STMICROELECTRONICS STV0196B

STV0196B
QPSK/BPSK DEMODULATOR AND FEC IC
..
.
.
.
FRONT-END INTERFACE
I AND Q 6 BITS DIGITAL INPUTS AT 2Fs
QPSK DEMODULATION (Two Modes : A and B)
INPUT SYMBOL FREQUENCY (Fs) UP TO
30MSYMBOLS/S
DIGITAL NYQUIST ROOT FILTER :
ROLL-OFF VALUE OF 0.35 IN MODE A
DIGITAL CARRIER LOOP :
- ON-CHIP DEROTATOR AND TRACKING
LOOP
- CARRIER OFFSET INDICATOR
- LOCK DETECTOR
- C/N INDICATOR FOR DISH POSITIONING
.
.
.
DIGITAL TIMING RECOVERY :
- INTERNAL TIMING ERROR EVALUATION
AND FILTER
- OUTPUT CONTROL SIGNAL FOR A 2Fs
EXTERNAL VCO OR VCXO
DIGITAL AGC :
- INTERNAL SIGNAL POWER ESTIMATION
AND FILTER
- OUTPUT CONTROL SIGNAL FOR AGC
(1 BIT PULSE DENSITY MODULATION)
DESCRIPTION
Designed for the fast growing direct broadcast
satellite (DBS) digital TV receiver market,
the SGS-THOMSON STV0196B Digital Satellite
Receiver Front-end integrates all the functions
needed to demodulate incoming digital satellite TV
signals from the tuner : Nyquist filters, QPSK/BPSK
demodulator, signal power estimator, automatic
gain control, Viterbi decoder, deinterleaver, ReedSolomon decoder and energy dispersal descrambler. This high level of integration greatly reduces
the package count and cost of a set top box. The
demodulator blocks are suitable for a wide range
of symbol rates while the advanced error correction
functions guaranteea low error rate even with small
receiver antennas or low power transmitters.
The STV0196B has multistandard capability.
It is fully compliant with the recently defined Digital
Video Broadcasting (DVB) standard (already
adopted by satellite TV operators in the USA,
Europe and Asia) and also compatible with the
main consumer digital satellite TV standards in use.
FORWARD ERROR CORRECTION
INNER DECODER :
- VITERBI SOFT DECODER FOR CONVOLUTIONAL CODES, CONSTRAINT LENGTH
M = 7, RATE 1/2
- PUNCTURED CODES 1/2, 2/3, 3/4, 5/6 AND
7/8 IN MODE A
- AUTOMATIC OR MANUAL RATE AND
PHASE RECOGNITION
.
.
.
DEINTERLEAVER :
- WORD SYNCHRO EXTRACTION
- CONVOLUTIVE DEINTERLEAVER
OUTER DECODER :
- IN MODE A : REED-SOLOMON DECODER
FOR 16 PARITY BYTES ; CORRECTION OF
UP TO 8 BYTE ERRORS
- BLOCK LENGTHS : 204 IN MODE A
- ENERGY DISPERSAL DESCRAMBLER
PQFP64
(Plastic Package)
ORDER CODE : STV0196B
CONTROL
I2C SERIAL BUS
September 1996
1/23
STV0196B
VSS
VDD
I5
I4
I3
I2
I1
I0
Q5
Q4
Q3
Q2
Q1
Q0
TEST
TEST
PIN CONNECTIONS
64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49
TEST
1
48
M_CLK
TEST
2
47
MODE
VS S
3
46
CLKREC
VDD
4
45
VDD
TEST
5
44
AGC
TEST
6
43
VDD
VS S
7
42
VS S
VDD
8
41
VS S
VS S
9
40
SDA
VDD
10
39
SCL
VS S
11
38
VDD
VDD
12
37
VS S
TEST
13
36
NRES
TEST
14
35
D60
TEST
15
34
ERROR
TEST
16
33
D/P
2/23
0196B-01.EPS
VSS
VDD
STR_OUT
CK_OUT
VSS
VDD
D7
D6
D5
D4
D3
D2
D1
D0
TEST
TEST
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
STV0196B
PIN LIST
Pin Number
Pin Name
Type
Pin Description
I [5..0]
Q [5..0]
M_CLK
I
I
I
In Phase Component, at twice the symbol frequency (2Fs).
