The STA350BW is an integrated solution of digital audio processing, digital amplifier control,
and FFX-power output stage, thereby creating a high-power single-chip FFX
™
solution
comprising high-quality, high-efficiency, and all-digital amplification.
The STA350BW is based on an FFX (fully flexible amplification) processor, a proprietary
technology from STMicroelectronics. FFX is the evolution and the enlargement of ST’s
ternary technology: the new processor can be configured to work in ternary, binary, binary
differential and phase-shift PWM modulation schemes.
The STA350BW contains the ternary, binary and binary differential implementations, a
subset of the full capability of the FFX processor.
The STA350BW is part of the Sound Terminal
®
family that provides full digital audio
streaming to the speaker, offering cost effectiveness, low power dissipation and sound
enrichment.
The STA350BW power section consists of four independent half-bridges. These can be
configured via digital control to operate in different modes. 2.1 channels can be provided by
two half-bridges and a single full-bridge, providing up to 2 x 18 W + 1 x 40 W of music output
power, by using standard 4 and 8 Ω speakers. Two channels can be provided by two fullbridges, providing up to 2 x 50 W of music power, by using standard 6 Ω speaker or
2 x 40 W by using 8 Ω speakers at 25 V. The IC can also be configured as 2.1 channels with
2 x 40 W provided by the device and external power for FFX power drive. If configured as
mono-BTL, the latter is capable of providing up to 1 x 90 W on a standard 3 Ω load or
1 x 75 W by using a 4 Ω, setting the supply voltage at 25 V. Please refer to the package
thermal characteristics and application suggestions for more details.
Also provided in the STA350BW are a full assortment of digital processing features. This
includes up to 8 programmable biquads (EQ) per channel. Special digital signal processing
techniques are available in order to manage low-frequency quantization noise in case of
very low frequency cutoff filter thresholds. The coefficient range -4..4 allows the easy
implementation of high shelf filters. Available presets allow the advantage of earlier time-tomarket by substantially reducing the amount of software development needed for certain
functions. This includes audio preset volume loudness, preset volume curves and preset EQ
settings. There are also new advanced AM radio interference reduction modes. Dual-band
DRC dynamically equalizes the system to provide speaker linear frequency response
regardless of output power level. This feature independently processes the two bands,
controlling dynamically the output power level in each band and so providing better sound
clarity.
The serial audio data input interface accepts all possible formats, including the popular I
2
S
format. Three channels of FFX processing are provided. This high-quality conversion from
PCM audio to FFX PWM switching waveform provides over 100 dB SNR and dynamic
range.
10/86Doc ID 018572 Rev 3
STA350BWDescription
1.1 Block diagram
Figure 1.Block diagram
I2S
in t er face
Vo lu me
contr ol
PLL
FFX
I2C
Power
control
Protection
cur re nt/the rm al
Log ic
Regulators
Bias
PowerDigita l DSP
Channel
1A
Chann el
1B
Ch anne l
2A
Chann el
2B
AM045167v1
Doc ID 018572 Rev 311/86
Pin connectionsSTA350BW
2 Pin connections
2.1 Connection diagram
Figure 2.Pin connection PowerSSO-36 (top view)
GND_SUB
SA
TEST_MODE
VSS
VCC_REG
OUT2B
GND2
VCC2
OUT2A
OUT1B
VCC1
GND1
OUT1A
GND_REG
VDD
CONFIG
OUT3B/FFX3B
OUT3A/FFX3A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
D05AU1638
VDD_DIG
GND_DIG
SCL
SDA
INT_LINE
RESET
SDI
LRCKI
BICKI
XTI
GND_PLL
FILTER_PLL
VDD_PLL
PWRDN
GND_DIG
VDD_DIG
TWARN/OUT4A
EAPD/OUT4B
AM045168v1
2.2 Pin description
Table 1.Pin description
PinTypeNameDescription
1GNDGND_SUBSubstrate ground
2I SAI
3ITEST_MODEThis pin must be connected to ground (pull-down)
4I/OVSSInternal reference at Vcc-3.3 V
5I/OVCC_REGInternal Vcc reference
6OOUT2BOutput half-bridge 2B
7GNDGND2Power negative supply
8PowerVCC2Power positive supply
9OOUT2AOutput half-bridge 2A
10OOUT1BOutput half-bridge 1B
12/86Doc ID 018572 Rev 3
2
C select address (pull-down)
STA350BWPin connections
Table 1.Pin description (continued)
PinTypeNameDescription
11PowerVCC1Power positive supply
12GNDGND1Power negative supply
13OOUT1AOutput half-bridge 1A
14GNDGND_REGInternal ground reference
15PowerVDDInternal 3.3 V reference voltage
16ICONFIGParallel mode command
17OOUT3B/FFX3BPWM out CH3B / external bridge driver
18OOUT3A/FFX3APWM out CH3A / external bridge driver
19OEAPD/OUT4BPower-down for external bridge / PWM out CH4B
20I/OTWARN/OUT4A
21PowerVDD_DIGDigital supply voltage
22GNDGND_DIGDigital ground
23IPWRDNPower down (pull-up)
24PowerVDD_PLLPositive supply for PLL
Thermal warning from external bridge (pull-up when input)
/ PWM out CH4A
25IFILTER_PLLConnection to PLL filter
26GNDGND_PLLNegative supply for PLL
27IXTIPLL input clock
2
28IBICKII
29ILRCKII
30ISDII
S serial clock
2
S left/right clock
2
S serial data channels 1 and 2
31IRESETReset (pull-up)
32OINT_LINEFault interrupt
2
33I/OSDAI
34ISCLI
C serial data
2
C serial clock
35GNDGND_DIGDigital ground
36PowerVDD_DIGDigital supply voltage
Doc ID 018572 Rev 313/86
Electrical specificationsSTA350BW
3 Electrical specifications
3.1 Absolute maximum ratings
Table 2.Absolute maximum ratings
Symbol Parameter MinTypMaxUnit
V
VDD_DIG Digital supply voltage-0.3 4 V
VDD_PLL PLL supply voltage-0.34
T
T
Power supply voltage (VCCxA, VCCxB) -0.330 V
cc
Operating junction temperature -20150 °C
op
Storage temperature -40 150 °C
stg
Warning:Stresses beyond those listed in Tabl e 2 above may cause
permanent damage to the device. These are stress ratings
only, and functional operation of the device at these or any
other conditions beyond those indicated under
“Recommended operating conditions” are not implied.
Exposure to AMR conditions for extended periods may affect
device reliability. In the real application, power supplies with
nominal values rated within the recommended operating
conditions may rise beyond the maximum operating
conditions for a short time when no or very low current is
sunk (amplifier in mute state). In this case the reliability of the
device is guaranteed, provided that the absolute maximum
ratings are not exceeded.
3.2 Thermal data
Table 3.Thermal data
SymbolParameterMinTypMaxUnit
R
th j-case
T
th-sdj
T
th-w
T
th-sdh
R
th j-amb
1. See Section 8: Package thermal characteristics on page 82 for details.
Table 6.Electrical specifications - power section (continued)
SymbolParameter ConditionsMin Typ Max Unit
FFX stereo mode,
<5 kHz
X
TA L K
Crosstalk
PSRRPower Supply Rejection Ratio
Peak efficiency, FFX mode
η
Peak efficiency, binary modes
1. Refer to Figure 3: Test circuit.
2. Limit current if the register (OCRB Section 7.3.3) overcurrent warning detect adjustment bypass is enabled. When disabled
refer to the Isc.
One channel driven
at 1 W
Other channel
measured
FFX stereo mode,
<5 kHz
VRipple01V RMS
Audio input = dither
only
Po = 2 x 25 W
into 8 Ω
Po = 2 x 10W into 4Ω
+ 1 x 20W into 8 Ω
80dB
80dB
90
%
86
Figure 3.Test circuit
Low cu rrent dea d tim e = M A X (tr,tf)
Low cu rrent dea d tim e = M A X (tr,tf)
D u ty cycle = 50%
D u ty cycle = 50%
INx Y
INx Y
MP
MP
MN
MN
GND
GND
+VCC
+VCC
OUTxY
OUTxY
OUTxY
OUTxY
trtf
trtf
R= 8
R= 8
Ω
Ω
+
+
-
vdc = VCC/2
vdc = VCC/2
VCC
VCC
0.9*VCC
0.9*VCC
VCC/2
VCC/2
0.1*VCC
0.1*VCC
t
t
AM045169v1
Doc ID 018572 Rev 317/86
Characterization curvesSTA350BW
4 Characterization curves
The following characterization curves were made using the STA350BW demonstration
board with 2.0 channels (refer to the schematic in Figure 6) under the following test
conditions:
V
= 25 V, f = 1 kHz, fSW = 384 kHz, Tamb = 25 °C and RL = 6 Ω, unless otherwise
CC
specified.
Figure 4.Demonstration board, 2.0 channels
18/86Doc ID 018572 Rev 3
AM045290v1
STA350BWCharacterization curves
Figure 5.Mono parallel BTL schematic
AM045170v1
Doc ID 018572 Rev 319/86
Characterization curvesSTA350BW
Figure 6.THD+N vs. output power (VCC = 25 V, load = 6 Ω)
10
5
2
1
0.5
%
0.2
0.1
0.05
0.02
0.01
1m1002m5m10m 20m50m 100m 200m500m 125102050
Figure 7.THD+N vs. output power (V
10
5
Vcc=25V , Load = 6Ω,
Freq= 1KHz
W
= 18 V, load = 8 Ω)
CC
AM045171v1
2
1
0.5
%
0.2
0.1
0.05
0.02
0.01
1m502m5m10m 20m50m 100m 200m500m1251020
Vcc=18V , Load = 8Ω,
Freq= 1KHz
W
AM045172v1
20/86Doc ID 018572 Rev 3
STA350BWCharacterization curves
Figure 8.Output power vs. V
60
55
50
45
40
35
W
30
25
20
15
10
5
+6+26+8+10+12+14+16+18+20+22+24
Load= 6 Ω,
Freq=1KHz
Figure 9.Output power vs. V
(load = 6 Ω)
CC
(load = 8 Ω)
CC
THD = 10%
Vdc
THD = 1%
AM045173v1
50
45
40
Load= 8 Ω,
Freq=1KHz
35
30
W
25
20
15
THD = 10%
THD = 1%
10
5
+6+26+8+10+12+14+16+18+20+22+24
Vdc
AM045174v1
Doc ID 018572 Rev 321/86
Characterization curvesSTA350BW
Figure 10. Efficiency vs. output power (VCC = 25 V, load = 6 Ω)
+1
+0.9
+0.8
+0.7
+0.6
η %
+0.5
+0.4
Vcc= 25V,
Load= 6 Ω,
Freq=1KHz
+0.3
+0.2
+0.1
+0
555101520253035404550
Figure 11. Efficiency vs. output power (V
+1
+0.9
+0.8
+0.7
+0.6
η %
+0.5
+0.4
Vcc= 25V,
Load= 8 Ω,
Freq=1KHz
+0.3
+0.2
+0.1
+0
54510152025303540
W
= 25 V, load = 8 Ω)
CC
W
AM045175v1
AM045176v1
22/86Doc ID 018572 Rev 3
STA350BWCharacterization curves
4.1 Mono parallel BTL characteristics
Figure 12. THD+N vs. output power (VCC = 25 V, load = 3 Ω)
10
5
2
Vcc=25V
Load= 3 Ω,
1
0.5
%
0.2
0.1
0.05
0.02
0.01
1m1002m5m 10m 20m50m 100m 200m500m125102050
Figure 13. Output power vs. V
110
100
90
80
Load= 3 Ω,
Freq=1KHz
(load = 3 Ω)
CC
Freq=1KHz
W
THD=10%
AM045177v1
70
60
W
50
40
30
20
10
+6+26+8+10+12+14+16+18+20+22+24
Vdc
THD=1%
AM045178v1
Doc ID 018572 Rev 323/86
Characterization curvesSTA350BW
Figure 14. Efficiency vs. output power (VCC = 26 V, load = 3 Ω)
+1
+0.9
+0.8
+0.7
+0.6
+0.5
η %
+0.4
+0.3
Vcc= 26V
Load= 3 Ω,
Freq=1KHz
+0.2
+0.1
+0
101102030405060708090100
Figure 15. Efficiency vs. output power (V
+1
+0.9
+0.8
+0.7
+0.6
+0.5
η %
+0.4
+0.3
Vcc=18V
Load= 3 Ω,
Freq=1KHz
+0.2
+0.1
+0
5501015202530354045
W
= 18 V, load = 3 Ω)
CC
W
AM045179v1
AM045180v1
24/86Doc ID 018572 Rev 3
STA350BWProcessing data paths
5 Processing data paths
Figure 16 and 17 illustrate the data processing paths inside the STA350BW.
