Text preview for : msp3410b.pdf part of ITT Intermetall MSP 3410 B MSP 3410 B Multistandard Sound Processor



Back to : msp3410b.rar | Home

PRELIMINARY DATA SHEET

MSP 3410 B Multistandard Sound Processor

Edition Nov. 20, 1995 6251-366-9PD

MSP 3410 B
Contents Page 4 5 5 5 5 6 6 6 9 9 9 9 10 10 10 10 10 11 11 11 11 11 12 13 14 14 14 14 14 14 14 15 15 15 16 16 17 17 18 18 19 20 2 Section 1. 2. 2.1. 2.2. 2.3. 3. 3.1. 3.2. 4. 4.1. 4.1.1. 4.1.2. 4.1.3. 4.1.4. 4.1.5. 4.1.6. 4.1.7. 4.1.8. 4.1.9. 4.1.10. 4.2. 4.3. 4.4. 5. 5.1. 5.2. 5.2.1. 5.2.2. 5.2.3. 5.2.4. 5.3. 6. 7. 8. 9. 10. 11. 11.1. 11.2. 11.2.1. 11.2.2. 11.2.3. Title Introduction Features of the MSP 3410 B Features of the Demodulator and Decoder Sections Features of the DSP-Section Features of the Analog Section Application Fields of the MSP 3410 B NICAM plus FM-Mono German 2-Carrier System (DUAL FM System) Architecture of the MSP 3410 B Demodulator Block Analog Sound IF ­ Input Section Quadrature Mixers Lowpass Filtering Block for Mixed Sound IF Signals CORDIC Block Differentiate Lowpass Filter Block for Demodulated Signals High Deviation FM Mode MSP-Mute Function in the Dual Carrier FM Mode DQPSK-Decoder NICAM-Decoder Analog Section and SCART Switches MSP 3410 B Audio Baseband Processing Dual Carrier FM Stereo/Bilingual Detection Control Bus Interface Protocol Description Proposal for MSP 3410 B I2C Telegrams Symbols Write Telegrams Read Telegrams Examples Start Up Sequence N-Bus Interface Pay-TV Interface Audio PLL and Crystal Specifications S-Bus Interface I2S Bus Interface Programming the Demodulator Part Write Registers: Table and Addresses Write Registers: Functions and Values Setting of Parameter AD_CV Control Register 'MODE_REG' FIR-Parameter

PRELIMINARY DATA SHEET

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B

Contents, continued Page 21 22 22 23 24 24 24 24 24 25 25 26 26 31 32 34 34 35 38 39 41 41 42 47 54 54 54 55 56 57 58 60 61 63 63 63 63 64 65 Section 11.2.4. 11.3. 11.4. 11.5. 11.6. 11.6.1. 11.6.2. 11.6.3. 11.6.4. 11.6.5. 11.6.6. 12. 12.1. 12.2. 12.3. 13. 13.1. 13.2. 13.3. 13.4. 13.5. 13.5.1. 13.5.2. 13.5.3. 14. 14.1. 14.2. 14.3. 14.4. 15. 16. 17. 18. 19. 19.1. 19.2. 19.3. 20. 21. Title DCO-Increments Read Registers: Listing and Addresses Read Registers: Functions and Values Sequences to Transmit Parameters and Start of Processing Software Proposals for Multistandard TV-Sets Multistandard Including System B/G with NICAM/FM-Mono only Multistandard Including System I with NICAM/FM-Mono only Multistandard Including System B/G with NICAM/FM-Mono and German DUAL FM Satellite Mode Automatic Search Function for FM-Carrier Detection Automatic Standard Detection Programming the Audio Processing Part Summary of the DSP Control Registers Exclusions Summary of Readable Registers Specifications Outline Dimensions Pin Connections and Descriptions Pin Configuration Pin Circuits Electrical Characteristics Absolute Maximum Ratings Recommended Operating Conditions Characteristics Timing Diagrams Power-up Sequence I2C Bus Timing Diagram I2S Bus Timing Diagram SBUS Timing Diagram Application Circuit DMA Application I2S Bus in Master/Slave Configuration with Standby Mode APPENDIX A: MSP 3410/3400B Technical Code History APPENDIX B: Documentation History MSP 3400 MSP 3410 and MSP 3400 MSP 3410 B and MSP 3400 B APPENDIX C: Documentation of known hardware restrictions for TC15 Index

ITT Semiconductors

3

MSP 3410 B
Multistandard Sound Processor Release Notes:

PRELIMINARY DATA SHEET

The hardware description in this document is valid for the MSP 3410 B version F7 and following versions. The suffix "B" in the name denotes the requirements of the crystal with modified specifications. For a brief history survey, please see appendix "MSP 3410 B Technical Code History". The present document is version 0.8. Revision bars indicate significant changes to revision 0.7.

1. Introduction The MSP 3410 B is a single-chip Multistandard Sound Processor for applications in analog and digital TV sets, satellite receivers and video recorders. The MSP-family, which goes back to the MSP 2400, demonstrates in an impressive way the progressive development towards highly integrated ICs, offering more and more features and flexibility. The development of the MSP 2410 included an automatic gain control but reduced the amount of external components. The MSP 2410 reached a high level of performance and is the basis for the new generation.

The MSP 3410 B increases function integration in a spectacular way. By including the MSP2410 as a library cell and combining it with AD/DA converters and high performance digital signal processing, the chip offers a wide range of features. The complete TV-sound-processing, starting at the Sound-IF domain, will be performed by one single IC. The inputs of the IC are analog audio signals in baseband and at intercarrier position. The MSP 3410 B covers the sound processing of a wide range of TV-standards. Some examples are listed in Table 3­1. The MSP 3410 B is produced in 1.0 µm CMOS technology and is available in 68-pin PLCC and in 64-pin PSDIP packages.

SBUS

I2S

I2C

4

4

2

Sound IF 1 Sound IF 2 MONO IN SCART1 IN SCART2 IN SCART3 IN 2 2 2

2 2

LOUDSPEAKER OUT HEADPHONE OUT

MSP 3410 B

2 2

SCART1 OUT SCART2 OUT

Fig. 1­1: Main I/O Signals MSP 3410 B

4

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
2.2. Features of the DSP-Section ­ flexible selection of audio sources to be processed

2. Features of the MSP 3410 B 2.1. Features of the Demodulator and Decoder Sections The MSP 3410 B is designed to simultaneously perform digital demodulation and decoding of NICAM-coded TV stereo sound, as well as demodulation of FM-mono TV sound. Alternatively, two carrier FM systems according to the German or Korean terrestrial specs or the satellite specs can be processed with the MSP 3410 B. Since it is simple and economic to demodulate AM sound carriers with conventional sound-IF-mixing units, the AM demodulation feature of the MSP will seldom be used. However, for FM carrier detection in satellite operation the AM demodulation offers a powerful feature to calculate the carrier field strength, which can be used for automatic search algorithms. So the IC facilitates a first step towards multistandard capability with its very flexible application and may be used in TV-sets, satellite tuners and video recorders. The MSP 3410 B facilitates profitable multistandard capability, offering the following advantages: ­ two selectable analog inputs (TV- and SAT-IF sources) ­ Automatic Gain Control (AGC) for analog input: input range: 0.14 ­ 3 Vpp ­ integrated A/D converter for sound-IF inputs ­ all demodulation and filtering is performed on chip and is individually programmable ­ simple realization of both digital NICAM standards (UK/Scandinavia) ­ no external filter hardware is required ­ only one crystal clock (18.432 MHz) is necessary ­ Pay-TV for NICAM-mode ­ FM carrier level calculation for automatic search algorithms and carrier mute function ­ high deviation FM-mono mode (max. deviation: approx. ±360 kHz)

­ digital input and output interfaces via S-Bus for DMAvia AMU, and via I2S-Bus for external DSP-Processors featuring Graphic Equalizer, Surround Sound etc. ­ performance of all deemphasis systems including adaptive Wegener Panda 1 without external components or controlling ­ performance of D2MAC audio together with an AMU 2481 ­ digitally performed FM-identification decoding and dematrixing ­ digital baseband processing: volume, bass, treble, pseudostereo and basewidth enlargement ­ simplified controlling of volume, bass, treble etc. ­ increased audio bandwidth for FM-Audio-signals (20 Hz ­ 15 kHz , ±1 dB) 2.3. Features of the Analog Section ­ three selectable analog pairs of audio baseband inputs (=three SCART inputs) Input level: 2 V RMS; input impedance: 25 k ­ one selectable analog mono input (i.e. AM sound); Input level: 2 V RMS; input impedance: 10 k ­ two high quality A/D converters; S/N-Ratio: 85 dB ­ 20 Hz to 20 kHz Bandwidth for SCART-to-SCARTCopy facilities ­ MAIN (loudspeaker) and AUX (headphones): two pairs of 4-fold oversampled D/A-converters Output level per channel: max. 1.4 VRMS Output resistance: max. 5 k S/N-Ratio: 85 dB at maximum volume max. noise voltage in mute mode: 10 µV (BW: 20 Hz ... 16 kHz) ­ one pair of four-fold oversampled D/A-converters supplying two selectable pairs of SCART-Outputs. Output level per channel: max. 2 V RMS, output resistance: max. 0.5 k, S/N-Ratio: 85 dB (20 Hz ... 16 kHz)

ITT Semiconductors

5

MSP 3410 B
3. Application Fields of the MSP 3410 B In the following sections, a brief overview about the two main TV sound standards, NICAM 728 and German FMStereo, demonstrates the complex requirements of a multistandard audio IC. 3.1. NICAM plus FM-Mono According to the British, Scandinavian and Spanish TVstandards, high quality stereo sound is transmitted digitally. The systems allow two high quality digital sound channels to be added to the already existing FM channel. The sound coding follows the format of the so-called Near Instantaneous Companding System (NICAM 728). Transmission is performed using Differential Quadrature Phase Shift Keying (DQPSK). Table 3­2 gives some specifications of the sound coding (NICAM); Table 3­3 offers an overview of the modulation parameters.