In Quadrature Component, at twice the symbol frequency (2Fs).
Master Clock Input, 2Fs. Sampling Clock of the External A to D Converters.
CLKREC
AGC
D60
O
O
O
1 Bit Control Signal for the External CLK VCO. It must be Low-pass Filtered.
1 Bit Control Signal for the External AGC. It must be Low-pass Filtered.
M_CLK Divided by 60
D [7..0]
O
Output Data
CK_OUT
STR_OUT
D/P
ERROR
O
O
O
O
Output Byte Clock
Output Synchronization Byte Signal
Data/Parity Signal
Output Error Signal. Set in Case of uncorrected Block.
SIGNAL INPUTS
51, 52, 53, 54, 55, 56
57, 58, 59, 60, 61, 62
48
FRONT END CONTROLS
46
44
35
SIGNAL OUTPUTS
26, 25, 24, 23,
22, 21, 20, 19
29
30
33
34
2
I C MICRO INTERFACE
39
40
SCL
SDA
I
I/O
Serial Clock
Serial Data Bus
47
1, 2, 5, 6, 13, 14, 15,
16, 17, 18, 63, 64
3, 7, 9, 11, 28, 32,
37, 41, 42, 49
4, 8, 10, 12, 27,
31, 38, 43, 45, 50
36
MODE
TEST
I
O
0 = Mode A, 1 = Mode B
Reserved for Manufacturing Test. It must remain unconnected
VSS
I
Ground References
VDD
I
3.3V Supply
NRES
I
Negative Reset
0196B-01.TBL
OTHER
BLOCK DIAGRAM
I[5...0]
Q[5...0]
CLKREC
D60
NYQUIST
FILTER
DEROTATOR
AGC
TIMING
RECOVERY
DCO
LOCK
INDICATOR
CARRIER
OFFSET
MEASURE
CARRIER PHASE
TRACKING LOOP
C/N
INDICATOR
DIVIDE BY 60
VITERBI DECODER
I2C BUS
INTERFACE
DEINTERLEAVER
AGC
M_CLK
SCL
SDA
D/P
REED SOLOMON DECODER
ERROR
STR_OUT
CK_OUT
0196B-02.EPS
ENERGY DESCRAMBLER
MODE
STV0196B
VDD
VSS
D[7..0]
3/23
STV0196B
FUNCTIONAL DESCRIPTION
I - I2C BUS SPECIFICATION
This is the standard I2C protocol.
The device address is ”1101000” ; the first byte is therefore Hex D0 for a write operation and Hex D1 for a
read operation.
I.1 The
The
The
The
The
The
Write Operation
first byte is the device address plus the direction bit (R/W = 0).
second byte contains the internal address of the first register to be accessed.
next byte is written in the internal register.
following (if any) bytes are written in successive internal registers.
transfer lasts until stop conditions are encountered.
STV0196B acknowledges every byte transfer.
I.2 - Read Operation
The address of the first register to read is programmed in a write operation without data, and terminated
by stop condition.
Then another start is followed by the device address and R/W = 1 ; all successive bytes are now data read
at successive positions starting from the initial address.
The STV0196B acknowledges every byte transfer.
Example :
Write registers 0 to 3 with AA,BB,CC,DD
Device Address,
Write D0
Start
Internal
Address
ACK
ACK
Data
AA
ACK
Data
BB
ACK
Data
CC
ACK
Stop
ACK
Stop
ACK
Stop
Read registers 2 and 3
Device Address,
Write D0
Start
Device Address,
Read D1
Start
ACK
ACK
Register Address 01
Data Read
BB
Data Read
CC
ACK
I.3 - Identification Register
This read only register gives the release number of the circuit in order to ensure software compatibility.
The read value is Hex 83 for STV0196B and Hex 81 for STV0196.
Internal Address : Hex 0B
1
0
0
0
Notes : - Unspecified register addresses must not be used.
