• Smoothed transitions before and after digital mute (soft
mute)
• Fade function: duration-programmable (6 ms to 22.4 s
at 44.1 kHz) digital volume control (attenuation as well
as gain): +6 dB to −90 dB in steps of 0.375 dB with
automatic soft mute
• Digital balance: 0 dB to −22.5 dB in steps of −1.5 dB
(maximum overall attenuation combined with volume
control: −90 dB)
OLUME CONTROL FEATURES
) 256fs or 384f
sys
s
TDA1546T
1.3.2S
• Digital de-emphasis filter for three sample rates (32 kHz,
44.1 kHz or 48 kHz)
• Digital treble: −10.5 dB to +12 dB at 20 kHz; 16 steps
spaced at 1.5 dB
• Digital bass: −9 dB to +13.5 dB at 20 Hz; 16 steps
spaced at 1.5 dB
• Distortion-free digital dynamic bass boost: 0 dB to
+37 dB at 10 Hz; 15 steps spaced at 2 dB
• Can be used for loudness or dynamic digital bass boost
• Double-speed mode (e.g. for high-speed dubbing)
• Pseudo double-speed mode (for power saving
application)
• Digital speaker system mode including digital crossover
filter.
1.3.3S
• Spectrum analyzer for seven different frequency ranges
• Digital silence detection. Level (−48 dB to∞dB, in steps
of 3 dB) and duration (200 ms to 3.2 s, in steps of
200 ms at 44.1 kHz) programmable. Output via versatile
pins.
• Peak level detection and readout to microcontroller
(dB linear, 0 dB to −90 dB in steps of 1.5 dB)
• Digital overload detection. Level-programmable (dB
linear, −1.5 dB to−46.5 dB, in steps of 3 dB). Output via
versatile pins.
• Digital spectrum analyzer by combination of peak
detection and 7-band selective filter
• Optional combination spectrum analyzer and overload
detection for frequency-dependent overload detection.
OUND PROCESSING FEATURES
OUND MONITOR FEA TURES
2ORDERING INFORMATION
TYPE
NUMBER
TDA1546TSO28plastic small outline package; 28 leads; body width 7.5 mmSOT136-1
January 19954
NAMEDESCRIPTIONVERSION
PACKAGE
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
TDA1546T
digital sound processing (BCC-DAC)
3QUICK REFERENCE DATA
SYMBOLSPARAMETERCONDITIONSMIN.TYP.MAX.UNIT
V
DD
I
DDD
I
DDA
I
DDO
I
DDX
V
FS(rms)
(THD+N)/Stotal harmonic distortion
S/Nsignal-to-noise ratio at
t
dg
BRinput bit rate at data inputfs= 48 kHz;
f
sys
TC
FS
T
amb
supply voltagenote 13.85.05.5V
digital supply currentnote 2−40−mA
analog supply currentnote 2−5.5−mA
operational amplifier
note 2−6.5−mA
supply current
clock circuitry supply
note 2−1−mA
current
full-scale output voltage
VDD= 5 V1.4251.51.575V
(RMS value)
at 0 dB signal level−−88−81dB
plus noise-to-signal ratio
at −60 dB signal level;
−0.0040.009%
−−44−40dB
A-weighted
A-weighted; at code
bipolar zero
00000H
group delayfs= sample rate;
normal-speed
100108−dB
−−s
24
-----f
s
−−3.072Ms
normal-speed
fs= 48 kHz;
−−6.144Ms
double-speed
system clock frequency6.4−18.432MHz
full-scale temperature
−±100× 10−6−
coefficient at analog
outputs (VOL and VOR)
operating ambient
−20−+70°C
temperature
−1
−1
Notes
1. All V
and VSS pins must be connected to the same supply or ground respectively.
DD
2. Measured at input code 00000H and VDD=5V.
January 19955
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
4GENERAL DESCRIPTION
The TDA1546T is the first Bitstream Continuous
Calibration digital-to-analog converter (BCC-DAC) to
feature unique signal processing functions. In addition to
the basic functions of digital filtering and digital-to-analog
conversion, it offers such advanced digital signal
processing functions as volume control, tone control, bass
boost, peak or spectrum analyzer readout and many more
convenient functions. The digital processing features are
of high sound quality due to the wide dynamic range of the
bitstream conversion technique.
2
The TDA1546T accepts I
word lengths of up to 20 bits and various Japanese serial
data input formats with word lengths of 16, 18 and 20 bits.
The circuit can operate as a master or slave with different
system clocks (256fs or 384fs) and is therefore, eminently
suitable for use in various applications such as DCC, CD,
DAT and MD.
S-bus data input formats with
TDA1546T
The range of applications is further extended by an
incorporated Digital Speaker System mode (DSS) with
digital crossover filter.
Four cascaded FIR filters and a sample-and-hold function
increase the oversampling rate from 1f
(384fssystem clock) or 128fs (256fs system clock).
A second-order noise shaper converts this oversampled
data to a bitstream for the 5-bit DACs.
The DACs are of the continuous calibration type and
incorporate a special data coding technique, which
contributes to a high signal-to-noise ratio and dynamic
range.