PRELIMINARY DATA SHEET

In the case of NICAM/FM mode there are three different audio channels available: NICAM A,NICAM B and FMmono. NICAM A and B may belong either to a stereo or to a dual language transmission. Information about operation mode and about the quality of the NICAM signal can be read by the CCU via the control bus. In the case of low quality (high bit error rate) the CCU may decide to switch to the analog FM-mono sound.

3.2. German 2-Carrier System (DUAL FM System) Since September 1981, stereo and dual sound programs have been transmitted in Germany using the 2-carrier system. Sound transmission consists of the already existing first sound carrier and a second sound carrier additionally containing an identification signal. Some more details of this standard are given in Table 3­4.

Table 3­1: European TV standards TV-System B/G B/G L I D,K M Satellite Satellite Position of Sound Carrier /MHz 5.5/5.74 5.5/5.85 6.5/5.85 6.0/6.552 6.5 4.5 6.5 7.02/7.2 Sound Modulation FM-Stereo FM-Mono/NICAM AM-Mono/NICAM FM-Mono/NICAM FM-Mono FM-Mono FM-Mono FM-Stereo Color System PAL PAL SECAM PAL SECAM NTSC PAL PAL Country Germany Scandinavia,Spain France UK USSR USA Europe (ASTRA) Europe (ASTRA)

Table 3­2: Summary of NICAM 728 sound coding characteristics Characteristics Audio sampling frequency Number of channels Initial resolution Companding characteristics Coding for compressed samples Preemphasis Audio overload level Values 32 kHz 2 14 bit/sample near instantaneous, with compression to 10 bits/sample in 32-samples (1 ms) blocks 2's complement CCITT Recommendation J.17 (6.5 dB attenuation at 800 Hz) +12 dBm0 measured at the unity gain frequency of the preemphasis network (2 kHz)

6

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B

Table 3­3: Summary of NICAM 728 sound modulation parameters Specification Carrier frequency of digital sound Transmission rate Type of modulation Spectrum shaping Roll-off Roll off factor 1.0 Carrier frequency of analog sound component Power ratio between vision carrier and analog sound carrier Power ratio between analog and modulated digital sound carrier 6.0 MHz FM mono 10 dB 10 dB UK 6.552 MHz Scandinavia/Spain 5.85 MHz 728 kBit/s 1 part/million Differentially encoded quadrature phase shift keying (DQPSK) by means of Roll-off filters 0.4 5.5 MHz FM mono 13 dB 7 dB 0.4 6.5 MHz AM mono terrestric 10 dB 17 dB cable 16 dB 11 dB France 5.85 MHz

Table 3­4: Key parameters for German 2-carrier sound system Sound Carriers Intercarrier frequencies Vision/sound power difference Sound bandwidth Pre-emphasis Frequency deviation Sound Signal Components Mono transmission Stereo transmission Dual sound transmission mono (L+R)/2 language A mono R language B Channel FM1 5.5 MHz 13 dB Channel FM2 5.7421875 MHz 20 dB 40 Hz to 15 kHz 50 µs ±50 kHz

Identification of Transmission Mode on Channel 2 Pilot carrier frequency Type of modulation Modulation depth Modulation frequency 54.6875 kHz AM 50% mono: stereo: dual: unmodulated 117.5 Hz 274.1 Hz

ITT Semiconductors

7

MSP 3410 B
33 34 39 MHz 5 9 MHz

PRELIMINARY DATA SHEET

SAW Filter Sound IF Mixer

Sound IF Filter

Tuner

Loudspeaker

Vision Demodulator

AM Sound

MSP 3410 B
SCART1
2 2 2 2 2

Headphone

SCART1 SCART2

Composite Video

SCART Inputs

SCART Outputs

SCART2 SCART3

I 2S optional Feature Processor

SBUS

AMU

DMA

According to the mixing characteristics of the Sound-IF mixer, the Sound-IF filter may be omitted. . Fig. 3­1: Typical MSP 3410 B application

8

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
sound IF mixer ICs however show large picture components on their outputs. In this case filtering is recommended. It was found, that the high pass filters formed by the coupling capacitors at pins ANA_IN1+ and ANA_IN2+ (as shown in the application diagram) are sufficient in most cases. 4.1.2. Quadrature Mixers The digital input coming from the integrated A/D converter may contain audio information at a frequency range of theoretically 0 to 9 MHz corresponding to the selected standards. By means of two programmable quadrature mixers two different audio sources, for example NICAM and FM-mono, may be shifted into baseband position. In the following, the two main channels are provided to process either: ­ NICAM (channel 1) and FM mono (channel 2) simultaneously or, alternatively, ­ FM2 (channel 1) and FM1 (channel 2). Two independent digital oscillators are provided to generate two pairs of sin/cos-functions. Two programmable increments, to be divided up into Low- and High Part, determine frequency of the oscillator, which corresponds to the frequency of the desired audio carrier. In section 11.1., format and values of the increments are listed.

4. Architecture of the MSP 3410 B Fig. 4­1 shows a simplified block diagram of the IC. Its architecture is split into three functional blocks: 1. demodulator and decoder section 2. digital signal processing (DSP) section performing audio baseband processing 3. analog section containing two A/D-converters, 6 D/A-converters, and channel selection 4.1. Demodulator Block 4.1.1. Analog Sound IF ­ Input Section The input pins ANA_IN1+, ANA_IN2+ and ANA_IN­ offer the possibility to connect two different sound IF sources to the MSP 3410 B. By means of bit [8] of AD_CV (see Table 11­2) either terrestrial or satellite sound IF signals can be selected. The analog-to-digital conversion of the preselected sound IF signal is done by a flash-converter, whose output can be used to control an analog automatic gain circuit (AGC), providing optimum level for a wide range of input levels. It is possible to switch between automatic gain control and a fixed (setable) input gain. In the optimum case, the input range of the AD converter is completely covered by the sound if source. Some combinations of SAW filters and

S_DA_OUT S_CL S_ID S_DA_IN

I2S_DA_OUT I2S_CL I2S_WS I2S_DA_IN

SBUS Interface

I2S Interface
S1...4 I2SL/R I2SL/R LOUDSPEAKER L LOUDSPEAKER R

Sound IF
ANA_IN1+ ANA_IN2+

Demodulator

FM1 FM2 NICAM A NICAM B

D/A D/A

DACM_L

Loudspeaker
DACM_R

DFP
Mono
MONO_IN

IDENT

IDENT HEADPHONE L

D/A D/A

DACA_L

Headphone
DACA_R

SC1_IN_L

HEADPHONE R

SCART1
SC1_IN_R

A/D A/D

SCART_L

SC2_IN_L

SCART_R

SCART2
SC2_IN_R

SCART_L SCART_R

D/A D/A

SC1_OUT_L

SCART 1
SC1_OUT_R SC2_OUT_L

SC3_IN_L

SCART3
SC3_IN_R

SCART Switching Facilities

SCART 2
SC2_OUT_R

Fig. 4­1: Architecture of the MSP 3410 B ITT Semiconductors 9

MSP 3410 B

PRELIMINARY DATA SHEET

N_DA N_CL DCO1 MODE_REG[6,7,10] Oscillator FIR_REG_1 Phase Mixer VREFTOP Lowpass CORDIC Differentiator Mute Lowpass FM2 DQPSK Decoder NICAM Decoder NICAMA NICAMB FRAME CW_DA CW_CL

MSP sound IF channel 1
AD_CV[7:1] ANA_IN1+ AGC ANA_IN2+ AD_CV[8] AD

Amplitude Carrier Detect Mixer AD_CV[9] IDENT

Carrier Detect

ANA_IN-

MSP sound IF channel 2
Mixer Lowpass

Amplitude

Mute CORDIC Phase FRAME NICAMA DCO2 Pins Internal signal lines (see fig. 4­5) Control registers DCO2 Oscillator FIR_REG_2 MODE_REG[8] Differentiator