The whole processing chain is composed of two consecutive sections. In the first one dualchannel processing is implemented, as described below, and then each channel is fed into
the post-mixing block allowing to generate either a third channel (typically used in 2.1 output
configuration and with crossover filters enabled) or to have the channels processed by the
dual-band DRC block (2.0 output configuration with crossover filters used to define the cutoff
frequency of the two bands).
The first section begins with a 2x oversampling FIR filter allowing for 2*Fs audio processing.
Then a selectable high-pass filter removes the DC level (enabled if HFB=0).
The channel 1 and 2 processing chain can include up to 8 filters, depending on the selected
configuration (bits BQL, BQ5, BQ6, BQ7 and XO[3:0]).
By default 4 independent filters per channel are enabled, plus the pre-configured DeEmphasis, Bass and Treble controls (BQL=0, BQ5=0, BQ6=0, BQ7=0).
If the coefficient sets are linked (BQL=1) it’s then possible to use De-Emphasis, Bass and
Treble filter in a user-defined configuration (provided the relevant BQx bits are set). In other
words both channels will use the same processing coefficients and can have up to 7 filters
each. Note that if BQL=0 the BQx bits are ignored and the 5th, 6th and 7th filters are
configured as, respectively, De-Emphasis, Bass and Treble controls.
Moreover the common 8th filter, from the subsequent processing section, can be available
on both channels (provided the pre-defined crossover frequencies are not used, XO[3:0]=0,
and the dual-band DRC is not used).
In the second section mixing and crossover filters are available. If B
(Figure 17), they are fully user-programmable and allow generating a third channel (2.1
outputs). Alternatively, in B
2
DRC mode, those blocks will be used to split the sub-band and
2
DRC is not enabled
define the cutoff frequencies of the two bands. A prescaler and a final post scaler allow full
control over the signal dynamic respectively before and after the filtering stages. A mixer
function is also available.
Figure 16. Left and right processing - part 1
Sampling
Sampling
frequency=Fs
frequency=Fs
From
From
I2S input
I2S input
interface
interface
x2
FIR
x2
over
FIR
sampling
over
x2
FIR
x2
over
FIR
sampling
over
Sampling
Sampling
frequency=2xFs
frequency=2xFs
PreScale
PreScale
PreScale
PreScale
Hi-Pass
Hi-Pass
Filter
Filter
If HPB=0
Hi-Pass
Hi-Pass
Filter
Filter
If HPB=0
Biquad#1Biquad#2Biquad#3Biquad
Biquad#1Biquad#2Biquad#3Biquad
User Defined Filters
If DSPB=0 and C1EQBP=0
Biquad#1Biquad#2Biquad#3Biquad
Biquad#1Biquad#2Biquad#3Biquad
User Defined Filters
If DSPB=0 and C2EQBP=0
#4
#4
#4
#4
If BQ5=1
and BQL=1
Biquad
#5
De-Emph.
If DEMP=0
If BQ5=1
and BQL=1
Biquad
#5
De-Emph.
If DEMP=0
IF BQ7=1
If BQ6=1
and BQL=1
and BQL=1
Biquad
Biquad
#7
#6
BassTreble
If C1TCB=0
BTC: Bass Boost/Cut
TTC: Treble Boost/Cut
If BQ6=1
IF BQ7=1
and BQL=1
and BQL=1
Biquad
Biquad
#6
#7
BassTreble
If C2TCB=0
BTC: Bass Boost/Cut
TTC: Treble Boost/Cut
L
L
L
R
AM045181v1
Doc ID 018572 Rev 325/86
Processing data pathsSTA350BW
Figure 17. Processing - part 2
Dual-band DRC enabled
C1Mx1
C1Mx1=
L
L
R
R
0x7fffff
+
+
C1Mx2
C1Mx2=
0x00000
C2Mx1
C2Mx1=
0x000000
+
+
C2Mx2
C2Mx2=
0x7fffff
C3Mx1
C3Mx1=
0x40000
+
+
C3Mx2
C3Mx2=
0x400000
User-Defined Mix Coefficients
User-Defined Mix Coefficients
Dual-band DRC disabled
C1Mx1
L
L
R
R
C1Mx1
+
+
C1Mx2
C1Mx2
Hi-Pass XO
B2DRC
Filter
Hi-pass
Hi-Pass XO
B2DRC
Hi-pass
Crossover Frequency determined by XO Setting
User Defined If XO=0000
User Defined If XO=0000
filter
Filter
filter
Channel ½
Hi-Pass XO
Biquad#5
Filter
--------------
Hi-pass XO
filter
+
+
CH3
Volume
CH1
Volume
CH2
Volume
CH3
Volume
DRC2
DRC1
DRC1
Vol
Vol
And
And
Limiter
Limiter
Vol
And
Limiter
DRC2
+
+
Postscale
Post scale
Postscale
Post scale
Postscale
Post scale
C2Mx1
C2Mx1
+
+
C2Mx2
C2Mx2
C3Mx1
C3Mx1
+
+
C3Mx2
C3Mx2
User-Defined Mix Coefficients
User-Defined Mix Coefficients
Channel ½
Hi-Pass XO
Biquad#5
Filter
--------------
Hi-pass XO
filter
Channel 3
Lo-Pass XO
Biquad
Filter
--------------
Low-pass XO
filter
Crossover Frequency determined by XO Setting
User Defined If XO=0000
User Defined If XO=0000
26/86Doc ID 018572 Rev 3
Vol
Vol
And
And
Limiter
Limiter
Vol
Vol
And
And
Limiter
Limiter
Postscale
Post scale
Postscale
Post scale
B2DRC Disabled
AM045182v1
STA350BWI2C bus specification
6 I2C bus specification
The STA350BW supports the I2C protocol via the input ports SCL and SDA_IN (master to
slave) and the output port SDA_OUT (slave to master). This protocol defines any device that
sends data to the bus as a transmitter and any device that reads the data as a receiver. The
device that controls the data transfer is known as the master and the other as the slave. The
master always starts the transfer and provides data to the serial clock for synchronization.
The STA350BW is always a slave device in all of its communications. It supports up to
400 kb/sec rate (fast-mode bit rate). The STA350BW I
interface works properly only in the case that the master clock generated by the PLL has a
frequency 10 times higher compared to the frequency of the applied SCL signal.
6.1 Communication protocol
6.1.1 Data transition or change
Data changes on the SDA line must only occur when the SCL clock is low. An SDA transition
while the clock is high is used to identify a START or STOP condition.
2
C is a slave-only interface. The I2C
6.1.2 Start condition
START is identified by a high-to-low transition of the data bus SDA signal while the clock
signal SCL is stable in the high state. A START condition must precede any command for
data transfer.
6.1.3 Stop condition
STOP is identified by a low-to-high transition of the data bus SDA signal while the clock
signal SCL is stable in the high state. A STOP condition terminates communication between
the STA350BW and the bus master.
6.1.4 Data input
During data input the STA350BW samples the SDA signal on the rising edge of clock SCL.
For correct device operation the SDA signal must be stable during the rising edge of the
clock and the data can change only when the SCL line is low.
6.2 Device addressing
To start communication between the master and the STA350BW, the master must initiate a
start condition. Following this, the master sends 8 bits (MSB first) corresponding to the
device select address and read or write mode to the SDA line.
The seven most significant bits are the device address identifiers, corresponding to the I
bus definition. In the STA350BW the I
the SA port configuration, 0x38 when SA = 0, and 0x3A when SA = 1.
2
C interface has two device addresses depending on
2
C
The eighth bit (LSB) identifies the read or write operation RW, this bit is set to 1 in read
mode and to 0 for write mode. After a START condition the STA350BW identifies on the bus
the device address and if a match is found, it acknowledges the identification on the SDA
Doc ID 018572 Rev 327/86
I2C bus specificationSTA350BW
bus during the 9th bit time. The byte following the device identification byte is the internal
space address.
6.3 Write operation
Following the START condition the master sends a device select code with the RW bit set
to 0. The STA350BW acknowledges this and then writes the byte of the internal address.
After receiving the internal byte address the STA350BW again responds with an
acknowledgement.
6.3.1 Byte write
In the byte write mode the master sends one data byte which is acknowledged by the
STA350BW. The master then terminates the transfer by generating a STOP condition.
6.3.2 Multi-byte write
The multi-byte write modes can start from any internal address. The master generating a
STOP condition terminates the transfer.
6.4 Read operation
6.4.1 Current address byte read
Following the START condition the master sends a device select code with the RW bit set
to 1. The STA350BW acknowledges this and then responds by sending one byte of data.
The master then terminates the transfer by generating a STOP condition.
6.4.2 Current address multi-byte read
The multi-byte read modes can start from any internal address. Sequential data bytes are
read from sequential addresses within the STA350BW. The master acknowledges each
data byte read and then generates a STOP condition, terminating the transfer.
6.4.3 Random address byte read
Following the START condition the master sends a device select code with the RW bit set
to 0. The STA350BW acknowledges this and then the master writes the internal address
byte. After receiving the internal byte address, the STA350BW again responds with an
acknowledgement. The master then initiates another START condition and sends the device
select code with the RW bit set to 1. The STA350BW acknowledges this and then responds
by sending one byte of data. The master then terminates the transfer by generating a STOP
condition.
6.4.4 Random address multi-byte read
The multi-byte read modes can start from any internal address. Sequential data bytes are
read from sequential addresses within the STA350BW. The master acknowledges each
data byte read and then generates a STOP condition, terminating the transfer.
Selects the ratio between the input I
frequency and the input clock.
2
S sample
2R/W 0MCS2
The STA350BW supports sample rates of 32 kHz, 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz,
176.4 kHz, and 192 kHz. Therefore the internal clock is:
●32.768 MHz for 32 kHz
●45.1584 MHz for 44.1 kHz, 88.2 kHz, and 176.4 kHz
●49.152 MHz for 48 kHz, 96 kHz, and 192 kHz
The external clock frequency provided to the XTI pin must be a multiple of the input sample
frequency (f
).
s
The relationship between the input clock and the input sample rate is determined by both
the MCSx and the IR (input rate) register bits. The MCSx bits determine the PLL factor
generating the internal clock and the IR bit determines the oversampling ratio used
internally.
Selects internal interpolation ratio based on input I
sample frequency
The STA350BW has variable interpolation (oversampling) settings such that internal
processing and FFX output rates remain consistent. The first processing block interpolates
by either 2-times or 1-time (pass-through) or provides a 2-times downsample. The
oversampling ratio of this interpolation is determined by the IR bits.
