Philips TDA8046 Service Manual

INTEGRATED CIRCUITS

DATA SH EET

TDA8046

Multi-mode QAM demodulator

Product speciﬁcation Supersededs data of 1996 Jul 23 File under Integrated Circuits, IC02

1996 Nov 19

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

CONTENTS

1 FEATURES 2 APPLICATION 3 QUICK REFERENCE DATA 4 ORDERING INFORMATION 5 BLOCK DIAGRAM 6 PINNING 7 FUNCTIONAL DESCRIPTION

7.1 Functional description of the individual blocks

7.1.1 Quadrature demodulator and half Nyquist filter

7.1.2 Equalizer

7.1.3 Lock detector

7.1.4 Carrier recovery

7.1.5 Clock recovery

7.1.6 AGC

7.1.7 Offset control

7.1.8 Loop amplifiers

7.1.9 Output formatter

7.1.10 Boundary scan

7.1.11 I2C-bus interface

7.1.12 I2C-bus write parameters

7.1.13 I2C-bus read parameters 8 LIMITING VALUES 9 THERMAL CHARACTERISTICS 10 DEMODULATOR AND HALF NYQUIST

FILTER CHARACTERISTICS 11 LOCK DETECTOR CHARACTERISTICS 12 CARRIER RECOVERY CHARACTERISTICS 13 CLOCK RECOVERY CHARACTERISTICS 14 AGC CHARACTERISTICS 15 INTEGRATED LOOP AMPLIFIERS

CHARACTERISTICS 16 CHARACTERISTICS OF DIGITAL INPUTS

AND OUTPUTS 17 PACKAGE OUTLINE 18 SOLDERING

18.1 Introduction

18.2 Reflow soldering

18.3 Wave soldering

18.4 Repairing soldered joints 19 DEFINITIONS 20 LIFE SUPPORT APPLICATIONS 21 PURCHASE OF PHILIPS I2C COMPONENTS

1996 Nov 19 2

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

1 FEATURES

• Different modulation schemes: 4, 16, 32,

64 and 256-QAM

• Digital demodulator and square root raised cosine

Nyquist filter with roll-off of 15% or 20%

• High performance adaptive equalizer (no training

sequence needed)

• Digital detectors for generation of required control

voltages for carrier recovery, clock recovery and AGC

• Input format: Straight binary or 2’s complement (up to 9 bits, TTL compatible)

• Output format: 8-bit wide bus (CMOS compatible)

C-bus interface to initialize and monitor the

• I demodulator. When no I2C-bus usage; 64-QAM, 20% roll-off factor in default mode

• 5 V peripheral and analog supply voltage

• 3.3 V core supply voltage

• Boundary scan test.

• Digital-to-analog converters and operational amplifiers

allowing high flexibility for selection of the (PLL) loop

2 APPLICATION

time constants

• High maximum symbol rate (r

) of 7 Msymbols/s

Demodulation for digital cable TV and cable modem.

3 QUICK REFERENCE DATA

SYMBOL PARAMETER CONDITIONS MIN. TYP. MAX. UNIT

DDD(core)

DDD

DDA

DDD(core)

DDD

DDA

core supply voltage 3.00 3.30 3.60 V digital peripheral supply voltage 4.75 5.00 5.25 V analog supply voltage 4.75 5.00 5.25 V core supply current V digital peripheral supply current V analog supply current V

DDD(core) DDD DDA

= 3.3 V; note 1 − 100 − mA = 5 V; note 1 − 14 − mA = 5 V; note 1 − 16 − mA

symbol rate −−7 Msym/s IL implementation loss note 2 − 0.7 − dB α Nyquist roll-off (programmable) − 15 or 20 − % SNR

lock

signal-to-noise ratio for locking a

21 −−dB

64-QAM constellation

signal-to-noise ratio for locking a

27 −−dB

256-QAM constellation

Notes

1. The supply currents are specified for the maximum symbol frequency.

2. The implementation loss (IL) of the demodulator is defined as the distance between the measured and theoretical BER curve as function of signal-to-noise ratio at a BER = 10 This performance depends on the chosen loop parameters (see

−6

for a back-to-back measurement at the IF frequency.