Lowpass

FM1/AM

Fig. 4­2: Demodulator architecture 4.1.3. Lowpass Filtering Block for Mixed Sound IF Signals By means of decimation filters the sampling rate is reduced. Then, data shaping and/or FM bandwidth limitation is performed by a linear phase Finite Impulse Response (FIR-filter). Just like the oscillators' increments the filter coefficients are programmable and are written into the IC by the CCU via the control bus. Thus, for example, different NICAM versions can easily be implemented. Two not necessarily different sets of coefficients are required, one for channel 1 (NICAM or FM2) and one for channel 2 (FM1=FM-mono). In section 11.2.3. several coefficient sets are proposed. Since both MSP channels are designed to process the German FM Stereo System with the same FIR coefficient set (despite 7 dB power level difference of the two sound carriers), the MSP channel 1 has an internal overall gain of 6 dB. To process two carriers of identical power level these 6 dBs have to be taken into account by decreasing the values of the channel 1 coefficient set, which has already been done in table 11­7. 4.1.4. CORDIC Block The filtered sound IF signals are demodulated by transforming the incoming signals from Cartesian into polar format by means of a CORDIC processor block. On the output, the phase and amplitude is available for further 10 processing. AM signals are derived from the amplitude information whereas the phase information serves for FM and NICAM (DQPSK) demodulation. 4.1.5. Differentiators FM demodulation is completed by differentiating the phase information output of the CORDIC block. 4.1.6. Lowpass Signals Filter Block for Demodulated

The demodulated FM and AM signals are further lowpass filtered and decimated to a final sampling frequency of 32 kHz. The usable bandwidth of the final baseband signals is about 15 kHz. 4.1.7. High Deviation FM Mode By means of MODE_REG [9], the maximum FM-deviation can be extended to approximately ±360 kHz. Since this mode can be applied only for the MSP sound IF channel 2, the corresponding matrices in the baseband processing must be set to sound A. Apart from this, the coefficient sets 380 kHz FIR_REG2 or 500 kHz FIR_REG2 must be chosen for the FIR_REG_2. In relation to the normal FM-mode, the audio level of the highdeviation mode is reduced by 6 dB. ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
S2 and S3 maintain their position and function. This facilitates the copying from selected SCART-inputs to SCART-outputs in the TV-set's standby mode.
SCART_IN SC1_IN_L/R MONO_IN SC1_IN_L/R 11 SC3_IN_L/R S1 ACB[3:2] 00 01 10 from Audio Baseband Processing (DFP) SCARTL/R SC1_OUT_L/R D A 11 S2 ACB[5:4] 00 01 10 S3 SC2_OUT_L/R SCART_OUT ACB[1:0] 00 01 10 to Audio Baseband Processing (DFP_IN) A D SCARTL/R

4.1.8. MSP-Mute Function in the Dual Carrier FM Mode To prevent noise effects or FM identification problems in the absence of one of the two FM carriers the MSP 3410 B offers a carrier detection feature, which must be activated by means of AD_CV[9], see section 11.2.1. If no FM carrier is available at the MSP channel 1, the corresponding channel FM2 (and S-Bus output samples 3 and 4) are muted. If no FM carrier is available at the MSP channel 2, the corresponding channel FM1 (and S-Bus output samples 1 and 2) are muted. In case of the absence of both FM carriers pure noise will be amplified by the input AGC. Therefore a proper mute function depends on the noise quality of the TV set's IF part and cannot be guaranteed. The mute function is not recommended for the satellite mode. 4.1.9. DQPSK-Decoder In case of NICAM-mode the phase samples are decoded according the DQPSK-Coding scheme. The output of this block contains the original NICAM-bitstream, which is available at the N-Bus interface. 4.1.10. NICAM-Decoder Before any NICAM decoding can start, the MSP must lock to the NICAM frame structure by searching and synchronizing to the so-called Frame Alignment Words (FAW). To reconstruct the original digital sound samples the NICAM-bitstream has to be descrambled, deinterleaved and rescaled. Also bit error detection and correction (concealment) is performed in this NICAM specific block. To facilitate the Central Control Unit CCU to switch the TV-set to the actual sound mode, control information on the NICAM mode and bit error rate are supplied by the the NICAM-Decoder. It can be read out via the I2C-Bus. 4.2. Analog Section: SCART Switches and Standby Mode The analog input and output sections offer a wide range of switching facilities, which are shown in Fig. 4­3.To design a TV set with 3 pairs of SCART-inputs and two pairs of SCART-outputs, no external switching hardware is required. The switches are controlled by the ACB bits defined in the audio processing interface (see section 12. Programming the Audio Processing Part). If the MSP 3410 B is switched off by first pulling STANDBYQ low and then disconnecting the 5V but keeping the 8V power supply (`Standby'-mode), the switches S1, ITT Semiconductors

Fig. 4­3: SCART-Switching Facilities Bold lines determine the default configuration In case of power-on start or starting from standby, the IC switches automatically to the default configuration, shown in the figure above. This action takes place after the first I2C transmission into the DFP part. By transmitting the ACB register first, the individual default setting mode of the TV set can be defined. 4.3. MSP 3410 B Audio Baseband Processing By means of the DFP processor all audio baseband functions are performed by digital signal processing (DSP). The DSP functions are grouped into three processing parts: Input preprocessing, channel selection and channel postprocessing. The input preprocessing is intended to prepare the various signals of all input sources in order to form a standardized signal at the input to the channel selector. The signals can be adjusted in volume, are processed with the appropriate deemphasis and are dematrixed if necessary. Having prepared the signals that way, the channel selector makes it possible to distribute all possible source signals to the desired output channels. The ability to route in an external coprocessor for special effects like graphic equalizer, surround processing, and sound field processing is of special importance. Routing can be done with each input source and output channel via the I2S inputs and outputs. All input and output signals can be processed simultaneously with the exception that FM2 cannot be pro11

MSP 3410 B
cessed at the same time as NICAM. Note that the NICAM input signals are only available in the MSP 3410 B version. While processing the adaptive deemphasis, no dual carrier stereo (German or Korean) or NICAM processing is possible. Identification values are not valid either.

PRELIMINARY DATA SHEET

operation mode, the MSP 3410 B detects the so-called identification signal. Information is supplied via the Stereo Detection Register to an external CCU.
Stereo Detection Filter IDENT AM Demodulation Bilingual Detection Filter Level Detect

Level Detect Stereo Detection Register

4.4. Dual Carrier FM Stereo/Bilingual Detection In the German and Korean TV standard, audio information can be transmitted in three modes: Mono, stereo or bilingual. To obtain information about the current audio

­

Fig. 4­4: Stereo/bilingual detection

Analog Inputs

SCARTL SCARTR DC level readout FM1 FM1 FM2 Adaptive Deemphasis Deemphasis 50/75 µs J17

Prescale

Loudspeaker Channel Matrix

Bass Treble Loudness Spatial Effects

Volume Balance

Loudspeaker L Loudspeaker Outputs Loudspeaker R

Prescale FM-Matrix Beeper

Demodulated IF Inputs

DC level readout FM2 NICAMA NICAMB Prescale Channel Select Deemphasis J17 Headphone Channel Matrix Volume Headphone L Headphone Outputs Headphone R

SBUS1 SBUS2 SBUS Inputs SBUS3 SBUS4

SCART Channel Matrix

Volume

SCARTL SCART Outputs SCARTR

I 2S Channel Matrix

I 2SL I 2SR

I 2S Outputs

I 2SL I 2S Bus Inputs I 2SR

Quasi peak readout L Quasi-Peak Detector Quasi peak readout R

Note: Actually, the source of the Quasi-Peak Detector is always the signal of the loudspeaker channels.