Table 11.IR bit settings as a function of input sample rate
Input sample rate fs (kHz)IR1st stage interpolation ratio
If the thermal warning adjustment is enabled (TWAB = 0), then the thermal warning
recovery determines if the -3 dB output limit is removed when thermal warning is negative.
If TWRB = 0 and TWAB = 0, then when a thermal warning disappears the -3 dB output limit
is removed and the gain is added back to the system. If TWRB = 1 and TWAB = 0, then
when a thermal warning disappears the -3 dB output limit remains until TWRB is changed to
zero or the device is reset.
The on-chip STA350BW power output block provides feedback to the digital controller using
inputs to the power control block. Input TWARN is used to indicate a thermal warning
condition. When TWARN is asserted (set to 0) for a period of time greater than 400 ms, the
power control block forces a -3 dB output limit (determined by TWOCL in the coefficient
RAM) to the modulation limit in an attempt to eliminate the thermal warning condition. Once
the thermal warning output limit adjustment is applied, it remains in this state until reset,
unless FDRB = 0.
The on-chip STA350BW power output block provides feedback to the digital controller using
inputs to the power control block. The FAULT input is used to indicate a fault condition (either
over-current or thermal). When FAULT is asserted (set to 0), the power control block
attempts a recovery from the fault by asserting the tri-state output (setting it to 0 which
directs the power output block to begin recovery), holds it at 0 for period of time in the range
of 0.1 ms to 1 second as defined by the fault-detect recovery constant register (FDRC
registers 0x29-0x2A), then toggles it back to 1. This sequence is repeated as long as the
fault indication exists. This feature is enabled by default but can be bypassed by setting the
FDRB control bit to 1.
7.2 Configuration register B (addr 0x01)
D7D6D5D4D3D2D1D0
C2IMC1IMDSCKESAIFBSAI3SAI2SAI1SAI0
10000000
7.2.1 Serial audio input interface format
Table 15.Serial audio input interface
BitR/WRSTNameDescription
0R/W0SAI0
1R/W0SAI1
2R/W0SAI2
3R/W0SAI3
Doc ID 018572 Rev 335/86
Determines the interface format of the input serial
digital audio interface.
Register descriptionSTA350BW
7.2.2 Serial data interface
The STA350BW audio serial input was designed to interface with standard digital audio
components and to accept a number of serial data formats. The STA350BW always acts as
the slave when receiving audio input from standard digital audio components. Serial data for
two channels is provided using three inputs: left/right clock LRCKI, serial clock BICKI, and
serial data 1 and 2 SDI12.
The SAI bits (D3 to D0) and the SAIFB bit (D4) are used to specify the serial data format.
The default serial data format is I
that follow.
7.2.3 Serial data first bit
Table 16.Serial data first bit
SAIFBFormat
0MSB-first
1LSB-first
Table 17.Support serial audio input formats for MSB-first (SAIFB = 0)
BICKISAI [3:0]SAIFBInterface format
32 * fs
48 * fs
2
S, MSB-first. Available formats are shown in the tables
00000I
2
S 15-bit data
00010Left/right-justified 16-bit data
00000I
2
S 16 to 23-bit data
00010Left-justified 16 to 24-bit data
00100Right-justified 24-bit data
01100Right-justified 20-bit data
10100Right-justified 18-bit data
11100Right-justified 16-bit data
00000I2S 16 to 24-bit data
00010Left-justified 16 to 24-bit data
00100Right-justified 24-bit data
64 * fs
01100Right-justified 20-bit data
10100Right-justified 18-bit data
11100Right-justified 16-bit data
36/86Doc ID 018572 Rev 3
STA350BWRegister description
Table 18.Supported serial audio input formats for LSB-first (SAIFB = 1)
BICKISAI [3:0]SAIFBInterface format
32 * fs
11001I
11101Left/right-justified 16-bit data
01001I
01001I
10001I
11001LSB first I
00011Left-justified 24-bit data
01011Left-justified 20-bit data
48 * fs
10011Left-justified 18-bit data
11011Left-justified 16-bit data
00101Right-justified 24-bit data
01101Right-justified 20-bit data
10101Right-justified 18-bit data
11101Right-justified 16-bit data
2
S 15-bit data
2
S 23-bit data
2
S 20-bit data
2
S 18-bit data
2
S 16-bit data
00001I2S 24-bit data
2
01001I
10001I
11001LSB first I
S 20-bit data
2
S 18-bit data
2
S 16-bit data
00011Left-justified 24-bit data
01011Left-justified 20-bit data
64 * fs
10011Left-justified 18-bit data
11011Left-justified 16-bit data
00101Right-justified 24-bit data
01101Right-justified 20-bit data
10101Right-justified 18-bit data
11101Right-justified 16-bit data
To make the STA350BW work properly, the serial audio interface LRCKI clock must be
synchronous to the PLL output clock. It means that:
■ the frequency of PLL clock / frequency of LRCKI = N ±4 cycles,
where N depends on the settings in Table 11 on page 34
■ the PLL must be locked.
If these two conditions are not met, and the IDE bit (reg 0x05 bit 2) is set to 1, the
STA350BW will immediately mute the I
2
S PCM data out (provided to the processing block)
and it will freeze any active processing task.
Doc ID 018572 Rev 337/86
Register descriptionSTA350BW
To avoid any audio side effects (like pop noise), it is strongly recommended to soft-mute any
audio streams flowing into the STA350BW data path before the desynchronization event
happens. At the same time any processing related to the I
2
C configuration should be issued
only after the serial audio interface and the internal PLL are synchronous again.
Note:Any mute or volume change causes some delay in the completion of the I
2
C operation due
to the soft-volume feature. The soft-volume phase change must be finished before any clock
desynchronization.
7.2.4 Delay serial clock enable
Table 19.Delay serial clock enable
BitR/WRSTNameDescription
0: No serial clock delay
5R/W 0DSCKE
1: Serial clock delay by 1 core clock cycle to tolerate
anomalies in some I2S master devices
7.2.5 Channel input mapping
Table 20.Channel input mapping
BitR/WRSTNameDescription
2
6R/W 0C1IM
7R/W 1C2IM
0: Processing channel 1 receives Left I
1: Processing channel 1 receives Right I
0: Processing channel 2 receives Left I
1: Processing channel 2 receives Right I2S Input
S Input
2
S Input
2
S Input
Each channel received via I2S can be mapped to any internal processing channel via the
Channel Input Mapping registers. This allows for flexibility in processing. The default
settings of these registers map each I
2
S input channel to its corresponding processing
channel.
38/86Doc ID 018572 Rev 3
STA350BWRegister description
7.3 Configuration register C (addr 0x02)
D7D6D5D4D3D2D1D0
OCRBCSZ3CSZ2CSZ1CSZ0OM1OM0
1011111
7.3.1 FFX power output mode
Table 21.FFX power output mode
BitR/WRSTNameDescription
0R/W 1OM0
Selects configuration of FFX output.
1R/W 1OM1
The FFX power output mode selects how the FFX output timing is configured.
Different power devices use different output modes.
Table 22.Output modes
OM[1,0]Output stage mode
00Drop compensation
01Discrete output stage - tapered compensation
10Full power mode
11Variable drop compensation (CSZx bits)
7.3.2 FFX compensating pulse size register
Table 23.FFX compensating pulse size bits
BitR/WRSTNameDescription
2R/W 1CSZ0
3R/W 1CSZ1
4R/W 1CSZ2
5R/W 0CSZ3
When OM[1,0] = 11, this register determines the
size of the FFX compensating pulse from 0 clock
ticks to 15 clock ticks.
Table 6:
Table 24.Compensating pulse size
CSZ[3:0]Compensating pulse size
00000 ns (0 tick) compensating pulse size
000120 ns (1 tick) clock period compensating pulse size
……
1111300 ns (15 tick) clock period compensating pulse size
The OCWARN input is used to indicate an overcurrent warning condition. When OCWARN
is asserted (set to 0), the power control block forces an adjustment to the modulation limit
(default is -3 dB) in an attempt to eliminate the overcurrent warning condition. Once the
overcurrent warning volume adjustment is applied, it remains in this state until a reset
occurs. The level of adjustment can be changed via the TWOCL (thermal warning/over
current limit) setting which is address 0x37 of the user-defined coefficient RAM.
7.4 Configuration register D (addr 0x03)
D7D6D5D4D3D2D1D0
SMEZDEDRCBQLPSLDSPBDEMPHPB
01000000
7.4.1 High-pass filter bypass
Table 26.High-pass filter bypass
BitR/WRSTNameDescription
0R/W 0HPB
Setting of one bypasses internal AC coupling digital
high-pass filter
The STA350BW features an internal digital high-pass filter for the purpose of AC coupling.
The purpose of this filter is to prevent DC signals from passing through an FFX amplifier. DC
signals can cause speaker damage. When HPB = 0, this filter is enabled.
7.4.2 De-emphasis
Table 27.De-emphasis
BitR/WRSTNameDescription
1R/W 0DEMP
40/86Doc ID 018572 Rev 3
0: No de-emphasis
1: Enable de-emphasis on all channels
STA350BWRegister description
7.4.3 DSP bypass
Table 28.DSP bypass
BitR/WRSTNameDescription
2R/W0DSPB
Setting the DSPB bit to 1 bypasses the EQ function of the STA350BW.
7.4.4 Post-scale link
Table 29.Post-scale link
BitR/WRSTNameDescription
3R/W0PSL
Post-scale functionality can be used for power-supply error correction. For multi-channel
applications running off the same power-supply, the post-scale values can be linked to the
value of channel 1 for ease of use and to update the values faster.
7.4.5 Biquad coefficient link
Table 30.Biquad coefficient link
BitR/WRSTNameDescription
4R/W 0BQL
0: Normal operation
1: Bypass of biquad and bass/treble functions
0: Each channel uses individual post-scale value
1: Each channel uses channel 1 post-scale value
0: Each channel uses coefficient values
1: Each channel uses channel 1 coefficient values
For ease of use, all channels can use the biquad coefficients loaded into the Channel-1
coefficient RAM space by setting the BQL bit to 1. Therefore, any EQ updates only have to
be performed once.
7.4.6 Dynamic range compression/anti-clipping bit
Table 31.Dynamic range compression/anti-clipping bit
BitR/WRSTNameDescription
5R/W0DRC
Both limiters can be used in one of two ways, anti-clipping or dynamic range compression.
When used in anti-clipping mode the limiter threshold values are constant and dependent on
the limiter settings. In dynamic range compression mode the limiter threshold values vary
with the volume settings allowing a nighttime listening mode that provides a reduction in the
dynamic range regardless of the volume level.
Doc ID 018572 Rev 341/86
0: Limiters act in anti-clipping mode
1: Limiters act in dynamic range compression mode
Register descriptionSTA350BW
7.4.7 Zero-detect mute enable
Table 32.Zero-detect mute enable
BitR/WRSTNameDescription
6R/W1ZDESetting of 1 enables the automatic zero-detect mute
Setting the ZDE bit enables the zero-detect automatic mute. The zero-detect circuit looks at
the data for each processing channel at the output of the crossover (bass management)
filter. If any channel receives 2048 consecutive zero value samples (regardless of fs) then
that individual channel is muted if this function is enabled.
