056 97 528 81 264 65 132 49 66 33 1 1 See Note 2 2048 11110000 See Note 2 6 16 x 64 1024 97 1088 96 1024 11100000 See Note 2 5 16 x 32 512 81 544 80 512 11010000 See Note 2 4 16 x 16 256 65 272 64 256 11000000 See Note 2 3 16 x 8 128 49 136 48 128 10110000 See Note 2 2 16 x 4 64 33 68 32 64 10100000 See Note 2 1 32 x 2 1 2 0 0 10000000 Notes: 1. 4096 normalized value units correspond to TMAX = 3.14 dBm0. 2. The character signals are obtained by inverting the even bits of the signals of column 6. Before this inversion, the character signal corresponding to positive input values between two successive decision values numbered n and n+1 (see column 4) is 128+n, expressed as a binary number. x +x 2 n–1 n 3. The value at the decoder output is y n = ---------------------- , for n = 1,...127, 128. 4. x128 is a virtual decision value. 5. Bit 1 is a 0 for negative input values. 58 Am79Q02/021/031 Data Sheet Table 3. µ-Law: Positive Input Values 1 Segment Number 2 3 # Intervals Value at x Interval Segment Size End Points 4 Decision Value Number n 5 6 7 Character Signal pre Quantized Decision Inversion of Value (at Even Bits Value xn Decoder (See Note 1) Output) yn Bit No. 8 Decoder Output Value No. 12345678 8159 8 (128) (8159) 127 7903 113 4319 112 4063 16 x 256 10000000 8031 127 4191 112 2079 96 1023 80 495 64 231 48 99 32 33 16 11111110 2 1 11111111 0 0 See Note 2 4063 10001111 See Note 2 7 16 x 128 2015 97 2143 96 2015 10011111 See Note 2 6 16 x 64 991 81 1055 80 991 10101111 See Note 2 5 16 x 32 479 65 511 64 479 10111111 See Note 2 4 16 x 16 223 49 239 48 223 11001111 See Note 2 3 16 x 8 95 33 103 32 95 11011111 See Note 2 2 16 x 4 31 17 35 16 31 11101111 See Note 2 1 15 x 2 2 3 1 1 0 0 1x1 Notes: 1. 8159 normalized value units correspond to TMAX = 3.17 dBm0. 2. The character signal corresponding to positive input values between two successive decision values numbered n and n+1 (see column 4) is 255-n, expressed as a binary number. x +x n+1 n - , for n = 1, 2,...127. 3. The value at the decoder is y0 = x0 = 0 for n = 0, and y n = ---------------------- 4. x128 is a virtual decision value. 2 5. Bit 1 is a 0 for negative input values. SLAC Products 59 APPLICATIONS The QSLAC device performs a programmable codec/ filter function for four telephone lines. It interfaces to the telephone lines through an Legerity SLIC device or a transformer with external buffering. The QSLAC device provides latched digital I/O to control and monitor four SLICs and provides access to time-critical information, like off/on-hook and ring trip, for all four channels via a single read operation. When various country or transmission requirements must be met, the QSLAC device enables a single SLIC design for multiple applications. The line characteristics (such as apparent impedance, attenuation, and hybrid balance) can be modified by programming each QSLAC device channel’s coefficients to meet desired performance. The QSLAC device requires an external buffer to drive transformer SLICs. Connection to a PCM back plane is implemented by means of a simple buffer chip. Several QSLAC devices can be tied together in one bus interfacing the back plane through a single buffer. An intelligent bus interface chip is not required because each QSLAC device provides its own buffer control (TSXA/B). The Q S L AC d ev i c e i s c o n t r o l l e d t h r o u g h t h e microprocessor interface, either by a microprocessor on the linecard or by a central processor. Controlling the SLIC The A m79 Q0 21 Q S LAC dev ic e ha s fi ve TTL compatible I/O pins (CD1, CD2, C3 to C5) for each channel. The Am79Q031 QSLAC device has only CD1 and CD2 available. The outputs are programmed using Command 19, and the status is read back using Command 20. CD1 and CD2 for all four channels can be read back using Command 16. The direction of the I/O pins (input or output) is specified by programming the SLIC I/O direction register (Commands 21 and 22). Default Filter Coefficients The default filter coefficients were calculated assuming an Am7920 SLIC with 50 Ω protection resistors, a 178 kΩ transversal impedance (ZT ), and a 90.5 k Ω receive impedance (ZRX). This SLIC has a transmit gain of 0.5 (GTX) and a current gain of 500 (K1). The transmit relative level was set to +0.28 dBr, and the receive 60 relative level was set to –4.39 dBr. The equalization filters (X and R) were not optimized. The balance filter was designed to give acceptable balance into a variety of impedances. The nominal input impedance was set to 815 Ω. If the SLIC circuit differs significantly from this design, the default filters cannot be used and must be replaced by programmed coefficients. Calculating Coefficients with WinSLAC Software The WinSLAC software is a program that models the QSLAC device, the line conditions, the SLIC, and the linecard components to obtain the coefficients of the programmable filters of the QSLAC device and some of the transmission performance plots. The following parameters relating to the desired line conditions and the components/circuits used in the linecard are to be provided as input to the program: 1. Line impedance or the balance impedance of the line is specified by the local PTT. 2. Desired two-wire impedance that is to appear at the linecard terminals of the exchange. 3. Tabular data for templates describing the frequency response and attenuation distortion of the design. 4. Relative analog signal levels for both the transmit and receive two-wire signals. 5. Component values and SLIC device selection for the analog portion of the line circuits. 6. Two-wire return loss template is usually specified by the local PTT. 7. Four-wire return loss template is usually specified by the local PTT. The output from the WinSLAC program includes the coefficients of the GR, GX, Z, R, X, and B filters as well as transmission performance plots of two-wire return loss, receive and transmit path frequency responses, and four-wire return loss. The software supports the use of the Legerity SLICs or allows entry of a SPICE netlist describing the behavior of any type of SLIC circuit. Am79Q02/021/031 Data Sheet PHYSICAL DIMENSIONS PL032 Dwg rev AH; 10/99 SLAC Products 61 PL044 Dwg rev. AN; 8/99 62 Am79Q02/021/031 Data Sheet PQT044 Dwg rev AS; 08/99 SLAC Products 63 REVISION SUMMARY Revision B to Revision C • In the Connection Diagrams section, “INT” was changed to “INT” for Am79Q021JC and Am79Q021VC. • “Frame sync” information was added to the first paragraph on page 31. Revision C to Revision D • Minor changes were made to the data sheet style and format to conform to Legerity standards. • Deleted the Am79Q02VC package and all references to it. • The physical dimensions (PL032, PL044, and PQT044) were added to the Physical Dimensions section. • In the Pin Description table, the second sentence was deleted from the MCLK/E1 row and the second to last sentence was deleted from the PCLK row. • On page 20, row 29 was deleted. • On pages 23–24, the reference to “29” was deleted. • In the Clock Mode Operation section on page 25, the second to last sentence was deleted. Revision D to Revision E • Page 59, Table 3, changed values in column 7. Revision E to Revision F • All the physical dimensions were updated. Revision F to Revision G • Page 28, deleted “Old Flag (CMD 47, Bit 0)” from Figure 9. • Page 53, “47. Read Transmit PCM Data”. Changed last row, last column from OLD to RSVD. Deleted text “Old Data Flag...” • Electrical characteristics table: “pk” added to units for Iout. Revision G to Revision H • 64 In the “Functional Description” section, deleted “VC” as an option for the Am79Q02 QSLAC device in the table listing the different configurations available. Am79Q02/021/031 Data Sheet Notes: www.legerity.com Legerity provides silicon solutions that enhance the performance, speeds time-to-market, and lowers the system cost of our customers' products. By combining process, design, systems architecture, and a complete set of software and hardware support tools with unparalleled factory and worldwide field applications support, Legerity ensures its customers enjoy a smoother design experience. It is this commitment to our customers that places Legerity in a class by itself. The contents of this document are provided in connection with Legerity, Inc. products. Legerity makes no representations or warranties with respect to the accuracy or completeness of the contents of this publication and reserves the right to make changes to specifications and product descriptions at any time