Fig. 4­5: Audio baseband processing (DFP-Firmware)

NICAMA

Internal signal lines (see fig. 4­2)

Table 4­1: Some examples for recommended channel assignments for demodulator and audio processing part Mode
B/G-Stereo B/G-Bilingual

MSP Sound IFChannel 1
FM2 (5.74 MHz): 2R FM2 (5.74 MHz): Sound B

MSP Sound IFChannel 2
FM1 (5.5 MHz): L+R FM1 (5.5 MHz): Sound A

FMMatrix
B/G Stereo No Matrix

ChannelSelect
Speakers: FM Speakers: FM H. Phone: FM Speakers: NICAM H. Phone: FM Speakers: FM Speakers: FM Speakers: FM H. Phone: FM Speakers: FM H. Phone: FM

Channel Matrix
Stereo Speakers: Sound A H. Phone: Sound B Speakers: Stereo H. Phone: Sound A Sound A Stereo Speakers: Sound A H. Phone: Sound B=C Speakers: Sound A H. Phone: Sound A

NICAM-I-ST/ FM-mono Sat-Mono Sat-Stereo Sat-Bilingual

NICAM (6.552 MHz)

FM (6.0 MHz): mono

No Matrix

not used 7.2 MHz: R 7.38 MHz: Sound C

FM (6.5 MHz): mono 7.02 MHz: L 7.02 MHz: Sound A

No Matrix No Matrix No Matrix

Sat-High Dev. Mode

don't care

6.552 MHz

No Matrix

12

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
read address (81 hex or 85 hex) and reading two bytes of data. Refer to Fig. 5­1: I2C Bus Protocol and section 5.2. Proposal for MSP 3410 B I2C Telegrams. Due to the internal architecture of the MSP 3410 B, the IC cannot react immediately to an I2C request. The typical response time is about 0.3 ms for the DFP processor part and 1 ms for the FP processor part if NICAM processing is active. If the receiver (MSP) can't receive another complete byte of data until it has performed some other functions, for example servicing an internal interrupt, it can hold the clock line I2C_CL LOW to force the transmitter into a wait state. The positions within a transmission where this may happen are indicated by 'Wait' in section 5.1. The maximum Wait-period of the MSP during normal operation mode is less than 7 ms. I2C-Bus error conditions (valid only from TC17 on): In case of any internal error, the MSPs wait-period is extended to 7.07 ms. Afterwards the MSP does not acknowledge (NAK) the device address. The data line will be left HIGH by the MSP and the clock line will be released. The master can then generate a STOP condition to abort the transfer. By means of NAK, the master is able to recognize the error state and to reset the IC via I2C-Bus. While transmitting the reset protocoll (s. 5.2.4.) to `CONTROL', the master must ignore the not acknowledge bits (NAK) of the MSP.

5. Control Bus Interface As a slave receiver, the MSP 3410 B can be controlled via I2C bus. Access to internal memory locations is achieved by subaddressing. The FP processor and the DFP processor parts have two separate subaddressing register banks. In order to allow for more MSP 3410 B IC's to be connected to the control bus, an ADR_SEL pin has been implemented. With ADR_SEL pulled to high, the MSP 3410 B responds to changed device addresses, thus two identical devices can be selected. Other devices of the same family will have different subaddresses (e.g. 34X0). By means of the RESET bit in the CONTROL register all devices with the same device address are reset. The IC is selected by asserting a special device address in the address part of a I2C transmission. A device address pair is defined as a write address (80 hex or 84 hex) and a read address (81 hex or 85 hex). Writing is done by sending the device write address first, followed by the subaddress byte, two address bytes, and two data bytes. For reading, the read address has to be transmitted first by sending the device write address (80 hex or 84 hex) followed by the subaddress byte and two address bytes. Without sending a stop condition, reading of the addressed data is done by sending the device

Table 5­1: I2C Bus Device and Subaddresses Name Binary Value Hex Value
ADR_SEL=low

Hex Value
ADR_SEL=high

Mode

Function

MSP CONTROL TEST WR_FP RD_FP WR_DFP RD_DFP

1000 000x 0000 0000 0000 0001 0001 0000 0001 0001 0001 0010 0001 0011

80/81 00 01 10 11 12 13

84/85

R/W W W W W W W

MSP device address software reset only for internal use write address FP read address FP write address DFP read address DFP

Table 5­2: Control Register Name CONTROL MSB RESET 14 0 13..1 0 LSB 0

ITT Semiconductors

13

MSP 3410 B
5.1. Protocol Description Write to DFP or FP
S hex 80 Wait ACK sub-addr ACK addr-byte high ACK addr-byte low ACK

PRELIMINARY DATA SHEET

data-byte high

ACK

data-byte low

ACK

P

Read from DFP or FP
S hex 80 Wait ACK sub-addr ACK addr-byte high ACK addr-byte low ACK S hex 81 Wait ACK data-byte high

Write to Control or Test Registers
S hex 80 Wait ACK sub-addr ACK data-byte high ACK

data-byte low

Note: S = P= ACK = NAK = Wait =

I2C-Bus Start Condition from master I2C-Bus Stop Condition from master Acknowledge-Bit: LOW on I2C_DA from slave (= MSP, grey) or master (= CCU, hatched) Not Acknowledge-Bit: HIGH on I2C_DA from master (= CCU, hatched) to indicate `End of Read' or from MSP indicating internal error state (not illustrated, only for version F7 on.) I2C-Clock line held low by the slave (= MSP) while interrupt is serviced (< 7 ms)

I2C_DA S I2C_CL Fig. 5­1: I2C bus protocol

1 0

P

(MSB first; data must be stable while clock is high)

5.2. Proposal for MSP 3410 B I2C Telegrams 5.2.1. Symbols < > aa dd Start Condition Stop Condition Address Byte Data Byte

5.2.4. Examples
<80 00 80 00> <80 00 00 00> <80 12 00 08 01 20>

RESET all MSP's statically clear RESET set loudspeaker channel source to NICAM and Matrix to STEREO

5.3. Start Up Sequence 5.2.2. Write Telegrams
<80 00 dd dd> <80 10 aa aa dd dd> <80 12 aa aa dd dd>

software RESET write data into FP register write data into DFP register

5.2.3. Read Telegrams
<80 11 aa aa <81 dd dd> read data from FP register <80 13 aa aa <81 dd dd> read data from DFP register

After power on or RESET the IC is in an inactive state. The CCU has to transmit the required coefficient set for a given operation via the I2C bus. Initialization must start with the demodulator part. If required for any reason, from version F7 on, the audio processing part can be loaded before the demodulator part.

14

ÇÇÇ ÇÇÇ ÇÇÇ ÇÇÇ ÇÇÇ ÇÇÇ
ACK data-byte low NAK P ACK P

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
8. Audio PLL and Crystal Specifications The MSP 3410 B requires a 18.432 MHz (10 pF, parallel) crystal. The clock supply of the whole system depends on the MSP 3410 B operation mode: 1. FM-Stereo: The system clock runs free on the crystal's 18.432 MHz. 2. D2-MAC operation: In this case, the system clock is locked to a synchronizing signal (DMA_SYNC) supplied by the D2-MAC chip. The DMA and the AMU chips can be driven by the MSP 3410 B audio clock (AUD_CL_OUT). 3. NICAM and FM_mono: An integrated clock PLL uses the 364 kHz baud-rate, accomplished in the NICAM demodulator block, to lock the system clock to the bit rate respective 32 kHz sampling rate of the NICAM transmitter. As a result, the whole audio system is supplied with a controlled 18.432 MHz clock. Remark on using the crystal: External capacitors at each crystal pin to ground are required. They are necessary for tuning the open-loop frequency of the internal PLL and for stabilizing the frequency in closed-loop operation. The higher the capacitors, the lower the clock frequency results. The nominal free running frequency should match the center of the tolerance range between 18.433 and 18.431 MHz as closely as possible.

6. N-Bus Interface The N-Bus interface consists of two lines, N-data and N-clock. The pure NICAM_728 data stream (before descrambling) is available together with a 728 kHz clock signal for the purpose of data transmission. N-Bus signals are based on TTL-levels. Data are latched with the falling clock edge. 7. Pay-TV Interface The MSP 3410 B facilitates the reception of encrypted NICAM sound, which is provided by Pay-TV systems. By means of bit 1 of the control word `MODE_REG' the operation mode `PAY-TV' can be activated. The MSP 3410 B inherent descrambler generally uses a 9-bit start sequence, which initializes a pseudo random sequence generator each ms. In normal operation mode the 9-bit sequence exists of 9 bits having each high level, which are loaded automatically into the descrambler's shift register. In the Pay-TV mode these bits have to be loaded via the two pins CW_DA and CW_CL into the mentioned shift register. The time window to load one complete 9-bit sequence is given by the high time of the frame signal which is available on pin 5. It is not necessary to load a new sequence at each ms, because if no new sequence has been transmitted, the old one is saved. If less than 9 new bits at each ms are loaded, one has to consider that any new incoming bit shifts the old ones by one position inside the shift register. A complete timing diagram is illustrated in Fig. 7­1.