Application notes

4 ORDERING INFORMATION

TYPE

NUMBER

NAME DESCRIPTION VERSION

TDA8046H QFP64 plastic quad ﬂat package; 64 leads (lead length 1.95 mm);

PACKAGE

SOT319-2

body 14 × 20 × 2.8 mm

1996 Nov 19 3

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

5 BLOCK DIAGRAM

book, full pagewidth

SSD1 to 12

DDD1 to 9

TMS

TDO

TDI

TRST

TCK

TEST3

TEST2

TEST1

7, 12, 14, 17,

24, 26, 31, 34,

46, 50, 61, 64

6, 13, 16, 

25, 33, 38, 

45, 51, 63

SCAN TEST

BOUNDARY

C-BUS 

TDA8046

CONTROL

DO0

DO7 to

27 to 30

20 to 23

FINE AGC

CONTROL

OFFSET

CONTROL

SQUARE ROOT

to DACs

internal clock for

digital processing

srs

CLOCK GENERATOR

CLKSDV

CLKOUT

OUTPUT

FORMATTER

FINE AGC

EQUALIZER

OFFSET

PHASE

DIGITAL

ROTATOR

SQUARE ROOT

RAISED COSINE

DEMODULATOR

INPUT

TATION

REPRESEN-

CARRIER

RECOVERY



NCO

CONTROL

CLOCK

RECOVERY

AGC

COARSE

ref

ref3

DAC

ref1Iref2Iref3

BIAS

GENERATOR

ref

ANALOG SECTION

ref

ref2

DAC

ref

ref1

DAC

565552

5857

5453

MGG198

CARREC

CARTC

BIAS

CLKREC

CLKTC

SSA

DDA

AGC

AGCTC

Fig.1 Block diagram.

SCL

SDA

CLK

CLKADC

1 to 5,

8 to 11

DIN0

to

DIN8

1996 Nov 19 4

PRESET

CLKT

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

6 PINNING

SYMBOL PIN I/O DESCRIPTION

DIN0 1 I digital input bit 0 (LSB) DIN1 2 I digital input bit 1 DIN2 3 I digital input bit 2 DIN3 4 I digital input bit 3 DIN4 5 I digital input bit 4 V

DDD1

SSD1

DIN5 8 I digital input bit 5 DIN6 9 I digital input bit 6 DIN7 10 I digital input bit 7 DIN8 11 I digital input bit 8 (MSB) V

SSD2

DDD2

SSD3

CLKADC 15 O clock output to ADC (4 × r V

DDD3

SSD4

CLKSDV 18 O clock symbol data valid output CLKT 19 I for test purpose only DO7 20 O parallel data output (bit 7) DO6 21 O parallel data output (bit 6) DO5 22 O parallel data output (bit 5) DO4 23 O parallel data output (bit 4) V

SSD5

DDD4

SSD6

DO3 27 O parallel data output (bit 3) DO2 28 O parallel data output (bit 2) DO1 29 O parallel data output (bit 1) DO0 30 O parallel data output (bit 0) V

SSD7

CLKOUT 32 I output formatter clock output V

DDD5

SSD8

SCL 35 I serial clock input (I SDA 36 I/O serial data input/output (I A0 37 I hardware address input (I V

DDD6

TEST3 39 I test input 3 (normally connected to ground) TEST2 40 I test input 2 (normally connected to ground)

6 supply digital peripheral supply voltage 1 (+5 V) 7 supply digital ground 1; for input peripheral and core

12 supply digital ground 2; for core and clock buffers 13 supply digital supply voltage 2; for core and clock buffers (+3.3 V) 14 supply digital peripheral ground 3

)

16 supply digital peripheral supply voltage 3 (+5 V) 17 supply digital ground 4; for core

24 supply digital peripheral ground 5 25 supply digital peripheral supply voltage 4 (+5 V) 26 supply digital ground 6; for core

31 supply digital peripheral ground 7

33 supply digital peripheral supply voltage 5 (+5 V) 34 supply digital peripheral ground 8

C-bus)

38 supply digital peripheral supply voltage 6 (+5 V)

1996 Nov 19 5

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

SYMBOL PIN I/O DESCRIPTION

TEST1 41 I test input 1 input (normally connected to ground) TRST 42 I optional asynchronous reset input TCK 43 I dedicated test clock input TMS 44 I input control signal V

DDD7

SSD9

TDO 47 O serial test data output TDI 48 I serial test data input PRESET 49 I set device into default mode input V