7.4.8 Submix mode enable
Table 33.Submix mode enable
BitR/WRSTNameDescription
7R/W 0SME
0: Sub Mix into Left/Right disabled
1: Sub Mix into Left/Right enabled
7.5 Configuration register E (addr 0x04)
D7D6D5D4D3D2D1D0
SVEZCEDCCVPWMSAMENSBWMPCMPCV
11000010
7.5.1 Max power correction variable
Table 34.Max power correction variable
BitR/WRSTNameDescription
0R/W 0MPCV
7.5.2 Max power correction
Table 35.Max power correction
BitR/WRSTNameDescription
0: Use standard MPC coefficient
1: Use MPCC bits for MPC coefficient
1R/W 1MPC
Setting of 1 enables power bridge correction for THD
reduction near maximum power output.
Setting the MPC bit turns on special processing that corrects the STA350BW power device
at high power. This mode should lower the THD+N of a full FFX system at maximum power
output and slightly below. If enabled, MPC is operational in all output modes except tapered
(OM[1,0] = 01) and binary. When OCFG = 00, MPC will not effect channels 3 and 4, the lineout channels.
42/86Doc ID 018572 Rev 3
STA350BWRegister description
7.5.3 Noise-shaper bandwidth selection
Table 36.Noise-shaper bandwidth selection
BitR/WRSTNameDescription
2R/W0NSBW
7.5.4 AM mode enable
Table 37.AM mode enable
BitR/WRSTNameDescription
3R/W 0AME
The STA350BW features an FFX processing mode that minimizes the amount of noise
generated in the frequency range of AM radio. This mode is intended for use when FFX is
operating in a device with an AM tuner active. The SNR of the FFX processing is reduced to
approximately 83 dB in this mode, which is still greater than the SNR of AM radio.
7.5.5 PWM speed mode
Table 38.PWM speed mode
BitR/WRSTNameDescription
4R/W 0PWMS
1: Third order NS
0: Fourth order NS
0: Normal FFX operation
1: AM reduction mode FFX operation
0: Normal speed (384 kHz) all channels
1: Odd speed (341.3 kHz) all channels
7.5.6 Distortion compensation variable enable
Table 39.Distortion compensation variable enable
BitR/WRSTNameDescription
5R/W 0DCCV
0: Use preset DC coefficient
1: Use DCC coefficient
7.5.7 Zero-crossing volume enable
Table 40.Zero-crossing volume enable
BitR/WRSTNameDescription
6R/W 1ZCE
The ZCE bit enables zero-crossing volume adjustments. When volume is adjusted on digital
zero-crossings, no clicks are audible.
Doc ID 018572 Rev 343/86
1: Volume adjustments only occur at digital zero-crossings
0: Volume adjustments occur immediately
Register descriptionSTA350BW
7.5.8 Soft-volume update enable
Table 41.Soft-volume update enable
BitR/WRSTNameDescription
7R/W 1SVE
1: Volume adjustments ramp according to SVR settings
0: Volume adjustments occur immediately
Note:To the left of the arrow is the processing channel. When using channel output mapping, any
of the three processing channel outputs can be used for any of the three inputs.
44/86Doc ID 018572 Rev 3
STA350BWRegister description
Figure 20. OCFG = 00 (default value)
OUT1A
OUT1A
Half
Half
Bridge
Bridge
Channel 1
LPF
LPF
LPF
LPF
Channel 1
Channel 2
Channel 2
LineOut1
LineOut1
LineOut2
LineOut2
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
OUT1B
OUT1B
OUT2A
OUT2A
OUT2B
OUT2B
OUT3A
OUT3A
OUT3B
OUT3B
OUT4A
OUT4A
OUT4B
OUT4B
AM045185v1
Figure 21. OCFG = 01
Half
Half
Bridge
Bridge
OUT1A
OUT1A
Channel 1
Channel 1
Figure 22. OCFG = 10
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
OUT1A
OUT1A
OUT1B
OUT1B
OUT2A
OUT2A
OUT2B
OUT2B
OUT3A
OUT3A
OUT3B
OUT3B
EAPD
EAPD
Power
Power
Device
Device
OUT1B
OUT1B
OUT2A
OUT2A
OUT2B
OUT2B
Channel 1
Channel 1
Channel 2
Channel 2
Channel 3
Channel 3
Channel 2
Channel 2
Channel 3
Channel 3
AM045186v1
AM045187v1
Doc ID 018572 Rev 345/86
Register descriptionSTA350BW
Figure 23. OCFG = 11
OUT1A
OUT1A
Half
Half
Bridge
Bridge
OUT1B
OUT2B
OUT2B
OUT3A
OUT3A
OUT3B
OUT3B
OUT4A
OUT4A
OUT4B
OUT4B
OUT1B
OUT2A
OUT2A
Channel 1
Channel 1
Channel 2
Channel 2
Channel 3
Channel 3
AM045188v1
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
Half
Half
Bridge
Bridge
The STA350BW can be configured to support different output configurations. For each PWM
output channel a PWM slot is defined. A PWM slot is always 1 / (8 * fs) seconds length. The
PWM slot defines the maximum extension for the PWM rising and falling edge, that is, the
rising edge as well as the falling edge cannot range outside the PWM slot boundaries.
Figure 24. Output mapping scheme
FFX1A
FFX1 B
FFX2 A
FFX™
FFX™
FFX™
FFX™
FFX™
FFX ™
modulator
modulator
modulator
modulator
modulator
modulator
FFX 2B
FFX3 A
FFX3B
FFX4 A
FFX 4B
REMAP
REMAP
REMAP
REMAP
REMAP
REMAP
OUT1A
OUT1B
OUT2A
OUT2B
OUT3A
OUT3B
OUT4A
OUT4B
Power
Power
Power
Power
Power
Power
Bridge
Bridge
Bridge
Bridge
Bridge
Bridge
OUT1A
OUT1B
OUT2A
OUT2B
For each configuration the PWM signals from the digital driver are mapped in different ways
to the power stage.
46/86Doc ID 018572 Rev 3
AM045189v1
STA350BWRegister description
2.0 channels, two full bridges (OCFG = 00)
●FFX1A -> OUT1A
●FFX1B -> OUT1B
●FFX2A -> OUT2A
●FFX2B -> OUT2B
●FFX3A -> OUT3A
●FFX3B -> OUT3B
●FFX4A -> OUT4A
●FFX4B -> OUT4B
●FFX1A/1B configured as ternary
●FFX2A/2B configured as ternary
●FFX3A/3B configured as lineout ternary
●FFX4A/4B configured as lineout ternary
On channel 3 line out (LOC bits = 00) the same data as channel 1 processing is sent. On
channel 4 line out (LOC bits = 00) the same data as channel 2 processing is sent. In this
configuration, neither volume control nor EQ has any effect on channels 3 and 4.
In this configuration the PWM slot phase is the following as shown in Figure 25.
Figure 25. 2.0 channels (OCFG = 00) PWM slots
OUT1A
OUT1A
OUT1B
OUT1B
OUT2A
OUT2A
OUT2B
OUT2B
OUT3A
OUT3A
OUT3B
OUT3B
OUT4A
OUT4A
OUT4B
OUT4B
AM045190v1
Doc ID 018572 Rev 347/86
Register descriptionSTA350BW
2.1 channels, two half-bridges + one full-bridge (OCFG = 01)
●FFX1A -> OUT1A
●FFX2A -> OUT1B
●FFX3A -> OUT2A
●FFX3B -> OUT2B
●FFX1A -> OUT3A
●FFX1B -> OUT3B
●FFX2A -> OUT4A
●FFX2B -> OUT4B
●FFX1A/1B configured as binary
●FFX2A/2B configured as binary
●FFX3A/3B configured as binary
●FFX4A/4B is not used
In this configuration, channel 3 has full control (volume, EQ, etc…). On OUT3/OUT4
channels the channel 1 and channel 2 PWM are replicated.
In this configuration the PWM slot phase is the following as shown in Figure 26.
Figure 26. 2.1 channels (OCFG = 01) PWM slots
OUT1A
OUT1A
OUT1A
OUT1A
OUT1A
OUT1A
OUT1B
OUT1B
OUT1B
OUT1B
OUT1B
OUT1B
OUT2A
OUT2A
OUT2A
OUT2A
OUT2A
OUT2A
OUT2B
OUT2B
OUT2B
OUT2B
OUT2B
OUT2B
OUT3A
OUT3A
OUT3A
OUT3A
OUT3A
OUT3A
OUT3B
OUT3B
OUT3B
OUT3B
OUT3B
OUT3B
OUT4A
OUT4A
OUT4A
OUT4A
OUT4A
OUT4B
OUT4B
OUT4B
OUT4B
OUT4B
AM045191v1
48/86Doc ID 018572 Rev 3
STA350BWRegister description
2.1 channels, two full-bridge + one external full-bridge (OCFG = 10)
●FFX1A -> OUT1A
●FFX1B -> OUT1B
●FFX2A -> OUT2A
●FFX2B -> OUT2B
●FFX3A -> OUT3A
●FFX3B -> OUT3B
●EAPD -> OUT4A
●TWARN -> OUT4B
●FFX1A/1B configured as ternary
●FFX2A/2B configured as ternary
●FFX3A/3B configured as ternary
●FFX4A/4B is not used
In this configuration, channel 3 has full control (volume, EQ, etc…). On OUT4 channel the
external bridge control signals are muxed.
In this configuration the PWM slot phase is the following as shown in Figure 27.
Figure 27. 2.1 channels (OCFG = 10) PWM slots
OUT1A
OUT1A
OUT1A
OUT1A
OUT1B
OUT1B
OUT1B
OUT1B
OUT2A
OUT2A
OUT2A
OUT2A
OUT2B
OUT2B
OUT2B
OUT2B
OUT3A
OUT3A
OUT3A
OUT3A
OUT3B
OUT3B
OUT3B
OUT3B
AM045192v1
Doc ID 018572 Rev 349/86
Register descriptionSTA350BW
7.6.2 Invalid input detect mute enable
Table 44.Invalid input detect mute enable
BitR/WRSTNameDescription
2R/W1IDESetting of 1 enables the automatic invalid input detect mute
Setting the IDE bit enables this function, which looks at the input I2S data and automatically
mutes if the signals are perceived as invalid.
7.6.3 Binary output mode clock loss detection
Table 45.Binary output mode clock loss detection
BitR/WRSTNameDescription
3R/W1BCLEBinary output mode clock loss detection enable
The BCLE bit detects loss of input MCLK in binary mode and will output 50% duty cycle.
7.6.4 LRCK double trigger protection
Table 46.LRCK double trigger protection
BitR/WRSTNameDescription
4R/W1LDTELRCLK double trigger protection enable
The LDTE bit actively prevents double triggering of the LRCLK.
7.6.5 Auto EAPD on clock loss
Table 47.Auto EAPD on clock loss
BitR/WRSTNameDescription
5R/W0ECLEAuto EAPD on clock loss
When active, the ECLE bit issues a power device power-down signal (EAPD) on clock loss
detection.
7.6.6 IC power-down
Table 48.IC power-down
BitR/WRSTNameDescription
7R/W 1PWDN
0: IC power-down low-power condition
1: IC normal operation
The PWDN register is used to place the IC in a low-power state. When PWDN is written
as 0, the output begins a soft-mute. After the mute condition is reached, EAPD is asserted
50/86Doc ID 018572 Rev 3
STA350BWRegister description
to power down the power-stage, then the master clock to all internal hardware except the
2
I
C block is gated. This places the IC in a very low power consumption state.
7.6.7 External amplifier power-down
Table 49.External amplifier power-down
BitR/WRSTNameDescription
7R/W0EAPD
0: External power stage power down active
1: Normal operation
The EAPD register directly disables/enables the internal power circuitry.
When EAPD = 0, the internal power section is placed in a low-power state (disabled). This
register also controls the FFX4B/EAPD output pin when OCFG = 10.