Frame

8 Bits

720 Bits Start of Descrambler End T

CW-Clock CW-Data CW-Clock Min: 10 kHz Max: 4 MHz 1 2 3 4 5 6 7 8 9

Period to load CW-Word T 7E-6 s

Fig. 7­1: Timing for Pay-TV signals

ITT Semiconductors

15

MSP 3410 B
9. S-Bus Interface Digital audio information provided by the DMA2381 via the AMU is serially transmitted to the MSP 3410 B via the S-Bus. The MSP 3410 B always has the master function. The S-Bus interface consists of four pins: 1. S_DA_IN: Four channels (4*16 bits) per sampling cycle (32 kHz) are transmitted. 2. S_CL: Gives the timing for the transmission of S-DATA (4.608 MHz). 3. S_ID: After 64 S-CLOCK cycles the S_ID determines the end of one sampling period. 4. S_DA_OUT: FM-Demodulator or NICAM decoder output for test purpose. 10. I2S Bus Interface

PRELIMINARY DATA SHEET

By means of this standardized interface, additional feature processors can be connected to the MSP 3410 B. Two possible formats are supported: The standard mode (MODE_REG[4]=0) selects the SONY format, where the I2S_WS signal changes at the word boundaries. The so-called PHILIPS format, which is characterized by a change of the I2S_WS signal one I2S_CL period before the word boundaries, is selected by setting MODE_REG[4]=1. The MSP 3410 B normally serves as the master on the I2S interface. Here the clock and word strobe lines are driven by the MSP 3410 B. By setting MODE_REG[3]=1, the MSP 3410 B is switched to a slave mode. Now these lines are input to the MSP 3410 B and the master clock is synchronized to 576 times the I2S_WS rate (32 kHz). No NICAM or D2MAC operation is possible in this mode. The I2S bus interface consists of four pins: 1. I2S_DA_IN: For input, two channels (2*16 bits) per sampling cycle (32 kHz) are transmitted. 2. I2S_DA_OUT: For output, two channels (2*16 bits) per sampling cycle (32 kHz) are transmitted. 3. I2S_CL: Gives the timing for the transmission of I2S serial data (1.024 MHz). 4. I2S_WS: The I2S_WS word strobe line defines the left and right sample.

16

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
Accessing a process address starts specific actions in the FP processor. For example addressing register 60hex activates the internal transfer of all preloaded data (MODE_REG, DCO1_LO/HI) into their final hardware registers. It's only the access of the address 60hex that counts, the two data bytes in the transmission have no meaning. Table 4­1 explains how to assign FM carriers to the MSP-Sound IF channels and the corresponding matrix modes in the audio processing part.

11. Programming the Demodulator Part 11.1. Write Registers: Table and Addresses In Table 11­1 all Write Registers are listed. All transmissions on the control bus are 16 bits wide. Data for the demodulator part (FP) have 8 or 12 significant bits. These data have to be inserted LSB bound and filled with zero bits into the 16 bit transmission word.

Table 11­1: MSP 3410 B write registers Register Write Address (hex) 00BB 0083 0001 0005 0093 009B 00A3 00AB 0107 010F Function

AD_CV MODE_REG FIR_REG_1 FIR_REG_2 DCO1_LO DCO1_HI DCO2_LO DCO2_HI FAWCT_SOLL FAW_ER_TOL

input selection, configuration of AGC and Mute Function and selection of A/Dconverter mode register serial shift register for 6 8 bit, filter coefficient channel 1 (48 bit) serial shift register for 6 8 bit, + 2 12 bit off set (total 72 bit) increment channel 1 Low Part increment channel 1 High Part increment channel 2 Low Part increment channel 2 High Part To synchronize to the frame structure of the NICAM bit stream, the MSP checks the data for Frame Alignment Words (FAWs). After having captured the first one, the MSP continues to check for n frame periods. On having found at least n-m FAWs after this period, the frame synchronism is achieved and the MSP switches to active NICAM-decoding. The value for n has to be loaded into FAWCT_SOLL; the one for m into FAW_ER_TOL. Proposal : n=12; m=2 audio PLL in case of NICAM operation mode Function After switch on or changing the TV system (B/G to I, I to B/G) all write-parameters have to be transmitted via I2C-Bus into the MSP 3410 B. Then `Load_REG_1/2' writes them into the corresponding registers. FM-processing starts. These are MODE_REG, DCO1/2_LO/HI. In the case of a TV-Standard change in MSP channel 1, only new channel 1 parameters have to be transmitted into the IC via I2C-Bus. These are: MODE_REG, DCO1_LO/HI. LOAD_REG_1 sets up the MSP channel 1 without interrupting the MSP channel 2 (FM1 or MONO channel). To start the NICAM-processing, this address has to be transmitted into the FP. Check of the FP ALU (for testing only) 0 always open 1 to be closed = default

AUDIO_PLL Process LOAD_REG_1/2

02D7 Address (hex) 0056

LOAD_REG_1

0060

SEARCH_NICAM SELF_TEST

0078 0792

Note: The WRITE-Addresses cannot be used to read back the corresponding register values.

ITT Semiconductors

17

MSP 3410 B
11.2. Write Registers: Functions and Values In the following, the functions of some registers are explained and their (default) values are defined: 11.2.1. Setting of Parameter AD_CV Table 11­2: AD_CV Register AD_CV 00BBhex Bit AD_CV [0] AD_CV [6:1] Meaning test Reference level in case of Automatic Gain Control = on (see Table 11­3). Constant gain factor when Automatic Gain Control = off (see Table 11­4). Determination of Automatic Gain or Constant Gain Selection of analog input MSP-Carrier-Mute Function

PRELIMINARY DATA SHEET

Settings 0 = on (default) 1 = off (for testing)

AD_CV [7] AD_CV [8] AD_CV [9]

0 = constant gain 1 = automatic gain 0 = ANALOG IN1 1 = ANALOG IN2 0 = off (no mute) 1 = on (mute as described in section 4.1.) 0 = on (default) 1 = off (for testing) 0

AD_CV[10] AD_CV[15:11]

NICAM-FIFO-Watchdog (only for test mode) reserved

Table 11­3: Reference values for active AGC (AD_CV[7] = 1) Application Terrestrial TV Input Signal Contains 2 FM Carriers or 1 FM and 1 NICAM Carrier 1 or more FM Carriers 1 NICAM Carrier only AD_CV [6:1] Ref. Value 101000 AD_CV [6:1] in integer 40 Range of Input Signal at pin 41 or 43 0.14 ­ 3 Vpp1)

SAT NICAM only
1)

100011 010100

35 20

0.14 ­ 3 Vpp1) 0.07 ­ 1.0 Vpp

For signals above 1.4 Vpp, the minimum gain of 3 dB is switched and overflow of the AD converter may result. Due to the robustness of the internal processing, the IC works up to and even more than 3 Vpp, if norm conditions of FM/ NICAM or FM1/FM2 ratio are supposed. In this overflow case, a loss of FM-S/N-ratio of about 10 dB may appear.

18

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B

Table 11­4: AD_CV parameters for constant input gain (AD_CV[7]=0) Step 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1)

AD_CV [6:1] Constant Gain 000000 000001 000010 000011 000100 000101 000110 000111 001000 001001 001010 001011 001100 001101 001110 001111 010000 010001 010010 010011 010100

Gain 3.00 dB 3.85 dB 4.70 dB 5.55 dB 6.40 dB 7.25 dB 8.10 dB 8.95 dB 9.80 dB 10.65 dB 11.50 dB 12.35 dB 13.20 dB 14.05 dB 14.90 dB 15.75 dB 16.60 dB 17.45 dB 18.30 dB 19.15 dB 20.00 dB

Input Level at pin ANA_IN1+ and ANA_IN2+ maximum input level: 3 Vpp (FM) or 1 Vpp (NICAM)1)

maximum input level: 0.14 Vpp

For signals above 1.4 Vpp, the minimum gain of 3 dB is switched and overflow of the AD converter may result. Due to the robustness of the internal processing, the IC works up to and even more than 3 Vpp, if norm conditions of FM/ NICAM or FM1/FM2 ratio are supposed. In this overflow case, a loss of FM-S/N-ratio of about 10 dB may appear. 11.2.2. Control Register `MODE_REG' The register `MODE_REG' contains the control bits determining the operation mode of the MSP 3410 B; Table 11­5 explains all bit positions. Table 11­5: Control word `MODE_REG': all bits are "0" after power-on-reset MODE_REG 0083hex Bit [0] [1] [2] [3] [4] [5] [6] Function DMA_SYNC1) PAYTV_EN DESCR_DIS I2S Mode1) I2S_WS Mode Audio_CL_OUT NICAM1) Comment Synchronization to DMA Pay-TV NICAM-Descrambler Master/Slave mode of the I2S bus WS due to the Sony or Philips-Format Switch Audio_Clock_Output to tristate MSP-channel 1 mode Definition 0 = NICAM (intern. Sync) 1 = D2MAC (ext. Sync) 0 = off 1 = on 0 = on 1 = off 0 = Master 1 = Slave 0 = Sony 1 = Philips 0 = on 1 = tristate 0 = FM 1 = Nicam Recommendation X 0 0 X X X X

ITT Semiconductors

19

MSP 3410 B

PRELIMINARY DATA SHEET

MODE_REG 0083hex Bit [7] [8] [9] Function FM1 FM2 FM AM HDEV Comment MSP-channel 1 mode MSP-channel 1/2 mode High Deviation Mode (channel matrix must be sound A) configuration of internal sound bus mode of sound bus3) reserved Definition 0 = Nicam 1 = FM 0 = FM 1 = AM 0 = normal 1 = high deviation mode 0 = Nicam/FM-Mono 1 = Two Carrier FM 0 = Tristate 1 = Active must be 0 Recommendation X 0 0

[10] [11] [15:12]
1)

S-Bus Setting S-Bus Mode2) reserved

X 0 0 X: Depending on mode

In case of NICAM operation, I2S-slave mode or synchronization to DMA not possible. In case of synchonization to DMA, no I2S-slave mode or NICAM is allowed. In case of I2S-slave mode, no synchonization to DMA or NICAM is allowed. 2) The normal operation mode is `Active' 3) To reduce radiation, the pins S_DA_OUT, S_CL, and S_ID should be switched to tristate if not used. IF S-Bus Mode = `tristate', pins `Frame', N_CL, and N_DA are also switched to tristate.