without notice. No license, whether express, implied, arising by estoppel or otherwise, to any intellectual property rights is granted by this publication. Except as set forth in Legerity's Standard Terms and Conditions of Sale, Legerity assumes no liability whatsoever, and disclaims any express or implied warranty, relating to its products including, but not limited to, the implied warranty of merchantability, fitness for a particular purpose, or infringement of any intellectual property right. Legerity's products are not designed, intended, authorized or warranted for use as components in systems intended for surgical implant into the body, or in other applications intended to support or sustain life, or in any other application in which the failure of Legerity's product could create a situation where personal injury, death, or severe property or environmental damage may occur. Legerity reserves the right to discontinue or make changes to its products at any time without notice. © 2001 Legerity, Inc. All rights reserved. Trademarks Legerity, the Legerity logo and combinations thereof, DSLAC, QSLAC, SLAC, and WinSLAC are trademarks of Legerity, Inc. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies. Americas Mailing: P.O. Box 18200 Austin, TX 78760-8200 Shipping: 4509 Freidrich Lane Austin, TX 78744-1812 ATLANTA 6465 East Johns Crossing, Suite 400 Duluth, GA USA 30097 MainLine: 770-814-4252 Fax: 770-814-4253 AUSTIN 4509 Freidrich Lane Austin, TX USA 78744-1812 MainLine: 512-228-5400 Fax: 512-228-5510 BOSTON 6 New England Executive Park Suite 400 Burlington, MA USA 01803 MainLine: 781-229-7320 Fax: 781-272-3706 CHICAGO 8770 W. Bryn Mawr, Suite 1300 Chicago, IL USA 60631 MainLine: 773-867-8034 Fax: 773-867-2910 DALLAS 4965 Preston Park Blvd., Suite 280 Plano, TX USA 75093 MainLine: 972-985-5474 Fax: 972-985-5475 HUNTSVILLE 600 Boulevard South, Suite 104 Huntsville, AL USA 35802 MainLine: 256-705-3504 Fax: 256-705-3505 IRVINE 1114 Pacifica Court, Suite 250 Irvine, CA USA 92618 MainLine: 949-753-2712 Fax: 949-753-2713 NEW JERSEY 3000 Atrium Way, Suite 270 Mt. Laurel, NJ USA 08054 MainLine: 856-273-6912 Fax: 856-273-6914 OTTAWA 600 Terry Fox Drive Ottawa, Ontario, Canada K26 4B6 MainLine: 613-599-2000 Fax: 613-599-2002 RALEIGH 2500 Regency Parkway, Suite 226 Cary, NC USA 27511 MainLine: 919-654-6843 Fax: 919-654-6781 SAN JOSE 1740 Technology Drive, Suite 290 San Jose, CA USA 95110 MainLine: 408-573-0650 Fax: 408-573-0402 Telephone: (512) 228-5400 Fax: (512) 228-5510 North America Toll Free: (800) 432-4009 Worldwide Sales Offices Asia Europe HONG KONG Units 2401-2, 24th Floor Jubilee Centre, 18 Fenwick Street Wanchai, Hong Kong MainLine: 852-2864-8300 Fax: 852-2866-1323 BELGIUM Baron Ruzettelaan 27 8310 Brugge Belgium MainLine: 32-50-28-88-10 Fax: 32-50-27-06-44 KOREA 135-090 18th Fl., Kyoung Am Bldg 157-26, Samsung-dong, Kangnam-ku Seoul, Korea MainLine: 82-2-565-5951 Fax: 82-2-565-3788 FRANCE 7, Avenue G. Pompidou Suite 402 92300 Levallois-Perret, France MainLine: 33-1-47-48-2206 Fax: 33-1-47-48-2568 SHANGHAI Shanghai P.O. Box 232022 Shanghai PR China 200232 MainLine: 86-21-54233253 Fax: 86-21-54233254 GERMANY Elisabethstrasse 89-91 80797 München, Germany MainLine: 49-89-5908-0 Fax: 49-89-5908-1308 SHENZHEN Room 310, Tower 9 Jinxiu Street 30 Futian District Shenzhen, PR China 518040 MainLine: 86-755-3706-667 Fax: 86-755-3706-520 ITALY Via F. Rosselli 3/2 20019 Settimo Mse, Milano Italy MainLine: 39-02-3355521 Fax: 39-02-33555232 SINGAPORE Serangoon Central Post Office P.O. Box 537 Singapore 915502 MainLine: 65-2803267 Fax: 65-2855869 TOKYO Shinjuku NS Bldg. 5F 2-4-1 Nishi Shinjuku, Shinjuku-ku Tokyo, Japan 163-0805 MainLine: 81-3-5339-2011 Fax: 81-3-5339-2012 SWEDEN Frösundaviks Allé 15, 4tr SE-16970 Solna Sweden MainLine: 46-8-509-045-45 Fax: 46-8-509-046-36 UK Regus House, Windmill Hill Business Park Whitehill Way SN5 6QR Swindon Wiltshire UK MainLine: 44-(0)1793-441408 Fax: 44-(0)1793-441608 To download or order product literature, visit our website at www.legerity.com. To order literature in North America, call: (800) 572-4859 or 512-349-3193 or email: [email protected] To order literature in Europe or Asia, call: 44-0-1179-341607 or email: Europe — [email protected] Asia — [email protected]