7.7 Volume control registers (addr 0x06 - 0x0A)
7.7.1 Mute/line output configuration register
D7D6D5D4D3D2D1D0
LOC1LOC0ReservedReservedC3MC2MC1MMMUTE
00000000
Table 50.Line output configuration
LOC[1:0]Line output configuration
00Line output fixed - no volume, no EQ
01Line output variable - CH3 volume effects line output, no EQ
10Line output variable with EQ - CH3 volume effects line output
Line output is only active when OCFG = 00. In this case LOC determines the line output
configuration. The source of the line output is always the channel 1 and 2 inputs.
7.7.2 Master volume register
D7D6D5D4D3D2D1D0
MV7MV6MV5MV4MV3MV2MV1MV0
11111111
7.7.3 Channel 1 volume
D7D6D5D4D3D2D1D0
C1V7C1V6C1V5C1V4C1V3C1V2C1V1C1V0
01100000
Doc ID 018572 Rev 351/86
Register descriptionSTA350BW
7.7.4 Channel 2 volume
D7D6D5D4D3D2D1D0
C2V7C2V6C2V5C2V4C2V3C2V2C2V1C2V0
01100000
7.7.5 Channel 3 / line output volume
D7D6D5D4D3D2D1D0
C3V7C3V6C3V5C3V4C3V3C3V2C3V1C3V0
01100000
The volume structure of the STA350BW consists of individual volume registers for each
channel and a master volume register that provides an offset to each channels volume
setting. The individual channel volumes are adjustable in 0.5 dB steps from +48 dB
to -80 dB.
As an example if C3V = 0x00 or +48 dB and MV = 0x18 or -12 dB, then the total gain for
channel 3 = +36 dB.
The master mute, when set to 1, mutes all channels at once, whereas the individual channel
mute (CxM) mutes only that channel. Both the master mute and the channel mutes provide
a “soft mute” with the volume ramping down to mute in 4096 samples from the maximum
volume setting at the internal processing rate (approximately 96 kHz).
A “hard (instantaneous) mute” can be obtained by programming a value of 0xFF (255) to
any channel volume register or the master volume register. When volume offsets are
provided via the master volume register, any channel whose total volume is less than -80 dB
is muted.
All changes in volume take place at zero-crossings when ZCE = 1 (Configuration register E
(addr 0x04)) on a per channel basis as this creates the smoothest possible volume
transitions. When ZCE = 0, volume updates occur immediately.
Table 51.Master volume offset as a function of MV[7:0]
Using AMGC[3:0] bits, attack and release thresholds and rates are automatically configured
to properly fit application specific configurations. AMGC[3:2] is defined in register EQ
coefficients and DRC configuration register (addr 0x31) on page 71.
The AMGC[1:0] bits behave in two different ways depending on the value of AMGC[3:2].
When this value is 00, then bits AMGC[1:0] are defined below in Tabl e 5 3.
Table 53.Audio preset gain compression/limiters selection for AMGC[3:2] = 00
AMGC[1:0]Mode
00User-programmable GC
01AC no clipping 2.1
10AC limited clipping (10%) 2.1
11DRC nighttime listening mode 2.1
Doc ID 018572 Rev 353/86
Register descriptionSTA350BW
7.8.2 Audio preset register 2 (addr 0x0C)
D7D6D5D4D3D2D1D0
XO3XO2XO1XO0AMAM2AMAM1AMAM0AMAME
00000000
7.8.3 AM interference frequency switching
Table 54.AM interference frequency switching bits
BitR/WRSTNameDescription
Audio preset AM enable
0R/W0AMAME
Table 55.Audio preset AM switching frequency selection
0: switching frequency determined by PWMS setting
1: switching frequency determined by AMAM settings
1001.301 MHz - 1.480 MHz1.181 MHz - 1.340 MHz
1011.481 MHz - 1.600 MHz1.341 MHz - 1.500 MHz
1101.601 MHz - 1.700 MHz1.501 MHz - 1.700 MHz
7.8.4 Bass management crossover
Table 56.Bass management crossover
BitR/WRSTNameDescription
4R/W 0XO0
5R/W 0XO1
6R/W 0XO2
7R/W 0XO3
Table 57.Bass management crossover frequency
XO[3:0]Crossover frequency
0000User-defined
000180 Hz
0010100 Hz
Selects the bass-management crossover frequency.
A 1st-order hign-pass filter (channels 1 and 2) or a
2nd-order low-pass filter (channel 3) at the selected
frequency is performed.
0011120 Hz
54/86Doc ID 018572 Rev 3
STA350BWRegister description
Table 57.Bass management crossover frequency (continued)
Tone control (bass/treble) can be bypassed on a per-channel basis for channels 1 and 2.
Table 58.Tone control bypass
CxTCBMode
0Perform tone control on channel x - normal operation
1Bypass tone control on channel x
7.9.2 EQ bypass
EQ control can be bypassed on a per-channel basis for channels 1 and 2. If EQ control is
bypassed on a given channel, the prescale and all filters (high-pass, biquads, de-emphasis,
bass, treble in any combination) are bypassed for that channel.
Doc ID 018572 Rev 355/86
Register descriptionSTA350BW
Table 59.EQ bypass
CxEQBPMode
0Perform EQ on channel x - normal operation
1Bypass EQ on channel x
7.9.3 Volume bypass
Each channel contains an individual channel volume bypass. If a particular channel has
volume bypassed via the CxVBP = 1 register then only the channel volume setting for that
particular channel affects the volume setting, the master volume setting will not affect that
channel.
7.9.4 Binary output enable registers
Each individual channel output can be set to output a binary PWM stream. In this mode
output A of a channel is considered the positive output and output B is the negative inverse.
Table 60.Binary output enable registers
CxBOMode
0FFX 3-state output - normal operation
1Binary output
7.9.5 Limiter select
Limiter selection can be made on a per-channel basis according to the channel limiter select
bits.
.
Table 61.Channel limiter mapping as a function of CxLS bits
CxLS[1:0]Channel limiter mapping
00Channel has limiting disabled
01Channel is mapped to limiter #1
10Channel is mapped to limiter #2
7.9.6 Output mapping
Output mapping can be performed on a per channel basis according to the CxOM channel
output mapping bits. Each input into the output configuration engine can receive data from
any of the three processing channel outputs.
.
Table 62.Channel output mapping as a function of CxOM bits
CxOM[1:0]Channel x output source from
00Channel1
01Channel 2
10Channel 3
56/86Doc ID 018572 Rev 3
STA350BWRegister description
7.10 Tone control register (addr 0x11)
D7D6D5D4D3D2D1D0
TTC3TTC2TTC1TTC0BTC3BTC2BTC1BTC0
01110111
7.10.1 Tone control
Table 63.Tone control boost/cut as a function of BTC and TTC bits
BTC[3:0]/TTC[3:0]Boost/Cut
0000-12 dB
0001-12 dB
……
0111-4 dB
0110-2 dB
01110 dB
1000+2 dB
1001+4 dB
……
1101+12 dB
1110+12 dB
1111+12 dB
7.11 Dynamic control registers (addr 0x12 - 0x15)
7.11.1 Limiter 1 attack/release rate
D7D6D5D4D3D2D1D0
L1A3L1A2L1A1L1A0L1R3L1R2L1R1L1R0
01101010
7.11.2 Limiter 1 attack/release threshold
D7D6D5D4D3D2D1D0
L1AT3L1AT2L1AT1L1AT0L1RT3L1RT2L1RT1L1RT0
01101001
7.11.3 Limiter 2 attack/release rate
D7D6D5D4D3D2D1D0
L2A3L2A2L2A1L2A0L2R3L2R2L2R1L2R0
01101010
Doc ID 018572 Rev 357/86
Register descriptionSTA350BW
7.11.4 Limiter 2 attack/release threshold
D7D6D5D4D3D2D1D0
L2AT3L2AT2L2AT1L2AT0L2RT3L2RT2L2RT1L2RT0
01101001
The STA350BW includes two independent limiter blocks. The purpose of the limiters is to
automatically reduce the dynamic range of a recording to prevent the outputs from clipping
in anti-clipping mode or to actively reduce the dynamic range for a better listening
environment such as a night-time listening mode which is often needed for DVDs. The two
modes are selected via the DRC bit in Configuration register E (addr 0x04) on page 42.
Each channel can be mapped to either limiter or not mapped, meaning that the channel will
clip when 0 dBfs is exceeded. Each limiter looks at the present value of each channel that is
mapped to it, selects the maximum absolute value of all these channels, performs the
limiting algorithm on that value, and then if needed adjusts the gain of the mapped channels
in unison.
The limiter attack thresholds are determined by the LxAT registers if EATHx[7] bits are set
to 0 else the thresholds are determined by EATHx[6:0] . It is recommended in anti-clipping
mode to set this to 0 dBfs, which corresponds to the maximum unclipped output power of an
FFX amplifier. Since gain can be added digitally within the STA350BW, it is possible to
exceed 0 dBfs or any other LxAT setting. When this occurs, the limiter, when active,
automatically starts reducing the gain. The rate at which the gain is reduced when the attack
threshold is exceeded is dependent upon the attack rate register setting for that limiter. Gain
reduction occurs on a peak-detect algorithm. Setting the EATHx[7] bits to 1 selects the
anti-clipping mode.
The limiter release thresholds are determined by the LxRT registers if ERTHx[7] bits are set
to 0, else the thresholds are determined by ERTHx[6:0]. Setting the ERTHx[7] bits to 1
automatically selects the anti-clipping mode. The release of the limiter, when the gain is
again increased, is dependent on an RMS-detect algorithm. The output of the volume/limiter
block is passed through an RMS filter. The output of this filter is compared to the release
threshold, determined by the Release Threshold register. When the RMS filter output falls
below the release threshold, the gain is again increased at a rate dependent upon the
Release Rate register. The gain can never be increased past its set value and, therefore, the
release only occurs if the limiter has already reduced the gain. The release threshold value
can be used to set what is effectively a minimum dynamic range, this is helpful as
overlimiting can reduce the dynamic range to virtually zero and cause program material to
sound “lifeless”.
In AC mode, the attack and release thresholds are set relative to full-scale. In DRC mode,
the attack threshold is set relative to the maximum volume setting of the channels mapped
to that limiter and the release threshold is set relative to the maximum volume setting plus
the attack threshold.
The extended release threshold value is determined as follows:
release threshold = -12 + ERTH2 / 4
Note:Attack/release threshold step is 0.125 dB in the range -12 dB to 0 dB.