11.2.3. FIR-Parameter The following data values (see Table 11­6) are to be transferred 8 bits at a time embedded LSB-bound in a 16 bit word. Note: These sequences must be obeyed. To change a coefficient set, the complete block FIR_REG_1 or FIR_REG_2 must be transmitted. The new coefficient set will be active without a load_reg routine. Table 11­6: Loading sequence for FIR-coefficients
FIR_REG_1 0001hex No. 1 2 3 4 5 6 Symbol Name NICAM/FM2_Coeff. (5) NICAM/FM2_Coeff. (4) NICAM/FM2_Coeff. (3) NICAM/FM2_Coeff. (2) NICAM/FM2_Coeff. (1) NICAM/FM2_Coeff. (0) (Channel 1: NICAM/FM2) Bits 8 8 8 8 8 8 Value see Table 11­7.

FIR_REG_2 0005hex No. 1 2 3 4 5 6 7 8 9 Symbol Name

(Channel 2: FM1/FM mono) Bits 8 8 8 8 8 8 8 8 8 Value 04 HEX 40 HEX 00 HEX see Table 11­7.

* IMREG1 (8 LSBS) * IMREG1 / IMREG2 (4 MSBs / 4 LSBs) * IMREG2 (8 MSBs) FM_Coef (5) FM_Coef (4) FM_Coef (3) FM_Coef (2) FM_Coef (1) FM_Coef (0)

* IMREG_1/2: Two 12-bit off-set constants

IMREG1 and IMREG2 are used to compensate for DCoffset, which are inherent to the FIR filter structure. IMREG1 is valid for the FIR_REG_1, IMREG2 for FIR_REG_2. In the Table above, IMREG1= IMREG2 = 004. Due to the partitioning to 8 bit units, the values 04hex, 40hex, and 00hex arise. ITT Semiconductors

20

PRELIMINARY DATA SHEET

MSP 3410 B

Table 11­7: 8 bit FIR-coefficients (decimal integer) for MSP 3410 B
FIR_REG_1 0001hex and FIR_REG_2 0005hex NICAM FMTerrestrial B/G, I FM - Satellite FIR filtering corresponds to a bandpass filtering with a band width of B = 130 kHz, 180 kHz, 200 kHz, ... 380 kHz B fc frequency Autosearch or AM

Bandwidth (see also Table FM Volume Prescale) C (i) SC/ SP/ F FIR_ REG1 UK FIR_ REG1 German Dual FM FIR_ REG1 and 2 3 18 27 48 66 72 130 kHz FIR_ REG1
1)

130 kHz FIR_ REG2
1)

180 kHz FIR_ REG1

180 kHz FIR_ REG2

200 kHz FIR_ REG1

200 kHz FIR_ REG2

280 kHz FIR_ REG1

280 kHz FIR_ REG2

380 kHz FIR_ REG1

380 kHz FIR_ REG2

500 kHz FIR_ REG2

FIR_ REG2

0 1 2 3 4 5 1)

­2 ­8 ­10 10 50 86

2 4 ­6 ­4 40 94

37 27 32 60 51 65

73 53 64 119 101 127

4 9 14 23 27 32

9 18 28 47 55 64

1 9 14 24 33 37

3 18 27 48 66 72

­4 ­4 2 19 41 57

­8 ­8 4 36 78 107

­1 ­6 ­9 4 38 70

­1 ­9 ­16 5 65 123

­1 ­1 ­8 2 59 126

75 19 36 35 39 40

The 130 kHz coefficients are based on subcarriers, which are 7 dB below an existent main carrier. INCRdez = int(f/fs 224) with: int = integer function f = IF-frequency in MHz fS = sampling frequency (18.432 MHz) Conversion of INCR into hex-format and separation of the 12-bit low and high parts lead to the required increments. (DCO1_HI or _LO for channel 1, DCO2_HI or LO for channel 2).

11.2.4. DCO-Increments For a chosen TV standard a corresponding set of 24-bit increments determining the mixing frequencies of the quadrature mixers, has to be written into the IC. In Table 11­8 some examples of DCO increments are listed. It is necessary to divide them up into low part and high part. The formula for the calculation of the increments for any chosen IF-Frequency is as follows:

Table 11­8: DCO increments for the MSP 3410 B; frequency in MHz, increments in Hex DCO1_LO 0093hex, DCO1_HI 009Bhex; DCO2_LO 00A3hex, DCO2_HI 00ABhex Frq. MHz 4.5 5.04 5.5 5.58 5.7421875 6.0 6.2 6.5 6.552 7.02 7.38 DCO_HI 3E8 460 4C6 4D8 4FC 535 561 5A4 5B0 618 668 DCO_LO 000 000 38E 000 0AA 555 C71 71C 000 000 000 5.76 5.85 5.94 6.6 6.65 6.8 7.2 7.56 500 514 528 5BA 5C5 5E7 640 690 000 000 000 AAA C71 1C7 000 000 Frq. MHz DCO_HI DCO_LO

ITT Semiconductors

21

MSP 3410 B
11.3. Read Registers: Listing and Addresses The following 8-bit parameters can be read out of the RAM of the MSP 3410 B; functionally they all belong to the NICAM decoding process; their addresses are listed in Table 11­9. All transmissions take place in 16 bit words. The valid 8 bit data are the 8 LSBs of the received data word. To enable correct switching to NICAM sound, at least the register C_AD_BITS must be read and evaluated by the CCU. Additional data bits and CIB bits, if supplied by the NICAM transmitter, as well as information about the signal quality can be obtained by reading the remaining registers. Format:
MSB 7 A[10] 6 A[9] 5 A[8]

PRELIMINARY DATA SHEET

ADD_BITS 0038hex 4 A[7] 3 A[6] 2 A[5] 1 A[4]

LSB 0 A[3]

CIB_BITS: cib bits 1 and 2 (see NICAM 728 specifications) Format:
MSB 7 x 6 x 5 x CIB_BITS 003Ehex 4 x 3 x 2 x 1 CIB1 LSB 0 CIB2

Table 11­9: Addresses of read registers Read Registers C_AD_BITS FAWCT_IST ADD_BITS CIB_BITS CONC_CT HEX 0023 0025 0038 003E 0058

FAWCT_IST: The contents of this register give information on the actual position of the FAW-counter. For optimum NICAM performance, the value should be identical with or little below the value of 'FAW_SOLL'. If it reaches 0 the FP-software mutes and stops the NICAM-decoding automatically by searching for FAW synchronization once more. CONC_CT: This register contains the actual number of bit errors of the previous 728 bit data frame. It may happen that in spite of acceptable FAWCT_IST the bit error rate result is too high for appropriate sound performance. In this case the CCU can switch to the analog FMsound assumed to have the same program (Control bit C4). Table 11­10: NICAM operation modes as defined by the EBU NICAM 728 specification
C4 C3 0 0 0 0 0 0 0 C2 0 0 1 1 0 0 1 C1 0 1 0 1 0 1 0 Operation Mode Stereo sound (NICAMA/B), independent mono sound (FM1) Two independent mono signals (NICAMA, FM1) Three independent mono channels (NICAMA, NICAMB, FM1) Data transmission only; no audio Stereo sound (NICAMA/B), FM1 carries same channel One mono signal (NICAMA). FM1 carries same channel as NICAMA Two independent mono channels (NICAMA, NICAMB). FM1 carries same channel as NICAMA Data transmission only; no audio Unimplemented sound coding option (not yet defined by EBU NICAM 728 specification)

11.4. Read Registers: Functions and Values C_AD_BITS: NICAM operation mode control bits and A[0­2] of the additional data bits. Format:
MSB 7 A[2] 6 A[1] 5 A[0] C_AD_BITS 0023hex 4 C4 3 C3 2 C2 1 C1 LSB 0 0 S 1 0 0 0

Important: "S" = Bit[0] indicates correct NICAM-synchronization (S=1). If S = 0, no correct frame or sequence synchronization have been found yet and the read registers are not valid. The operation mode is coded by C4-C1 as shown in Table 11­10. ADD_BITS: Contains the remaining 8 of the 11 additional data bits. The additional data bits are yet not defined by the NICAM 728 system. 22

1 1

1 x

0 1

1 x

1 x

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
For NICAM operation the following steps listed in `NICAM_START, _READ and _Check' in Table 11­11 must be taken. For FM-stereo operation the evaluation of the identification signal must be performed. For positive identification check, the MSP 3410 B sound channels have to be switched corresponding to the detected operation mode.