SSD10

DDD8

BIAS

AGCTC

AGC

CARTC

CARREC

CLKTC

CLKREC

SSA

DDA

SSD11

CLK 62 I clock input (4 × r V

DDD9

SSD12

45 supply digital supply voltage 7; for core (+3.3 V) 46 supply digital ground 9; for core

50 supply digital ground 10; for the digital section of the analog block 51 supply digital supply voltage 8; for the digital section of the analog block (+5 V) 52 I input bias current for DACs 53 O inverted operational ampliﬁer input voltage for loop ﬁltering 54 O analog output voltage for AGC 55 O inverted operational ampliﬁer input voltage for carrier recovery loop

ﬁltering 56 O analog output voltage for carrier recovery 57 O inverted operational ampliﬁer input voltage for clock recovery loop

ﬁltering 58 O analog output voltage for clock recovery 59 supply analog ground 60 supply analog supply voltage (+5 V) 61 supply digital ground 11; for clock

)

63 supply digital supply voltage 9; for clock 64 supply digital peripheral ground 12

1996 Nov 19 6

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

DDD9

handbook, full pagewidth

DIN0 DIN1

DIN2 DIN3

DIN4

DDD1

SSD1

DIN5 DIN6 DIN7 DIN8

SSD2

DDD2

SSD3

CLKADC

DDD3

SSD4

CLKSDV

CLKT

SSD12

V 64

63 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19

CLK

SSD11

V 61

DDA

V 60

TDA8046

SSA

V 59

CLKREC

V 58

CLKTC

V 56

CARREC

CARTC

V 55

AGC

V 54

AGCTC

V 53

BIAS

I 52

51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33

MGG197

DDD8

SSD10

PRESET TDI TDO V

SSD9

DDD7

TMS TCK TRST TEST1 TEST2 TEST3 V

DDD6

A0 SDA SCL V

SSD8

DDD5

DO7

DO6

DO5

DO4

SSD5

DDD4

Fig.2 Pin configuration.

1996 Nov 19 7

SSD6

DO3

DO2

DO1

DO0

SSD7

CLKOUT

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

7 FUNCTIONAL DESCRIPTION

Figure 3 shows the application of the TDA8046 multi-mode QAM demodulator. The frequency of the IF signal (IF

) is down converted to a frequency that

QAM

equals the symbol rate (rs) by a mixer which is driven from a local oscillator with a frequency of f

CAR=fIF+rs

. After low pass filtering this baseband signal is applied to an external 8 or 9-bit ADC.

For 256-QAM, a 9-bit ADC is preferred, for the other modes an 8-bit ADC is sufficient.

The multi-mode QAM demodulator has digital detectors for AGC, carrier recovery and clock recovery. The on-chip DACs translate the detector values to analog control

handbook, full pagewidth

RF

signal

SAWTUNER

QAM

currents which are then integrated by a loop filter. To perform this loop filtering, an operational amplifier is integrated after each DAC.

The carrier recovery consists of a two-loop system. The outer loop is shown in Fig.3, and controls both phase and frequency at a low speed. The inner loop controls the carrier phase at a high speed (wide loop bandwidth).

The AGC also consists of two loops; the outer loop is the coarse AGC and one inner loop is the fine AGC.

The recovered symbols are converted into bits according to a demapping scheme and represented at the output in an 8-bit parallel output format. The QAM demodulator can

be initialized and monitored by the I

8 or 9 bits

clk

CAR

= fIF + r

LPF ADC

C-bus interface.

clock recovery

carrier recovery

AGC

TDA8046

C-BUS

Fig.3 Application with multi-mode QAM demodulator.

DO7 to DO0 CLKOUT CLKSDV

MGG167

1996 Nov 19 8

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

7.1 Functional description of the individual blocks

The functional block diagram of the multi-mode QAM demodulator is illustrated in Fig.1. This section describes the individual blocks in the demodulator. After adaptation for the used input format (2’s complement or binary), the input signal is demodulated in the I and Q baseband signals which are applied to the inputs of the half-Nyquist filter (equals square root raised cosine). To avoid overloading of the ADC, an AGC detector is placed after the adaptation for the input format. The control value for the clock recovery is generated after half Nyquist filtering. The echoes created in the cable network are reduced significantly in the equalizer.