7.12 User-defined coefficient control registers (addr 0x16 - 0x26)
7.12.1 Coefficient address register
D7D6D5D4D3D2D1D0
CFA5CFA4CFA3CFA2CFA1CFA0
000000
7.12.2 Coefficient b1 data register bits 23:16
D7D6D5D4D3D2D1D0
C1B23C1B22C1B21C1B20C1B19C1B18C1B17C1B16
00000000
7.12.3 Coefficient b1 data register bits 15:8
D7D6D5D4D3D2D1D0
C1B15C1B14C1B13C1B12C1B11C1B10C1B9C1B8
00000000
7.12.4 Coefficient b1 data register bits 7:0
D7D6D5D4D3D2D1D0
C1B7C1B6C1B5C1B4C1B3C1B2C1B1C1B0
00000000
62/86Doc ID 018572 Rev 3
STA350BWRegister description
7.12.5 Coefficient b2 data register bits 23:16
D7D6D5D4D3D2D1D0
C2B23C2B22C2B21C2B20C2B19C2B18C2B17C2B16
00000000
7.12.6 Coefficient b2 data register bits 15:8
D7D6D5D4D3D2D1D0
C2B15C2B14C2B13C2B12C2B11C2B10C2B9C2B8
00000000
7.12.7 Coefficient b2 data register bits 7:0
D7D6D5D4D3D2D1D0
C2B7C2B6C2B5C2B4C2B3C2B2C2B1C2B0
00000000
7.12.8 Coefficient a1 data register bits 23:16
D7D6D5D4D3D2D1D0
C1B23C1B22C1B21C1B20C1B19C1B18C1B17C1B16
00000000
7.12.9 Coefficient a1 data register bits 15:8
D7D6D5D4D3D2D1D0
C3B15C3B14C3B13C3B12C3B11C3B10C3B9C3B8
00000000
7.12.10 Coefficient a1 data register bits 7:0
D7D6D5D4D3D2D1D0
C3B7C3B6C3B5C3B4C3B3C3B2C3B1C3B0
00000000
7.12.11 Coefficient a2 data register bits 23:16
D7D6D5D4D3D2D1D0
C4B23C4B22C4B21C4B20C4B19C4B18C4B17C4B16
00000000
Doc ID 018572 Rev 363/86
Register descriptionSTA350BW
7.12.12 Coefficient a2 data register bits 15:8
D7D6D5D4D3D2D1D0
C4B15C4B14C4B13C4B12C4B11C4B10C4B9C4B8
00000000
7.12.13 Coefficient a2 data register bits 7:0
D7D6D5D4D3D2D1D0
C4B7C4B6C4B5C4B4C4B3C4B2C4B1C4B0
00000000
7.12.14 Coefficient b0 data register bits 23:16
D7D6D5D4D3D2D1D0
C5B23C5B22C5B21C5B20C5B19C5B18C5B17C5B16
00000000
7.12.15 Coefficient b0 data register bits 15:8
D7D6D5D4D3D2D1D0
C5B15C5B14C5B13C5B12C5B11C5B10C5B9C5B8
00000000
7.12.16 Coefficient b0 data register bits 7:0
D7D6D5D4D3D2D1D0
C5B7C5B6C5B5C5B4C5B3C5B2C5B1C5B0
00000000
7.12.17 Coefficient write/read control register
D7D6D5D4D3D2D1D0
ReservedRAR1WAW1
00000
Coefficients for user-defined EQ, mixing, scaling, and bass management are handled
internally in the STA350BW via RAM. Access to this RAM is available to the user via an I
register interface. A collection of I
coefficient base address, five sets of three store the values of the 24-bit coefficients to be
written or that were read, and one contains bits used to control the write/read of the
coefficient(s) to/from RAM.
2
C registers are dedicated to this function. One contains a
2
C
Three different RAM banks are embedded in the STA350BW. The three banks are managed
in paging mode using EQCFG register bits. They can be used to store different EQ settings.
For speaker frequency compensation, a sampling frequency independent EQ must be
implemented. Computing three different coefficients set for 32 kHz, 44.1kHz, 48 kHz and
downloading them into the three RAM banks, it is possible to select the suitable RAM block
depending on the incoming frequency with a simple I
64/86Doc ID 018572 Rev 3
2
C write operation in register 0x31.
STA350BWRegister description
For example, in case of different input sources (different sampling rates), the three different
sets of coefficients can be downloaded once at startup, and during normal playit is possible
to switch among the three RAM blocks allowing faster operation, without any additional
download from the microcontroller.
To write the coefficients in a particular RAM bank, this bank must be selected first, writing
bit 0 and bit 1 in register 0x31. Then the write procedure below can be used.
Note that as soon as a RAM bank is selected, the EQ settings are automatically switched to
the coefficients stored in the active RAM block.
Note:The read and write operation on RAM coefficients works only if LRCKI (pin 29) is switching.
Reading a coefficient from RAM
1.Select the RAM block with register 0x31 bit 1, bit 0.
2. Write 6 bits of address to I
3. Write 1 to R1 bit in I
4. Read top 8 bits of coefficient in I
5. Read middle 8 bits of coefficient in I
6. Read bottom 8 bits of coefficient in I
2
C register 0x16.
2
C address 0x26.
2
C address 0x17.
2
C address 0x18.
2
C address 0x19.
Reading a set of coefficients from RAM
1.Select the RAM block with register 0x31 bit1, bit0.
2. Write 6 bits of address to I
3. Write 1 to RA bit in I
4. Read top 8 bits of coefficient in I
5. Read middle 8 bits of coefficient in I
6. Read bottom 8 bits of coefficient in I
7. Read top 8 bits of coefficient b2 in I
8. Read middle 8 bits of coefficient b2 in I
9. Read bottom 8 bits of coefficient b2 in I
10. Read top 8 bits of coefficient a1 in I
11. Read middle 8 bits of coefficient a1 in I
12. Read bottom 8 bits of coefficient a1 in I
13. Read top 8 bits of coefficient a2 in I
14. Read middle 8 bits of coefficient a2 in I
15. Read bottom 8 bits of coefficient a2 in I
16. Read top 8 bits of coefficient b0 in I
17. Read middle 8 bits of coefficient b0 in I
18. Read bottom 8 bits of coefficient b0 in I
2
C register 0x16.
2
C address 0x26.
2
C address 0x17.
2
C address 0x18.
2
C address 0x19.
2
C address 0x1A.
2
C address 0x1B.
2
C address 0x1C.
2
C address 0x1D.
2
C address 0x1E.
2
C address 0x1F.
2
C address 0x20.
2
C address 0x21.
2
C address 0x22.
2
C address 0x23.
2
C address 0x24.
2
C address 0x25.
Doc ID 018572 Rev 365/86
Register descriptionSTA350BW
Writing a single coefficient to RAM
1.Select the RAM block with register 0x31 bit1, bit0.
2. Write 6 bits of address to I
3. Write top 8 bits of coefficient in I
4. Write middle 8 bits of coefficient in I
5. Write bottom 8 bits of coefficient in I
6. Write 1 to the W1 bit in I
2
C register 0x16.
2
C address 0x17.
2
C address 0x18.
2
C address 0x19.
2
C address 0x26.
Writing a set of coefficients to RAM
1.Select the RAM block with register 0x31 bit1, bit0.
2. Write 6 bits of starting address to I
3. Write top 8 bits of coefficient b1 in I
4. Write middle 8 bits of coefficient b1 in I
5. Write bottom 8 bits of coefficient b1 in I
6. Write top 8 bits of coefficient b2 in I
7. Write middle 8 bits of coefficient b2 in I
8. Write bottom 8 bits of coefficient b2 in I
9. Write top 8 bits of coefficient a1 in I
10. Write middle 8 bits of coefficient a1 in I
11. Write bottom 8 bits of coefficient a1 in I
12. Write top 8 bits of coefficient a2 in I
13. Write middle 8 bits of coefficient a2 in I
14. Write bottom 8 bits of coefficient a2 in I
15. Write top 8 bits of coefficient b0 in I
16. Write middle 8 bits of coefficient b0 in I
17. Write bottom 8 bits of coefficient b0 in I
18. Write 1 to the WA bit in I
2
C address 0x26.
2
C register 0x16.
2
C address 0x17.
2
C address 0x18.
2
C address 0x19.
2
C address 0x1A.
2
C address 0x1B.
2
C address 0x1C.
2
C address 0x1D.
2
C address 0x1E.
2
C address 0x1F.
2
C address 0x20.
2
C address 0x21.
2
C address 0x22.
2
C address 0x23.
2
C address 0x24.
2
C address 0x25.
The mechanism for writing a set of coefficients to RAM provides a method of updating the
five coefficients corresponding to a given biquad (filter) simultaneously to avoid possible
unpleasant acoustic side-effects. When using this technique, the 6-bit address specifies the
address of the biquad b1 coefficient (for example, 0, 5, 10, 20, 35 decimal), and the
STA350BW generates the RAM addresses as offsets from this base value to write the
complete set of coefficient data.
66/86Doc ID 018572 Rev 3
STA350BWRegister description
7.12.18 User-defined EQ
The STA350BW can be programmed for four EQ filters (biquads) per each of the two input
channels. The biquads use the following equation:
where Y[n] represents the output and X[n] represents the input. Multipliers are 24-bit signed
fractional multipliers, with coefficient values in the range of 0x800000 (-1) to 0x7FFFFF
(0.9999998808).
Coefficients stored in the user-defined coefficient RAM are referenced in the following
manner:
CxHy0 = b
CxHy1 = b
1
2
/ 2
CxHy2 = -a1 / 2
CxHy3 = -a
2
CxHy4 = b0 / 2
where x represents the channel and the y the biquad number. For example, C2H41 is the b
coefficient in the fourth biquad for channel 2.
Additionally, the STA350BW can be programmed for a high-pass filter (processing
channels 1 and 2) and a low-pass filter (processing channel 3) to be used for bassmanagement crossover when the XO setting is 000 (user-defined). Both of these filters
when defined by the user (rather than using the preset crossover filters) are second order
filters that use the biquad equation given above. They are loaded into the C12H0-4 and
C3Hy0-4 areas of RAM noted in Tab le 7 0 .
By default, all user-defined filters are pass-through where all coefficients are set to 0, except
the b
/2 coefficient which is set to 0x400000 (representing 0.5).
0
2
7.12.19 Pre-scale
The STA350BW provides a multiplication for each input channel for the purpose of scaling
the input prior to EQ. This pre-EQ scaling is accomplished by using a 24-bit signed
fractional multiplier, with 0x800000 = -1 and 0x7FFFFF = 0.9999998808. The scale factor
for this multiplication is loaded into RAM using the same I
coefficients and the bass-management. All channels can use the channel-1 pre-scale factor
by setting the Biquad link bit. By default, all pre-scale factors are set to 0x7FFFFF.
7.12.20 Post-scale
The STA350BW provides one additional multiplication after the last interpolation stage and
the distortion compensation on each channel. This post-scaling is accomplished by using a
24-bit signed fractional multiplier, with 0x800000 = -1 and 0x7FFFFF = 0.9999998808. The
scale factor for this multiplication is loaded into RAM using the same I
biquad coefficients and the bass-management. This post-scale factor can be used in
conjunction with an ADC equipped microcontroller to perform power-supply error correction.
All channels can use the channel-1 post-scale factor by setting the post-scale link bit. By
default, all post-scale factors are set to 0x7FFFFF. When line output is being used,
channel-3 post-scale will affect both channels 3 and 4.
2
C registers as the biquad
2
C registers as the
Doc ID 018572 Rev 367/86
Register descriptionSTA350BW
7.12.21 Overcurrent post-scale
The STA350BW provides a simple mechanism for reacting to overcurrent detection in the
power block. When the ocwarn input is asserted, the overcurrent post-scale value is used in
place of the normal post-scale value to provide output attenuation on all channels. The
default setting provides 3 dB of output attenuation when ocwarn is asserted.
The amount of attenuation to be applied in this situation can be adjusted by modifying the
Overcurrent Post-scale value. As with the normal post-scale, this scaling value is a 24-bit
signed fractional multiplier, with 0x800000 = -1 and 0x7FFFFF = 0.9999998808. By default,
the overcurrent post-scale factor is set to
applied, it remains until the device is reset.