11.5. Sequences to Transmit Parameters and to Start Processing

After having been switched on, the MSP has to be initialized by transmitting the parameters according to the LOAD_SEQ_1/2 of Table 11­11. To make the data active, the load routine LOAD_REG_1/2 must be activated.

Table 11­11: Sequences to initialize and start the MSP 3410 B
LOAD_SEQ_1/2: General Initialization, followed by LOAD_REG_1/2 Write into MSP 3410 B: 0. AD_CV 1. Audio_PLL 2. FAWCT_SOLL 3. FAW_ER_TOL 4. FIR_REG_1 5. FIR_REG_2 6. MODE_REG 7. DCO1_LO 8. DCO1_HI 9. DCO2_LO 10. DCO2_HI 11. start LOAD_REG_1/2 process; FM-processing starts NICAM_START: Start of the NICAM Software Write into MSP 3410 B: 1. Start SEARCH_NICAM Process 2. Wait at least 0.5 s NICAM_READ: Read NICAM specific information Read out of MSP 3410 B: 1. FAWCT_IST 2. C_AD_BITS 3. CONC_CT NICAM_CHECK: CCU checks for presence, operation mode and quality of NICAM signal 1. Evaluation of all three parameters in the CCU (see section 11.4.) 2. If necessary, switch the corresponding sound channels within the audio processing part FM_IDENT_CHECK: Decoding of the identification signal 1. Evaluation of the stereo detection register (DFP register 0018hex, high part) 2. If necessary, switch the corresponding sound channels within the audio processing part LOAD_SEQ_1: Reinitialization of Channel 1 without affecting Channel 2, followed by LOAD_REG_1 Write into MSP 3410 B: 1. FIR_REG_1 2. MODE_REG 3. DCO1_LO 4. DCO1_HI 5. start LOAD_REG_1 process (6 8 bit) (12 bit) (12 bit) In the case "NICAM only" operation, the steps 9. and 10. can be skipped Note: To ensure software compatibility to the MSP3400 B, before any modification of a demodulator parameter concerning an active output channel, this channel should be muted

(only for NICAM mode) (only for NICAM mode)

PAUSE: Duration of "Pause" determines the repetition rate of the NICAM or the FM_IDENT-check AUDIO PROCESSING INIT: Initialization of Audio Processing Part, which may be customer dependant (see section 12.)

ITT Semiconductors

23

MSP 3410 B
11.6. Software Proposals for Multistandard TV-Sets To familiarize the reader with the programming scheme of the MSP 3410 B demodulator part, three examples in the shape of flow diagrams are shown in the following sections.

PRELIMINARY DATA SHEET

the MSP 3410 B must be switched to the FM-mono sound.

11.6.2. Multistandard Including System I with NICAM/FM-Mono only This case is identical to the one above. The only difference consists in selecting the UK parameters for DCO1_LO/HI, DCO2_LO/HI and FIR_REG_1.

11.6.1. Multistandard Including System B/G with NICAM/FM-Mono only Fig. 11­1 shows a flow diagram for the CCU software, applied for the MSP 3410 B in a TV set, which facilitates NICAM and FM-mono sound. For the instructions, please refer to Table 11­11.

11.6.3. Multistandard Including System B/G with NICAM/FM-Mono and German DUAL FM Fig. 11­3 shows a flow diagram for the CCU software, applied for the MSP 3410 B in a TV set, which facilitates all standards according to System B/G. For the instructions used in the diagram, please refer to Table 11­11. After having switched on the TV-set and having initialized the MSP 3410 B (LOAD_SEQ_1/2), FM-mono sound is available. Fig. 11­3 shows that to check for any stereo or bilingual audio information in channel 1, its parameter should be loaded with NICAM and FM2 parameters alternately (LOAD_SEQ_1). In the case of success the MSP 3410 B has to switch to the desired audio mode.

START LOAD_SEQ_1/2 Channel 1: NICAM Parameter

Audio Processing Init

NICAM_START

11.6.4. Satellite Mode Fig. 11­2 shows the simple flow diagram to be used for the MSP 3410 B in a satellite receiver. For FM-mono operation the corresponding FM carrier should preferably be processed at the MSP-channel 2.

Pause

NICAM_READ

Yes NICAM_CHECK
NICAM ?

START LOAD_SEQ_1/2 MSP­Channel 1: FM2­Parameter MSP­Channel 2: FM1­Parameter

No LOAD_REG_1

Fig. 11­1: CCU software flow diagram: Standard B/G/I NICAM/FM mono only

Audio Processing Init

If the program is changed, resulting in another program within the Scandinavian System B/G no parameters of the MSP 3410 B have to be modified. To facilitate the check for NICAM the CCU has only to continue at the 'NICAM_START' instruction. During the 'NICAM_CHECK' 24

STOP Fig. 11­2: CCU software flow diagram: SAT-mode

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
11.6.5. Automatic Search Function for FM-Carrier Detection The AM demodulation ability of the MSP 3410 B offers the possibility to calculate the "field strength" of the momentarily selected FM carrier which can be read out by the CCU. In SAT receivers this feature can be used to realize an automatic FM carrier search. Therefore, the MSP has to be switched to AM-mode (Bit 8 of MODE_REG). The sound-IF frequency range must now be "scanned" in the MSP-channel 2 by means of the programmable quadrature mixer with an appropriate incremental frequency (i.e. 10 kHz).
1)

START LOAD_SEQ_1/2 Channel 1: NICAM Parameter

Audio Processing Init

NICAM_START

NICAM_READ
1)

After each incrementation there is a field strength value available at the DC level register FM1, which must be examined for relative maxima by the CCU. This results in either continuing search or switching the MSP back to FM demodulation mode. During the search process the FIR_REG_2 must be loaded with the coefficient set "AUTOSEARCH", which enables small bandwidth resulting in appropriate field strength characteristics. The absolute field strength value (can be read out of "DC Level Readout FM1") also gives information on whether a main FM carrier or a subcarrier was detected, and as a practical consequence the FM bandwidth (FIR_REG_1/2) and the deemphasis (50 µs or adaptive) can be switched automatically. For a detailed description of the automatic search function please refer to the corresponding MUBI program. 11.6.6. Automatic Standard Detection The AM demodulation ability of the MSP 3410 B enables also a simple method to decide between standard B/G (FM-carrier at 5.5 MHz) and standard I (FM-carrier at 6.0 MHz). It is achieved by tuning the MSP in the AM-mode to the two discrete frequencies and evaluating the field strength via the DC level register.

Pause

NICAM_READ

Yes NICAM_CHECK
NICAM ?

No LOAD_SEQ_1 Channel 1: FM2 Parameter

Pause

Yes

FM_ IDENT_CHECK
Pilot?

No LOAD_SEQ_1 Channel 1: NICAM Parameter

Fig. 11­3: CCU software flow diagram: Standard B/G with NICAM or FM stereo 1) The first READ could result in incorrect values.

ITT Semiconductors

25

MSP 3410 B
12. Programming the Audio Processing Part 12.1. Summary of the DSP Control Registers Control registers are 16 bit wide. Transmissions via I2C
Name Volume loudspeaker channel Balance loudspeaker channel [L/R] Bass loudspeaker channel Treble loudspeaker channel Loudness loudspeaker channel Spatial effect loudspeaker channel Volume headphone channel Volume SCART channel Loudspeaker channel source Loudspeaker channel matrix Headphone channel source Headphone channel matrix SCART channel source SCART channel matrix I2S I2S channel source channel matrix 000chex 000bhex 000ahex 0009hex I2C Bus Address 0000hex 0001hex 0002hex 0003hex 0004hex 0005hex 0006hex 0007hex 0008hex High/ Low H H H H H H H H H L H L H L H L H H H L

PRELIMINARY DATA SHEET

bus have to take place in 16 bit words. Single data entries are 8 bit. Some of the defined 16 bit words are divided into low and high byte, thus holding two different control entities.All control registers are readable.