- All the unused bits in the registers must be programmed to 0.
4/23
0
0
1
1
STV0196B
FUNCTIONAL DESCRIPTION (continued)
I.4 - Register Map
REGISTER HEX 00
INPUT CONFIGURATION REGISTER (R/W)
Reset Value : Hex 04
0 -Q(1) or Q(0) input
1 Signed (1) or positive (0) I & Q inputs
2 Nyquist filtering on (1) / off (0)
3 BPSK (1), QPSK(0)
4 To be set to 0.
5 To be set to 0.
6 To be set to 0.
7 To be set to 0.
REGISTER HEX 08
VSTATUS REGISTER (Read only)
0
PR[2..0] Current puncture rate identification
1
2
3
LK
(1) synchro found,
(0) searching puncture rate
4
PRF
(1) puncture rate found,
(0) searching puncture rate
5
unused set to (0)
6
unused set to (0)
7
CF
(1) carrier found, (0) searching carrier
REGISTERS HEX 01 TO HEX 05
VITERBI, PUNCTURE RATE THRESHOLDS (R/W)
Reset Value : Hex 20
rate
Hex01 VTH0 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0 1/2
Hex02 VTH1 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0 2/3
Hex03 VTH2 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0 3/4
Hex04 VTH3 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0 5/6
7/8
Hex05 VTH4 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0 or
6/7
REGISTER HEX 06
VSEARCH (VITERBI) (R/W)
Reset Value : Hex 19
0
H[1..0] Sync counter hysteresis value
1
2
T[1..0] Sync search time out
3
4
VITERBI error rate averaging period.
SN[1..0]
C/N indicator averaging period.
5
6
F
VITERBI operating status freeze (1)
7
A/M
(0) automatic, (1) manual
REGISTER HEX 07
VERROR REGISTER (Read only)
REGISTER HEX 09
PUNCTURE RATE ENABLE (R/W)
Reset Value : Hex 10 (mode A)
0
E0 (1) Puncture 1/2 enabled, (0) disabled
1
E1 (1) Puncture 2/3 enabled, (0) disabled
2
E2 (1) Puncture 3/4 enabled, (0) disabled
3
E3 (1) Puncture 5/6 enabled, (0) disabled
4
E4 (1) Puncture7/8 (mode A), 6/7 (mode B)
(0) disabled
5
unused
6
7
REGISTER HEX 0A
RS REGISTER (R/W)
Reset Value : Hex B8
0 RS0 (1) output clock stopped during parity,
(0) continuous
1 RS1 Output clock polarity
2 RS2 (1) all synchro words are Hex47,
(0) synchro inversion disabled
3 RS3 Write error bit
4 RS4 Descrambler on (1), off (0)
5 RS5 Reed-Solomon on (1), off (0)
6 RS6 Normal operation (0), Reed-Solomon
correction bytes to output (1)
7 RS7 De-interleaver on (1), off (0)
5/23
STV0196B
FUNCTIONAL DESCRIPTION (continued)
I.4 - Register Map (continued)
REGISTER HEX 0B
IDENTIFICATION REGISTER (Read only)
Reset Value : Hex83 for STV0196B,
Hex 81 for STV0196
REGISTER HEX 0C
TIMING LOOP : TIME CONSTANT (R/W)
Reset Value : Hex 45
0
1
beta_tmg coefficient
2
3
4
alpha_tmg coefficient
5
6
7
Istr
external VCO/VCXO slope polarity (0)
positive, (1) negative
REGISTER HEX 10
CARRIER OFFSET EVALUATOR (Read only)
Signed value ranging from 80 to 7F.
REGISTER HEX 11
AGC CONTROL REGISTER (R/W)
Reset Value : 18 Hex.
0
1
2
AGC reference level m
3
4
5
6
unused
7
Iagc
REGISTER HEX 0D
TIMING FREQUENCY REGISTER (R/W)
Signed value ranging from 80 to 7F.
REGISTER HEX 12
AGC INTEGRATOR (R/W)
Signed value ranging from 80 to 7F.