Table 70.RAM block for biquads, mixing, scaling and bass management
Index (decimal)Index (hex)CoefficientDefault
0x5A9DF7. Once the overcurrent attenuation is
00x00
10x01C1H11(b2)0x000000
20x02C1H12(a1/2)0x000000
30x03C1H13(a2)0x000000
40x04C1H14(b0/2)0x400000
50x05Channel 1 - Biquad 2C1H200x000000
……………
190x13Channel 1 - Biquad 4C1H440x400000
200x14
210x15C2H110x000000
……………
390x27Channel 2 - Biquad 4C2H440x400000
400x28
410x29C12H1(b2)0x000000
420x2AC12H2(a1/2)0x000000
430x2BC12H3(a2)0x000000
440x2CC12H4(b0/2)0x400000
450x2D
460x2EC3H1(b2)0x000000
470x2FC3H2(a1/2)0x000000
480x30C3H3(a2)0x000000
490x31C3H4(b0/2)0x400000
500x32Channel 1 - Pre-ScaleC1PreS0x7FFFFF
510x33Channel 2 - Pre-ScaleC2PreS0x7FFFFF
Channel 1 - Biquad 1
Channel 2 - Biquad 1
Channel 1/2 - Biquad 5
for XO = 000
nd
Hi-pass 2
Channel 3 - Biquad
Low-pass 2
Order filter
for XO≠000
for XO = 000
nd
Order filter
for XO≠000
C1H10(b1/2)0x000000
C2H100x000000
C12H0(b1/2)0x000000
C3H0(b1/2)0x000000
520x34Channel 1 - Post-ScaleC1PstS0x7FFFFF
530x35Channel 2 - Post-ScaleC2PstS0x7FFFFF
68/86Doc ID 018572 Rev 3
STA350BWRegister description
Table 70.RAM block for biquads, mixing, scaling and bass management (continued)
Index (decimal)Index (hex)CoefficientDefault
540x36Channel 3 - Post-ScaleC3PstS0x7FFFFF
550x37TWARN/OC - LimitTWOCL0x5A9DF7
560x38Channel 1 - Mix 1C1MX10x7FFFFF
570x39Channel 1 - Mix 2C1MX20x000000
580x3AChannel 2 - Mix 1C2MX10x000000
590x3BChannel 2 - Mix 2C2MX20x7FFFFF
600x3CChannel 3 - Mix 1C3MX10x400000
610x3DChannel 3 - Mix 2C3MX20x400000
620x3EUNUSED
630x3FUNUSED
7.13 Variable max power correction registers (addr 0x27 - 0x28)
D7D6D5D4D3D2D1D0
MPCC15MPCC14MPCC13MPCC12MPCC11MPCC10MPCC9MPCC8
D7D6D5D4D3D2D1D0
MPCC7MPCC6MPCC5MPCC4MPCC3MPCC2MPCC1MPCC0
00011010
11000000
The MPCC bits determine the 16 MSBs of the MPC compensation coefficient. This
coefficient is used in place of the default coefficient when MPCV = 1.
The DCC bits determine the 16 MSBs of the distortion compensation coefficient. This
coefficient is used in place of the default coefficient when DCCV = 1.
The FDRC bits specify the 16-bit fault detect recovery time delay. When FAULT is asserted,
the TRISTATE output is immediately asserted low and held low for the time period specified
by this constant. A constant value of 0x0001 in this register is approximately 0.083 ms. The
default value of 0x000C gives approximately 0.1 ms.
7.16 Device status register (addr 0x2D)
D7D6D5D4D3D2D1D0
PLLULFAULTUVFAULTOVFAULTOCFAULTOCWARNTFAULTTWARN
This read-only register provides fault and thermal-warning status information from the power
control block. Logic value 1 for faults or warning means normal state. Logic 0 means a fault
or warning detected on power bridge. The PLLUL = 1 means that the PLL is not locked.
Table 71.Status register bits
BitR/WRSTNameDescription
0:
7R-PLLUL
6R -FAULT
5R - UVFAULT
4R - OVFAULT
PLL locked
1: PLL not locked
0: fault detected on power bridge
1: normal operation
Bits AMGC[3:2] change the behavior of the bits AMGC[1:0] as given in Ta b le 7 3 below.
Table 73.Anti-clipping and DRC preset
AMGC[3:2]Anti clipping and DRC preset selected
00DRC/Anti-clipping behavior described in Ta bl e 5 3 on page 53 (default).
01DRC/Anti-clipping behavior is described Table 74 on page 71
10/11Reserved, do not use
When AMGC[3:2] = 01 then the bits 1:0 are defined as given here in Tabl e 74 .
Table 74.Anti-clipping selection for AMGC[3:2] = 01
AMGC[1:0]Mode
00AC0, stereo anticlipping 0 dB limiter
01AC1, stereo anticlipping +1.25 dB limiter
10AC2, stereo anticlipping +2 dB limiter
11Reserved do not use
The AC0, AC1, AC2 settings are designed for the loudspeaker protection function, limiting at
the minimum any audio artifacts introduced by typical anti-clipping/DRC algorithms. More
detailed information is available in the applications notes “Configurable output power rate
using STA335BW” and “STA335BWS vs STA335BW”.
The XOB bit can be used to bypass the crossover filters. Logic 1 means that the function is
not active. In this case, the high-pass crossover filter works as a pass-through on the data
path (b0 = 1, all the other coefficients at logic 0) while the low-pass filter is configured to
have zero signal on channel-3 data processing (all the coefficients are at logic 0).
Doc ID 018572 Rev 371/86
Register descriptionSTA350BW
7.18 Extended configuration register (addr 0x36)
D7D6D5D4D3D2D1D0
Mdrc[1]Mdrc[0]PS48DBXAR1XAR2BQ5BQ6BQ7
00000000
The extended configuration register provides access to B2DRC and biquad 5, 6 and 7.
7.18.1 Dual-band DRC
The STA350BW device provide a dual-band DRC (B2DRC) on the left and right channels
data path, as depicted in Figure 29. Dual-band DRC is activated by setting MDRC[1:0] = 1x.
R
R
R
R
2
L
L
L
L
DRC scheme
PassXO
PassXO
B2DRC
B2DRC
Filter
Filter
Hi-pass
Hi-pass
filter
filter
PassXO
PassXOB2DRC
B2DRC
Filter
Filter
Hi-pass
Hi-pass
filter
filter
CH1
CH1
Volume
Volume
CH3
CH3
Volume
-
-
-
-
Volume
CH2
CH2
Volume
Volume
CH3
CH3
Volume
Volume
VolAndLimiter
VolAndLimiter
DRC1
DRC1
VolAndLimiter
VolAndLimiter
DRC2
DRC2
DRC1
DRC1
VolAndLimiter
VolAndLimiter
DRC2
DRC2
+
+
+
+
AM045194v1
Figure 29. B
The low frequency information (LFE) is extracted from left and right channels, removing the
high frequencies using a programmable Biquad filter, and then computing the difference with
the original signal. Limiter 1 (DRC1) is then used to control Left/Right high-frequency
amplitude of the components, while limiter 2 (DRC2) is used to control the low-frequency
components (see Chapter 7.11).
The cutoff frequency of the high-pass filters can be user-defined, XO[3:0] = 0, or selected
from the pre-defined values.
DRC1 and DRC2 are then used to independently limit L/R high frequencies and LFE
channels amplitude (see Chapter 7.11) as well as their volume control. To be noted that, in
this configuration, the dedicated channel 3 volume control can actually act as a bass boost
enhancer as well (0.5 dB/step resolution).
The processed LFE channel is then recombined with the L and R channels in order to
reconstruct the 2.0 output signal.
Sub-band decomposition
The sub-band decomposition for B2DRC can be configured specifying the cutoff frequency.
The cutoff frequency can be programmed in two ways, using XO bits in register 0x0C, or
using the “user programmable” mode (coefficients stored in RAM addresses 0x28 to 0x31).
72/86Doc ID 018572 Rev 3
STA350BWRegister description
For the user programmable mode, use the formulas below to compute the high-pass filters:
b0 = (1 + alpha) / 2a0 = 1
b1 = -(1 + alpha) / 2a1 = -alpha
b2 = 0a2 = 0
where alpha = (1-sin(ω0))/cos(ω0), and ω0 is the cutoff frequency.
A first-order filter is suggested to guarantee that for every ω
filter obtained as the difference (as shown in Figure 29) will have a symmetric (relative to HP
filter) frequency response, and the corresponding recombination after the DRC has low
ripple. Second-order filters can be used as well, but in this case the filter shape must be
carefully chosen to provide good low-pass response and minimum ripple recombination. For
second-order filters it is not possible to give a closed formula to get the best coefficients, but
empirical adjustment should be done.
DRC settings
The DRC blocks used by B2DRC are the same as those described in Chapter 7.11. B2DRC
configure automatically the DRC blocks in anticlipping mode. Attack and release thresholds
can be selected using registers 0x32, 0x33, 0x34, 0x35, while attack and release rates are
configured by registers 0x12 and 0x14.
Band downmixing
The low-frequency band is down-mixed to the left and right channels at the B2DRC output.
Channel volume can be used to weight the bands recombination to fine-tune the overall
frequency response.
7.18.2 EQ DRC mode
Setting MDRC = 01, it is possible to add a programmable biquad (the XO biquad at RAM
addresses 0x28 to 0x2C is used for this purpose) to the Limiter/compressor measure path
(side chain). Using EQDRC the peak detector input can be shaped in frequency using the
programmable biquad. For example if a +2 dB bass boost is applied (using a low shelf filter
for example), the effect is that the EQDRC output will limit bass frequencies to -2 dB below
the selected attack threshold.
the corresponding low-pass
0
Generally speaking, if the biquad boosts frequency f with an amount of X dB, the level of a
compressed sine wave at the output will be TH - X, where TH is the selected attack
threshold.
Note:EQDRC works only if the biquad frequency response magnitude is ≥ 0dB for every
frequency.
Doc ID 018572 Rev 373/86
Register descriptionSTA350BW
Figure 30. EQDRC scheme
EQDRC
EQDRC
ATTENUATION
ATTENUATION
ATTENUATION
Channel In
Channel In
Channel In
Channel In
Channel In
Channel In
Channel In
Channel In
BIQUAD
BIQUADBIQUAD
BIQUAD
BIQUADBIQUAD
PEAK
PEAK
DETECTOR
DETECTOR
Standard DRC
Standard DRC
PEAK
PEAK
DETECTOR
DETECTOR
ATTENUATION
CLACULATOR
CLACULATOR
ATTENUATION
ATTENUATION
CLACULATOR
CLACULATOR
ATTENUATION
ATTENUATION
AM045195v1
7.18.3 Extended post-scale range
PS48DBMode
0Post-scale value is applied as defined in coefficient RAM
1
Post-scale is an attenuation by default. When PS48DB is set to 1, a 48-dB offset is applied
to the coefficient RAM value, so post-scale can act as a gain too.
7.18.4 Extended attack rate
The attack rate shown in Ta bl e 6 4 can be extended to provide up to an 8 dB/ms attack rate
on both limiters.
XAR1Mode
0Limiter1 attack rate is configured using Ta b le 6 4
1Limiter1 attack rate is 8 dB/ms
XAR2Mode
0Limiter2 attack rate is configured using Ta b le 6 4
1Limiter2 attack rate is 8 dB/ms
Post-scale value is applied with +48 dB offset with respect to the
coefficient RAM value
74/86Doc ID 018572 Rev 3
STA350BWRegister description
7.18.5 Extended BIQUAD selector
De-emphasis filter as well as bass and treble controls can be configured as user-defined
filters when the equalization coefficients link is activated (BQL = 1) and the corresponding
BQx bit is set to 1.
BQ5Mode
0Pre-set de-emphasis filter selected
1User-defined biquad 5 coefficients are selected
BQ6Mode
0Pre-set bass filter selected as per Ta bl e 6 3
1User-defined biquad 6 coefficients are selected
BQ7Mode
0Pre-set treble filter selected as per Ta bl e 6 3
1User-defined biquad 7 coefficients are selected
When filters from 5th to 7th are configured as user-programmable, the corresponding
coefficients are stored respectively in addresses 0x20-0x24 (BQ5), 0x25-0x29 (BQ6), 0x2A0x2E (BQ7) as in Ta bl e 7 0.