On-board amplifiers convert the output current to a voltage
signal capable of driving a line output. Externally
connected capacitors perform the required first-order
filtering. Additional post filtering is not required.
to 96f
s
s
Fig.1 Digital audio reconstruction system using the TDA1546T.
January 19956
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
5BLOCK DIAGRAM
handbook, full pagewidth
VERS1
VERS0
SPECTRUM ANALYSER
dB CONVERTER
SILENCE
DETECTION
21
VERSATILE
20
OUTPUT
BCK5WS6DATA
4
PEAK
DETECTION
OVERLOAD
DETECTION
SOURCE
SELECTION
MULTIPLE FORMAT
INPUT INTERFACE
DE-EMPHASIS
VOLUME CONTROL
BALANCE
SOFT MUTE
TONE CONTROL 1
BASS
TONE CONTROL 2
TREBLE
TONE CONTROL 3
BASS BOOST
CKSL
7
CRYSTAL
OSCILLATOR
CLOCK
GENERATION
AND
DISTRIBUTION
TDA1546T
12
XTAL1
13
XTAL2
16
CDEC
14
V
DDX
15
V
SSX
10
V
DDD
9
V
SSD
TEST1
TEST2
1 nF
C
EXT1
1 µF
V
V
GND
OL
FIR FILTER
3
11
23FILTCL
2.2 kΩ
22
26
ref
TDA1546T
LINEAR INTERPOLATOR 8f to 16f
6 x OVERSAMPLING (SAMPLE-AND-HOLD)
SECOND ORDER NOISE SHAPER
DATA ENCODER
LEFT OUTPUT SWITCHES
16 (4-bit)
R
CONV1
V
27
SSO
CALIBRATED
CURRENT
SOURCES
16 (4-bit)
CALIBRATED
CURRENT
28
V
DDO
SINKS
STAGE 1:1f to 2f
STAGE 2:2f to 4f
STAGE 3:4f to 8f
ss
REFERENCE
SOURCE
ss
FIR FILTER
ss
FIR FILTER
s
s
LINEAR INTERPOLATOR 8f to 16f
6 x OVERSAMPLING (SAMPLE-AND-HOLD)
SECOND ORDER NOISE SHAPER
RIGHT OUTPUT SWITCHES
16 (4-bit)
CALIBRATED
CURRENT
SOURCES
16 (4-bit)
CALIBRATED
CURRENT
SINKS
MICROCONTROLLER
INTERFACE
DATA ENCODER
1
V
DDA
ss
R
V
SSA
CONV2
2.2 kΩ
OP1OP1
2
8
19
18
17
24 FILTCR
25
MLC782
POR
L3DATA
L3CLK
L3MODE
V
OR
C
1 nF
EXT2
Fig.2 Block diagram.
January 19957
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
7FUNCTIONAL DESCRIPTION
The TDA1546T CMOS digital-to-analog bitstream
converter incorporates an up-sampling digital filter and
noise shaper which increase the oversample rate of 1f
input data to 128fs in the normal-speed mode. This
high-rate oversampling, together with the 5-bit DAC,
enables the filtering required for waveform smoothing and
out-of-band noise reduction to be achieved by simple
first-order analog post-filtering.
In the double-speed mode, the input sample frequency is
twice that of the normal-speed mode, as is the signal
bandwidth. The TDA1546T is able to distinguish between
the two modes (by means of a special programming bit),
so that in the double-speed mode, only half the amount of
oversampling is applied, and digital filtering is applied over
double the bandwidth compared to normal-speed. Thus in
the double-speed mode, the input sample rate of 1fs input
data is up-sampled by a factor 64f
, achieving the same
s
absolute output sample frequency as in normal-speed
mode.
In the block diagram, Fig.2, a general subdivision into main
functional Sections is illustrated. The actual signal
processing takes place in the digital signal processing
block. The two blocks named microcontroller interface and
clock generation and distribution fulfil a general auxiliary
function to the audio data processing path. The
microcontroller interface provides access to all the blocks
that require, or allow, configuration or selection and
processes the data readout from the peak detection block,
all via a simple three-line interface. The clock generation
and distribution section, which is driven by the external
system clock or crystal oscillator, provides the data
processing blocks with time bases and controls the system
mode dependent frequency settings. The following
sections give detailed explanations of the operation of
each block and their setting options processed by the
microcontroller interface, the use of the microcontroller
interface and of the operation of the clock section with its
various system settings.
s
TDA1546T
7.1Clock generation and distribution
The TDA1546T has an internal clock generator that may
be used by connecting a crystal of 11.2896 MHz (256f
16.9344 MHz (384fs) between pins XTAL1 and XTAL2.
This mode is used when the TDA1546T is the master in
the system. The circuit diagram of Fig.4 shows the typical
connection of the external oscillator circuitry for master
mode operation.
Alternatively, the TDA1546T can also operate in slave
mode. Figure 5 shows how to connect for slave mode
operation. In this mode, pin XTAL1 receives an input clock
of 256 or 384f
(fs= 32, 44.1 or 48 kHz) and voltage levels
s
of 0 V to 5 V by AC coupling and attenuation.