The equalizer produces a ‘clean’ constellation diagram from which the information for the carrier recovery is derived. This constellation is also applied to the output formatter which demaps the transmitted symbols in corresponding bits. The carrier recovery and lock detection functions are based on the equalizer output. The output of the equalizer is applied to an output formatter, which translates the symbol bits to a FEC input format. The digital outputs of the clock recovery, AGC, and carrier recovery section are converted into currents which are integrated by the loop filters. To make these loop filters active, operational amplifiers are integrated on the chip.

The TDA8046 can handle five different digital modulation schemes; 4, 16, 32, 64 and 256-QAM. These schemes

are selectable via the I

7.1.1 Q

UADRATURE DEMODULATOR AND HALF NYQUIST

FILTER

C-bus interface.

Quadrature demodulation is accomplished after selection of the appropriate input format via the I2C-bus. The in-phase and quadrature components are both applied to a half Nyquist filter. In default mode, this filter gives a 20% roll-off half Nyquist shaping. The basic schematic of the quadrature demodulator followed by the half Nyquist filter is shown in Fig.4. The signs of the multiplication factors in the Q-branch can be inverted (I2C-bus bit INVD).

When using an 8-bit ADC the LSB of the 9-bit input word should be connected to the positive supply (V

DDD

). This ensures a symmetrical 2’s complement representation which can be multiplied by −1 in a correct (2’s complement) way. The overall transfer function of the square root raised cosine filters is shown in Figs 5 and 6.

For characteristics see Chapter 10.

handbook, full pagewidth

COMPLEMENT

BINARY OR

TWO's

I2C-BUS

+1, 0, −1, 0 0, −1, 0, +1

I2C-BUS

DIN8

to

DIN0

Fig.4 Schematic diagram of the quadrature demodulator and half Nyquist filter.

1996 Nov 19 9

C-BUS

HALF NYQUIST

FILTER

HALF NYQUIST

FILTER

C-BUS

MGG168

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

handbook, full pagewidth

5 0

−5

relative

gain (dB)

−15

−25

−35

−45

−55

010.25 0.5 0.75 1.751.51.25 relative frequency

Fig.5 Half Nyquist receiver filter transfer function (20% roll-off).

MBG987

f )

(

handbook, full pagewidth

relative

gain (dB)

−10

−20

−30

−40

−50

0 0.5 1 1.50.25 0.75 1.25 1.75

Fig.6 Half Nyquist receiver filter transfer function (15% roll-off).

relative frequency

MGG169

f )

(

1996 Nov 19 10

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

7.1.2 EQUALIZER This function is realized with a T spaced 12 or 14 taps

(selected via the I2C-bus) adaptive filter with a feedback part. The equaliser is based on a Decision Feedback Equalizer (DFE) structure with Least Mean Square (LMS) coefficient updating algorithm. No training sequence is required. The block schematic of the total equalizer is shown in Fig.8. The main tap of the equalizer is adjustable for fine AGC function (6 dB AGC range). The settings of the equalizer taps can be read via the I2C-bus. If the

equalizer diverges, an alarm bit is set (I

C-bus bit ALEQ) and an automatic reset of the taps can be performed (I2C-bus bit EAR).

To improve acquisition time, the convergence steps of the FFE/DFE parts of the equalizer are programmable via the I2C-bus. When the system locks, the steps are automatically modified for optimum performances.

Besides reading the equalizer tap values, the main tap of the equalizer can also be programmed. After setting the main tap, the other coefficients can be set to zero. The equalizer settings can also be frozen via the I2C-bus.

The equalizer has been proven to work correctly under bad channel conditions as indicated in Table 1. It is guaranteed that all loops (including equalizer) converge at a SNR of 21 dB for a 64-QAM modulation format and 27 dB for a 256-QAM modulation format.

Table 1 Channel echo proﬁle

DELAY AMPLITUDE PHASE

⁄8× T

2 × T

⁄8× T

⁄8× T ⁄8× T

sym

sym sym

0.08 130°

0.20 60°

0.05 310°

0.10 200°

0.03 200°

Figure 7 represents the QAM spectrum seen by the equalizer. It corresponds (in the frequency domain) to the multiplication of a full nyquist spectrum by the impulse response of the channel specified in Table 1.

handbook, full pagewidth

relative

gain (dB)

−1

−3

−5

−7

−9

−11

−0.5 0.5

−0.375 0.375−0.125 0.125−0.25 0.250

Fig.7 QAM spectrum with echo profile as seen by the equalizer.

relative frequency

MGD636

f )

(

1996 Nov 19 11

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

handbook, full pagewidth

input

FEED

FORWARD

EQUALIZER

TAPS CALCULATION

−

Fig.8 DFE equalizer structure.