Note:BQx bits are ignored if BQL = 0 or if DEMP = 1 (relevant for BQ5) or CxTCB = 1 (relevant for
The soft-volume update has a fixed rate by default. Using register 0x37 and 0x38 it is
possible to override the default behavior allowing different volume change rates.
It is also possible to independently define the fade-in (volume is increased) and fade-out
(volume is decreased) rates according to the desired behavior.
SVUPEMode
0When volume is increased, use the default rate
1When volume is increased, use the rates defined by SVUP[4:0]
Doc ID 018572 Rev 375/86
Register descriptionSTA350BW
When SVUPE = 1 the fade-in rate is defined by the SVUP[4:0] bits according to the following
formula:
Fade-in rate = 48 / (N + 1) dB/ms
where N is the SVUP[4:0] value.
SVDWEMode
0When volume is decreased, use the default rate
1When volume is decreased, use the rates defined by SVDW[4:0]
When SVDWE = 1 the fade-out rate is defined by the SVDW[4:0] bits according to the
following formula:
Fade-in rate = 48 / (N + 1) dB/ms
where N is the SVDW[4:0] value.
Note:For fade-out rates greater than 6 dB/msec it is suggested to disable the CPWMEN bit
(Section 7.24.4 ) and ZCE bit (Section 7.5.7) in order to avoid any audible pop noise.
The extra volume resolution allows fine volume tuning by steps of 0.125dB.
The feature is enabled when VRESEN=1 , as depicted in Figure 31. The overall channel
volume in this case will be CxVol+CxVR (in dB). On top of the total volume range from
-80 dB to +48 dB, this extra volume resolution works in a volume range from -80 dB to +42
dB. For volumes greater than +42 dB, this function must not be selected.
Figure 31. Extra resolution volume scheme
Audio Data In
Audio Data In
CxVOL
CxVOL
Soft
Soft
Volume
Volume
0
0
Audio Data Out
Audio Data Out
X
X
1
1
VRESTG
VRESTG
VRESEN
VRESEN
MVOL or CxVOL’event
MVOL or CxVOL’event
CxVR
CxVR
X
X
01
01
If VRESEN = 0 the channel volume will be defined only by the CxVol registers.
Fine-tuning steps can be set according to the following table for channels 1, 2, 3:
CxVRMode
00 0 dB
01-0.125dB
10-0.25dB
11-0.375dB
AM045196v1
Doc ID 018572 Rev 377/86
Register descriptionSTA350BW
Two different behaviors can be configured by the VRESTG bit:
●If VRESTG=’0’ the CxVR contribution will be applied immediately after the
corresponding I
●If VRESTG=’1’ the CxVR bits will be effective on channel volume only after the
2
C bits are written.
corresponding CxVol register or master volume register is written (even to the previous
values).
VRESENVRESTGMode
00Extra Volume Resolution disabled
01Extra Volume Resolution disabled
10Volume fine-tuning enabled and applied immediately.
11
Volume fine-tuning enabled and applied when master or
channel volume is updated
A special feature inside the digital processing block is available. In case of poles positioned
at very low frequencies, biquad filters can generate some audible quantization noise or
unwanted DC level. In order to avoid this kind of effect, a quantization noise-shaping
capability can be used. The filter structure including this special feature, relative to each
biquad, is shown in Figure 32.
By default, this capability is not activated to maintain backward compatibility with all the
previous Sound Terminal products. The new feature can be enabled independently for each
biquad using the I
2
C registers. The D7 bit, when set, is responsible for activating this
function on the crossover filter while the other bits address any specific biquads according to
the previous table. Channels 1 and 2 share the same settings. Bit D7 is effective also for
channel 3 if the relative OCFG is used.
78/86Doc ID 018572 Rev 3
STA350BWRegister description
Figure 32. Biquad filter structure with quantization error noise-shaping
-1
In(t)
In(t)
-1
z
z
b
b
0
0
+
+
-
-
Q
Q
Out(t)
Out(t)
-1
-1
z
z
b
b
1
1
-1
-1
z
z
b
b
2
2
a
a
1
1
a
a
2
2
-1
-1
z
z
-1
-1
z
z
AM045197v1
7.23 Extended coefficient range up to -4...4 (address 0x49, 0x4A)
Biquads from 1 to 7 have in the STA350BW the possibility to extend the coefficient range
from [-1,1) to [-4..4). This allows the realization of high shelf filters that may require a
coefficients dynamic greater in absolute value than 1.
Three ranges are available, [-1;1) [-2;2) [-4;4). By default the extended range is not activated
to maintain backward compatibility with all the previous Sound Terminal products.
Each biquad has its independent setting according to the following table:
Table 75.Biquad filter settings
CEXT_Bx[1]CEXT_Bx[0]
00
01
10
11Reserved
Doc ID 018572 Rev 379/86
[-1;1)
[-2;2)
[-4;4)
Register descriptionSTA350BW
In this case the user can decide, for each filter stage, the right coefficients range. Note that
for a given biquad the same range will be applied to Left and Right (Channel 1 and Channel
2).
The crossover biquad does not have the availability of this feature, maintaining the [-1;1)
range unchanged.
7.24.1 Rate powerdown enable (RPDNEN) bit (address 0x4B, bit D7)
In the STA350BW, by default, the power-down pin and I2C power-down act on mute
commands to perform the fadeout. This default can be changed so that the fadeout can be
started using the master volume. The RPDNEN bit, when set, activates this feature.
7.24.2 Noise-shaping on DC cut filter enable (NSHHPEN) bit (address 0x4B,
bit D6)
Following the description in Section 7.22, this bit, when set, enables the noise-shaping
technique on the DC cutoff filter. Channels 1 and 2 share the same settings.
7.24.3 Bridge immediate off (BRIDGOFF) bit (address 0x4B, bit D5)
A fadeout procedure is started in the STA350BW once the PWDN function is enabled.
Independently from the fadeout time, after 13 million clock cycles (PLL internal frequency)
the bridge is put in powerdown (tristate mode). There is also the possibility to change this
behavior so that the power bridge will be switched off immediately after the PWDN pin is tied
to ground, without therefore waiting for the 13 million clock cycles. The BRIDGOFF bit, when
set, activates this function. Obviously the immediate power-down will generate a pop noise
at the output. therefore this procedure must be used only in case pop noise is not relevant in
the application. Note that this feature works only for hardware PWDN assertion and not for a
powerdown applied through the I
different number of clock cycles.
2
C interface. Refer to Section 7.24.5 in order to program a
80/86Doc ID 018572 Rev 3
STA350BWRegister description
7.24.4 Channel PWM enable (CPWMEN) bit (address 0x4B, bit D2)
This bit, when set, activates a mute output in case the volume will reach a value lower than
As per Section 7.24.3, the assertion of PWDN activates a counter that, by default, after 13
million clock cycles puts the power bridge in tristate mode, independently from the fadeout
time. Using these registers it is possible to program this counter according to the following
table:
PNDLSL[2]PNDLSL[1]PNDLSL[2]Fade out time
00
00
01
01
10
10
11
11 1
0Default time (13M clock cycles)
1Default time divided by 2
0Default time divided by 4
1Default time divided by 8
0Default time divided by 16
1Default time divided by 32
0Default time divided by 64
Default time divided by 128
Doc ID 018572 Rev 381/86
Package thermal characteristicsSTA350BW
8 Package thermal characteristics
Using a four-layer PCB the thermal resistance junction-to-ambient with 2 copper ground
areas of 6 x 4 cm
2
and with 24 via holes (see Figure 33) is 17 °C/W in natural air convection.
The dissipated power within the device depends primarily on the supply voltage, load
impedance and output modulation level.
Thus, the maximum estimated dissipated power for the STA350BW is:
2 x 40 W @ 8 Ω, 25.5 VPd max ~ 8 W
2 x 17 W + 1 x 35 W @ 4 Ω, 8 Ω, 25 VPd max < 7 W
Figure 33. Double-layer PCB with 2 copper ground areas and 24 via holes
AM045200v1
Figure 34 shows the power derating curve for the PowerSSO-36 slug-down package on
PCBs with copper areas of 5 x 4 cm
2
and 6 x 4cm2.
Figure 34. PowerSSO-36 power derating curve
Pd (W)
8
7
Copper Area 6x4 cm
Copper Area 6x4 cm
6
5
4
3
Copper Area 5x4 cm
Copper Area 5x4 cm
and 20 via holes
2
1
0
020406080100120140160
and 20 via holes
and 24via holes
and 24 via holes
Tamb ( °C)
STA350BW
STA350BW
Power-SSO36
Power-SSO36
AM045201v1
82/86Doc ID 018572 Rev 3
STA350BWPackage mechanical data
9 Package mechanical data
In order to meet environmental requirements, ST offers these devices in different grades of
ECOPACK
specifications, grade definitions and product status are available at: www.st.com.
ECOPACK
®
packages, depending on their level of environmental compliance. ECOPACK®
Updated min. and typ. values for Isc in Table 6: Electrical
specifications - power section
Doc ID 018572 Rev 385/86
STA350BW
Please Read Carefully:
Information in this document is provided solely in connection with ST products. STMicroelectronics NV and its subsidiaries (“ST”) reserve the
right to make changes, corrections, modifications or improvements, to this document, and the products and services described herein at any
time, without notice.
All ST products are sold pursuant to ST’s terms and conditions of sale.
Purchasers are solely responsible for the choice, selection and use of the ST products and services described herein, and ST assumes no
liability whatsoever relating to the choice, selection or use of the ST products and services described herein.
No license, express or implied, by estoppel or otherwise, to any intellectual property rights is granted under this document. If any part of this
document refers to any third party products or services it shall not be deemed a license grant by ST for the use of such third party products
or services, or any intellectual property contained therein or considered as a warranty covering the use in any manner whatsoever of such
third party products or services or any intellectual property contained therein.
UNLESS OTHERWISE SET FORTH IN ST’S TERMS AND CONDITIONS OF SALE ST DISCLAIMS ANY EXPRESS OR IMPLIED
WARRANTY WITH RESPECT TO THE USE AND/OR SALE OF ST PRODUCTS INCLUDING WITHOUT LIMITATION IMPLIED
WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE (AND THEIR EQUIVALENTS UNDER THE LAWS
OF ANY JURISDICTION), OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT.
UNLESS EXPRESSLY APPROVED IN WRITING BY TWO AUTHORIZED ST REPRESENTATIVES, ST PRODUCTS ARE NOT
RECOMMENDED, AUTHORIZED OR WARRANTED FOR USE IN MILITARY, AIR CRAFT, SPACE, LIFE SAVING, OR LIFE SUSTAINING
APPLICATIONS, NOR IN PRODUCTS OR SYSTEMS WHERE FAILURE OR MALFUNCTION MAY RESULT IN PERSONAL INJURY,
DEATH, OR SEVERE PROPERTY OR ENVIRONMENTAL DAMAGE. ST PRODUCTS WHICH ARE NOT SPECIFIED AS "AUTOMOTIVE
GRADE" MAY ONLY BE USED IN AUTOMOTIVE APPLICATIONS AT USER’S OWN RISK.
Resale of ST products with provisions different from the statements and/or technical features set forth in this document shall immediately void
any warranty granted by ST for the ST product or service described herein and shall not create or extend in any manner whatsoever, any
liability of ST.
ST and the ST logo are trademarks or registered trademarks of ST in various countries.
Information in this document supersedes and replaces all information previously supplied.
The ST logo is a registered trademark of STMicroelectronics. All other names are the property of their respective owners.