REGISTER HEX 0E
CARRIER LOOP REGISTER (R/W)
Reset Value : Hex A3
0
beta_carrier coefficient
1
2
3
unused
4
alpha_carrier coefficient
5
6
7
Deratator on (1), off (0)
REGISTER HEX 13
AGC COEFFICIENT
0
G[2..0] AGC coefficient
1
2
3
4
unused
5
6
7
REGISTER HEX 0F
DEROTATOR FREQUENCY REGISTER (R/W)
Signed value ranging from 80 to 7F.
REGISTER HEX 14
C/N INDICATOR (Read only)
Value ranging from 00 to FF.
6/23
STV0196B
FUNCTIONAL DESCRIPTION (continued)
II - ADC INTERFACE
II.1 - M_CLK Master Clock Input
IV - TIMING RECOVERY
This is the highest frequency clock of the chip, at
twice the symbol frequency; all other clocks are
derived from it.
This clock should be output from an external VCO
or VCXO, controlled by CLKREC output.
M_CLK divided by 60 is available to the system
(output D60).
II.2 - I and Q Signal Inputs
Those signals are coded on 6 bits, either in 2’s
complement or as positive values : the choice is
programmable via the Input Configuration register.
The π/2 ambiguity inherent in QPSK is solved in the
Error Correction part.
A programmable bit in a mode register allows to
multiply by -1 the data on Q input, in order to
accommodate QPSK modulation with another convention of rotation sense ; (this is equivalent to a
permutation of I and Q inputs, or a spectral symmetry).
The timing loop comprises an external VCO
or VCXO, running at twice the symbol frequency,
controlled by the output CLKREC ; this signal is a
pulse density modulated output, at the symbol
frequency, and represents the filtered timing
error.
The loop is parametrised by two coefficients : alpha_tmg and beta_tmg ; the 12 bit filter output is
converted into a pulse density modulation signal
which should be filtered by an analog low pass filter
before commanding the VCO.
IV.1 - Timing Loop Registers
Time Constant Register
Internal Address : Hex0C
Reset Value : Hex45
Istr
Invert
bit
III - NYQUIST ROOT FILTER
0
0
1
0
0
-Q(1) or Q(0) input
0
Signed (1) or positive (0)
I&Q Inputs
0
Nyquist filtering
on (1)/off (0)
0
BPSK(1), QPSK(0)
Input Configuration Register
(the written value of each bit is the reset value)
Internal Address : Hex00
0
0
alpha_tmg (1 to 6)
0
1
0
1
beta_tmg (0 to 9)
The bit ”Istr” allows to change the polarity of the
output signal, in order to accommodate both possibilities of external VCO :
Istr
0
The I and Q components are filtered by a digital
Nyquist root filter with the following features :
- Input : separate I and Q streams, two samples per
symbol.
- Excess bandwidth : 0.35 in Mode A.
- The filters may be bypassed ; in this case, the
input flow is connected to the carrier and clock
recovery section.
1
1
Loop Control
VCO frequency raises when output average
voltage raises
VCO frequency decreases when output
average voltage raises
Timing Frequency Register
Internal Address : Hex0D
Signed number
The value of this register, when the system is
locked, is an image of the frequencyoffset; it should
be as close as possible to 0 in order to have a
symmetric capture range ; reading it allows optimal
trimming of the timing VCO range.
IV.2 - Loop Equations
The external VCO is controlled by the output
CLKREC followed by a low pass filter.
The full analog swing of the output originates a
relative frequency shift of 2∆f , depending on the
characteristics of the external VCO (typically a
fraction of percent).
The frequency range is therefore f = f0 ( 1 ± ∆f).
Neglecting the analog low pass filter on the pulse
modulated output, this loop may be considered as
a second order loop.
7/23
STV0196B
FUNCTIONAL DESCRIPTION (continued)
The natural frequency and the damping factor may be calculated by the following formulas :
Fs
ωn
=

√
β K0 Kd
fn =
2π
2π
where β is programmed by the timing register
: β = 2beta_tmg.
∆f
.