Adjustable Range, Operational Modes [+12 dB ... ­94 dB, MUTE] [0..100% / 100% or 100% / 0..100%] [+12 dB ... ­12 dB] [+12 dB ... ­12 dB] [0 dB ... +17 dB] [OFF, ON] [+12 dB ... ­77 dB, MUTE] [00hex ... 7Fhex] [FM, NICAM, SCART, SBUS12, SBUS34, [SOUNDA, SOUNDB, STEREO] [FM, NICAM, SCART, SBUS12, SBUS34, [SOUNDA, SOUNDB, STEREO] [FM, NICAM, SCART, SBUS12, SBUS34, [SOUNDA, SOUNDB, STEREO] [FM, NICAM, SCART, SBUS12, SBUS34, [SOUNDA, SOUNDB, STEREO] [FM, NICAM, SCART, SBUS12, SBUS34, [00hex ... 7Fhex] [00hex ... 7Fhex] I2S] I2S] I2S] I2S] I2S]

Reset Mode MUTE
100%/100%

0 dB 0 dB 0 dB OFF MUTE 00hex FM SOUNDA FM SOUNDA FM SOUNDA FM SOUNDA FM (see note) 00hex 00hex

ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË Ë Ë ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË Ë ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË Ë ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË Ë ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË ËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËËË

ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ É É É É ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ É É É É
Quasi-peak detector source Prescale SCART Prescale FM FM matrix 000dhex 000ehex [NO_MAT, GSTEREO, KSTEREO] [OFF, 50 µs, 75 µs, J17] [OFF, WP1] [00hex ... 7Fhex] [OFF, J17] Bits [7..0] NO_MAT (see note) 50 µs OFF (s. note) 00hex 00hex Deemphasis FM Adaptive Deemphasis FM Prescale NICAM 0010hex 0011hex 0013hex 0014hex 0015hex 0016hex 000fhex H L H H H

ÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇ Ç Ç Ç Ç Ç Ç Ç ÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇ ÇÇÇÇÇÇ Ç Ç Ç Ç ÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇ ÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇÇ
Deemphasis NICAM J17 (s. note) ACB Register (SCART Switches and DIG_OUT Pins) Beeper H/L L [00hex ... 7Fhex]/[00hex ... 7Fhex] [B/G, M] reserved for future use 0/0 (s. note) B/G ­ Identification Mode Special SCART Mode

Unused parts of the 16 bit registers must be zero.
Note: For future compatibility to new technical codes of the MSP3410 B or the MSP3400 B some coefficients concerning features not implemented or not changeable yet must nevertheless be initialized. Please consider the following compatibility restrictions: ­ Quasi peak source must always be the same as the speaker source ­ NICAM deemphasis switching facility not yet implemented, NICAM deemphasis must be switched on ­ Panda1, if switched on, must always be activated together with 75 µs deemphasis ­ Panda1 must be switched off if NICAM is selected ­ FM dematrix must be switched off if Panda1 is selected ­ Beeper off: set frequency to 0 and volume to 0; ­ Beeper on: set frequency to 40hex and set volume; beeper frequency not yet variable

26

ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
Balance Loudspeaker Channel H 7Fhex 7Ehex 74hex 73hex 72hex Left 100%, Right 99.2% 26hex 15hex 0­ 14hex Left 100%, Right 0.8% Left 100%, Right muted 1000 0010 1000 0001 Balance loudspeaker channel [L/R] Left muted, Right 100% Left 0.8%, Right 100% Left 99.2%, Right 100% Left 100%, Right 100% 0001hex 0111 1111 0111 1110 0000 0001 0000 0000 RESET 1111 1111 H 7Fhex 7Ehex 01hex 00hex FFhex 82hex 81hex

Volume Loudspeaker Channel Volume loudspeaker channel +12 dB +11 dB +1 dB 0 dB ­1 dB ­77 dB ­94 dB Mute 0000hex 0111 1111 0111 1110 0111 0100 0111 0011 0111 0010 0010 0110 0001 0101 0000 0000... 0001 0100

The highest positive 8 bit number yields in a maximum possible gain of 12 dB. Decreasing the volume register by 1 LSB decreases volume by 1 dB. The minimum volume without loudness is ­77 dB. Together with loudness, the volume range can be increased by the actual loudness setting. Setting loudness to 17 dB, the lowest possible volume is ­94 dB. Volume settings lower than the given minimum mute the output. With large scale input signals, positive volume settings may lead to signal clipping. To prevent severe clipping effects with bass or treble boosts, the internal volume is automatically limited to a level where in combination with either bass or treble setting the amplification does not exceed 12 dB. For example: setting bass to +9 dB and treble to +5, the maximum possible volume is +3 dB. Values higher than +3 dB are internally limited to +3 dB. Please consider that even if the loudspeaker or the headphone or both channels are not used ( i.e. satellite receiver, video recorder), they must be initialized after reset according to the tables Volume Loudspeaker Channel shown above and Volume Headphone Channel on page 28.

Positive balance settings reduce the left channel without affecting the right channel, negative settings reduce the right channel leaving the left channel at 100%. A step by 1 LSB decreases or increases the balance by about 0.8% (exact figure: 100/127).

Bass Loudspeaker Channel Bass loudspeaker channel +12 dB +11 dB +1 dB 0 dB ­1 dB ­11 dB ­12 dB 0002hex 0110 0000 0101 1000 0000 1000 0000 0000 RESET 1111 1000 1010 1000 1010 0000 H 60hex 58hex 08hex 00hex F8hex A8hex A0hex

With positive bass settings internal overflow may occur even with overall volume less than 0 dB. This will lead to a clipped output signal. Therefore, it is not recommended to set bass to a value that, in conjunction with volume, would result in an overall positive gain.

ITT Semiconductors

27

MSP 3410 B
Treble Loudspeaker Channel Treble loudspeaker channel +12 dB +11 dB +1 dB 0 dB ­1 dB ­11 dB ­12 dB 0003hex 0110 0000 0101 1000 0000 1000 0000 0000 RESET 1111 1000 1010 1000 1010 0000 H 60hex 58hex 08hex 00hex F8hex A8hex A0hex Stereo Basewidth Enlargement (SBE) or Pseudo Stereo Effect (PSE)

PRELIMINARY DATA SHEET

Spatial Effects Loudspeaker Channel Spatial effect loudspeaker channel OFF 0005hex 0000 0000 RESET 0011 1111 H 00hex 3Fhex

The kind of spatial effect depends on the source mode. If the incoming signal is in mono mode, Pseudo Stereo Effect is active, for stereo signals Stereo Basewidth Enlargement is effective.

With positive treble settings internal overflow may occur even with overall volume less than 0 dB. This will lead to a clipped output signal. Therefore it is not recommended to set treble to a value that in conjunction with volume would result in a overall positive gain. Loudness Loudspeaker Channel Loudness loudspeaker channel +17 dB +16 dB +1 dB 0 dB 0004hex 0100 0100 0100 0000 0000 0100 0000 0000 RESET H 44hex 40hex

Volume Headphone Channel Volume Headphone Channel +12 dB +11 dB +1 dB 0 dB ­1 dB ­77 dB 0000hex 0111 1111 0111 1111 0111 0100 0111 0011 0111 0010 0010 0110 0000 0000... 0010 0101 H 7Fhex 7Ehex 74hex 73hex 72hex 26hex 0­ 25hex

04hex 00hex

Mute

Loudness increases the volume of low and high frequency signals while keeping the amplitude of the 1 kHz reference frequency constant. The intended loudness has to be set according to the actual volume setting. Because loudness introduces gain, it is not recommended to set loudness to a value that in conjunction with volume would result in a overall positive gain. Mode Loudness Normal (constant volume at 1 kHz) Super Bass (constant volume at 2 kHz) 00004hex 0000 0000 Reset 0000 0100 L 00hex 04hex

Volume SCART Channel Volume SCART channel OFF 0 dB gain (digital full scale (FS) to 2 VRMS output) +6 dB gain (­6 dBFS to 2 VRMS output) 0007hex 00hex RESET 40hex H

7Fhex

By means of `Mode Loudness', the corner frequency for bass amplification can be set to two different values. In Super Bass mode, the corner frequency is shifted up. The point of constant volume is shifted from 1 kHz to 2 kHz. 28

The highest positive 8 bit number yields in a maximum possible gain of 12 dB. Decreasing the volume register by 1 LSB decreases volume by 1 dB. The minimum volume is ­77 dB. Lower volume settings mute the output. With large scale input signals, positive volume settings may lead to signal clipping. ITT Semiconductors

PRELIMINARY DATA SHEET

MSP 3410 B
SCART Prescale 0008hex 0009hex 000ahex 000bhex 000chex 0000 0000 RESET 0000 0001 0000 0010 0000 0011 0000 0100 0000 0101 H H H H H 00hex 01hex 02hex 03hex 04hex 05hex Maximum Volume (28 kHz deviation 1) recommended FIRbandwidth: 130 kHz) Deviation 50 kHz1) recommended FIRbandwidth: 200 kHz Deviation 75 kHz1) recommended FIRbandwidth: 200 or 280 kHz Deviation 150 kHz1) recommended FIRbandwidth: 380 kHz Maximum deviation 192 kHz1) recommended FIRbandwidth: 380 kHz Prescale for adaptive deemphasis WP1 recommended FIRbandwidth: 130 kHz Volume Prescale SCART OFF 0 dB gain (2 VRMS input to digital full scale) +14 dB gain (400 mVRMS input to digital full scale) FM Prescale Volume Prescale FM (normal FM mode) OFF 000ehex 00hex RESET 7Fhex H 000dhex 00hex RESET 19hex 7Fhex H

Channel Source Modes Loudspeaker channel source Headphone channel source SCART channel source I2S channel source Quasi-peak d