The CDEC output (pin 16) contains a buffered version of
the system clock for external use. The clock selection pin
CKSL is used to select between system clock frequency
ratios. Its effect is shown in Table 1.
Table 1 System clock selection
PIN CKSLSYSTEM CLOCKCDEC OUTPUT
0256f
1384f
s
s
256f
384f
Fig.4 External crystal oscillator circuit.
) or
s
s
s
January 19959
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
Fig.5 External clock input connection.
7.2Power-on reset
The internal register file of the TDA1546T is initialized by a
power-on reset sequence which can be instigated via the
POR input pin 8. A LOW input on POR causes the reset
sequence to be active. This input has an internal
resistance to VDD to allow for passive use with only an
external capacitor connected between this pin and ground.
For correct detection by the TDA1546T internal controller,
the system clock must be running, andPOR should remain
LOW for at least one audio sample period before being
returned HIGH. Following detection another audio sample
period is needed to complete the initialization procedure,
after which the values of the various control bits in the
internal register file are at their predefined initial values
(see Section 7.3).
TDA1546T
7.3Microprocessor interface
The exchange of data and control information between the
TDA1546T and a microcontroller is accomplished through
a serial hardware interface comprising the following pins:
L3DATA: microcontroller interface bidirectional data
line.
Information transfer through the microcontroller bus is
organized according to the so-called ‘L3’ format, in which
two different modes of operation can be distinguished;
address mode and data transfer mode.
The address mode is required to select a device
communicating via the L3-bus and to determine the
direction of data transfer in data transfer mode. Data
transfer for the TDA1546T can be in two directions, input
to the TDA1546T to program its sound processing and
other functional features, and output from the TDA1546T
for transfer of audio peak data, which it has acquired and
processed, to the system microcontroller.
7.3.1A
The address mode is used to select a device for
subsequent data transfer and to define the direction of that
transfer as well as the source or destination registers. The
address mode is characterized by L3MODE being LOW
and a burst of 8 clock pulses on L3CLK, accompanied by
8 data bits. The fundamental timing is shown in Fig.6.
DDRESS MODE
Fig.6 Timing address mode.
January 199510
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
Data bits 0 to 1 indicate the type of the subsequent data
transfer as shown below in Table 2. The direction of the
channel status and user data transfers depends on the
transmit/receive mode.
Table 2 Selection of data exchange
(1)
BIT 1
X0data to TDA1546Tinput
X1data from
Note
1. Where X = don't care.
Data bits 2 to 7 represent a 6-bit device address, with bit 7
being the MSB and bit 2 the LSB. The address of the
BIT 0TRANSFERDIRECTION
output
TDA1546T
TDA1546T
TDA1546T is 000100 (bit 7 to bit 2). In the event that the
TDA1546T receives a different address, it immediately
3-states the L3DATA pin and deselects its microcontroller
interface logic. A dummy address of 000000 is defined for
the deselection of all devices that are connected to the
serial microcontroller bus.
7.3.2D
The selection performed in the address mode remains
active during subsequent data transfers, until the
TDA1546T receives a new address command. The
fundamental timing of data transfers is shown in Fig.7,
where L3DATA denotes the data from the TDA1546T to
the microcontroller (L3DATA write). The timing for the
opposite direction is essentially the same as in the address
mode (L3DATA read). The maximum input clock and data
rate is 64f
ATA TRANSFER MODE
(or 32fs when in the double-speed mode).
s
Fig.7 Timing for data transfer mode.
All transfers are bytewise, i.e. they are based on groups of
8 bits. Data will be stored in the TDA1546T after the eighth
bit of a byte has been received.
A multi-byte transfer is illustrated in Fig.8. The definition of
the L3 protocol allows for a so-called “halt” mode, as some
devices which are expected to connect to the same
microcontroller bus lines may require an indication of when
January 199511
8 bits have been transferred. This halt mode option is
implemented in the TDA1546T, meaning that subsequent
byte transfers must be separated by a period identified as
halt mode. A halt mode period is characterized by the
following conditions:
L3MODE = LOW, L3DATA = 3-state and L3CLK = HIGH.
Philips SemiconductorsPreliminary specification
Bitstream Continuous Calibration DAC with
digital sound processing (BCC-DAC)
Fig.8 Multibyte transfer.
7.3.3ORGANIZATION AND PROGRAMMING OF THE
INTERNAL REGISTER FILE
Command data received from the microcontroller is stored
in an internal register file (see Table 3) which is organized
as a page of 16 registers, each containing a 4-bit
command data word (D3 to D0).
TDA1546T
Access to the words in the register file involves selection
of the address of a register location (by means of A3, A2,
A1 and A0). A second page of 4 registers is accessible by
means of the extended address register bits (EA2, EA1
and EA0) and extended data register bits (ED3, ED2, ED1
and ED0).
Table 3 Microcontroller control register file
ADDRESS
D3D2D1D0
A3A2A1A0
00 00BAL3BAL2BAL1BAL01 1 1 1balance left
01BAR3BAR2BAR1BAR01 1 1 1balance right
10VC3VC2VC1VC01 1 1 1volume control
11VC7VC6VC5VC41 1 1 0volume control