7.1.3 LOCK DETECTOR The lock detector indicates whether all algorithms in the

demodulator are converged or not. For a symbol error rate (at the input of the demodulator) smaller than 2 × 10−2, the detector will give the indication ‘LOCK’ (I2C-bus bit LK = 1). For larger symbol error rates, the detector will generate the ‘UNLOCK’ signal (I2C-bus bit LK = 0). It should ne noted that this ‘UNLOCK’ signal is generated before any other part of the demodulator loses lock. The lock detector is part of the carrier recovery loop, see Fig.9. The Lock Detector Threshold (LDT) can be changed

with the help of the I

C-bus. The estimation algorithm used in the lock detector also provides information about the SER ratio which can be read out via the I2C-bus interface.

For characteristics see Chapter 11.

7.1.4 C

ARRIER RECOVERY

The carrier recovery detector consists of a Phase-Frequency Detector (PFD) and Phase Detector (PD). Depending on the mode of operation, the carrier recovery is switched either between the phase frequency (no lock) or the phase detector (lock). The carrier recovery consists of the following two loops:

DECISION

FEEDBACK

EQUALIZER

TAPS CALCULATION

decision

MGG170

output

1. The outer loop; this loop controls the phase and frequency of the incoming QAM signal at the IF frequency in such a way that the constellation is optimally positioned for detection.

2. The inner loop; the bandwidth of this loop can be large and can therefore reduce the influence of large bandwidth phase noise.

A fully digital carrier recovery function is also possible and can be selected via the I

C-bus. Should this configuration be used, then the external components of the loop filter will not have to be implemented.

Four different maximum DAC output currents can be selected via the I2C-bus. The output currents of the DAC are defined in such a way that a VCO with a behaviour as shown in Fig.9 can be connected directly to the output of the integrated operational amplifier. Should the VCO slope be negative then the sign of the current can be inverted by the I2C-bus. Figure 10 defines the DAC output currents.

For characteristics see Chapter 12.

1996 Nov 19 12

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

handbook, full pagewidth

QAM

VCO

LPF

ADC

external

DAC

CAR

I2C-BUS

DEMODULATION

AND

FILTERING

I2C-BUS

ref

DIGITAL

INNER LOOP

EQUALIZER

ref1

lock

LOCK

PHASE

FREQUENCY

DETECTOR

PHASE

DETECTOR

lock

Fig.9 Schematic diagram of the carrier recovery.

I2C-BUS

C-BUS

MGG171

1996 Nov 19 13

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

handbook, full pagewidth

CARI = 1

= positive output current.

pos

= negative output current.

neg

–()

posIneg

------------------------------ -

∆ I

-------------------------------- -

+()

posIneg

–()

posIneg

CAR

DAC output

current

VCO

CARI = 0

CAR

CARREC

MGG180

−1/

−I

CAR

digital input

CAR

100×=

Fig.10 Definition of the DAC currents and the expected frequency behaviour of the VCO.

1996 Nov 19 14

Philips Semiconductors Product speciﬁcation

Multi-mode QAM demodulator TDA8046

7.1.5 CLOCK RECOVERY The clock recovery function uses the unequalized I and Q

signals, i.e. the half Nyquist filter outputs (see Fig.4). The clock recovery section generates a control value each symbol period. As this algorithm is based on the energy maximization, both main and mid symbols are required at the input. Consequently, the input data rate is twice the symbol rate. The schematic diagram of this detector is illustrated in Fig.11.

handbook, full pagewidth

external

CLOCK

RECOVERY

DETECTOR

DAC

rsI

The clock generator generates the required internal clocks from the VCXO clock signal at 4 × r

. The input stage

amplifier of this generator enables the designer to supply a low amplitude oscillator signal to the TDA8046. The DAC output current range (I

) can be varied via the I2C-bus.

CLK

The sign of the output current can also be inverted to adjust for the correct sign of the VCXO slope.

For characteristics see Chapter 13.

to

VCXO

ref3

CLK

ref

2 4

Fig.11 Schematic diagram of the clock recovery.

from

VCXO

MGG172

1996 Nov 19 15

+ 33 hidden pages