226
Kd is the phase detector ; its value depends on : Kd = 0.977m2 (in Mode A),
the roll-off value and on the power of the signal.
or Kd = 0.564m2 (in Mode B).
where m is the programmed reference level
(see AGC part), reset value : m = 24
Fs is the symbol frequency, ∆f is the half range of the VCO
K0 is the constant of the VCO
: K0 =
Therefore fn = 19.2 10−6 ⋅ m ⋅ Fs ⋅ √

∆f 2beta_tmg
 (Mode A)
−6
beta_tmg
or
fn = 14.6 10 ⋅ m ⋅ Fs ⋅ √

∆f 2
 (Mode B)
The damping factor is : ξ =
α
2
K0Kd
√

β with α = 2 alpha_tmg + 12
0.247 ⋅ m ⋅ √

∆f ⋅ 2 alpha_tmg
0.188 ⋅ m ⋅ √
∆f ⋅ 2alpha_tmg
(Mode
A)
or
ξ
=
(Mode B).

√
2 beta_tmg

√
2beta_tmg
beta_tmg can only take value from 0 to 9 ; if beta_tmg = 0, the loop becomes a first order one.
alpha_tmg can take any value from 1 to 6 ; if both alpha_tmg and beta_tmg are null, the loop is open ; the
duty cycle of the CLKREC output is controlled by writting the timing frequency register.
The next curve shows the natural frequency for a symbol frequency of 20Mbd, in Mode A, with nominal
reference level m = 24 as a function of the VCO relative frequency half range ∆f, for different values of the
register value beta_tmg.
The followingchart gives the value of the damping factor as a function of the VCO relative range, for different
combinations of alpha_tmg and beta_tmg, noticing that the damping factor only depends on the value of
α
or (2 . alpha_tmg - beta_tmg ).
β
√
or ξ =
8/23
STV0196B
FUNCTIONAL DESCRIPTION (continued)
Figure 1 : Natural Frequency for Fs = 20MBauds
100
be ta_tmg
9
8
7
6
5
4
3
2
1
10
1
0.1
0.0001
0.001
0196B-03.EPS
NATURAL FREQUENCY (kHz)
res e t value
0.01
VCO Re lative Fre quency Ra nge (∆f)
Figure 2 : Damping Factor
10
8
7
re se t value
6
5
4
KSI
3
2
1
1
0
2alpha_tmg
- beta _tmg
0.0001
0.001
0196B-04.EPS
0.1
0.01
VCO Relative Frequency Ra nge (∆f)
Example :
the VCO is trimmed from 39.9MHz to 40.1MHz when the VCO control output CLKREC goes from duty
cycle 0 to 100%. The peak-to-peak relative range is therefore 0.5% and ∆f = 0.0025 ; the reset values of
the parameters (alpha_tmg = 4, beta_tmg = 5) leads to a natural frequency of 2.6kHz, with a damping
factor of 0.84.
9/23
STV0196B
FUNCTIONAL DESCRIPTION (continued)
The shaded area correspond to the reset values.
V - CARRIER RECOVERY ; DEROTATOR
The input of the circuit is a pair of demodulated
signals ; however, there may subsist some phase
error not corrected by the front end loop.
Furthermore, the demodulation may be done at
constant frequency; the tuner is trimmed in order
to make the useful signal bandwidth centered on
this demodulation frequency ; in that case, a carrier
offset frequencymay subsist; it is fixed by the mean
of the on-chip derotator which acts as a fine tuning
carrier loop.
The derotator frequency range is limited to an
interval corresponding to ±Fs/16.
V.1 - Loop Parameters
Like the timing loop, the carrier loop is a second
order system where two parameters α and β may
be programmed respectively with alpha_car and
beta_car.
Carrier Loop Parameter Registers
Internal Address : Hex0E
Derotator
ON/OFF
1
0
1
0
0
alpha_carrier
0
1
1
beta_carrier
Derotator Frequency Register
Internal Address : Hex0F
beta_car
(reg. value)
0
1
2
3
4
5
6
7
Tn = 2π/ωn
(symb per)
NA
907
642
454
321
227
160
113
22
31
44
62
88
125
177
NA
NA
NA
fn (kHz) for
F = 20Mbd
alpha_car
(reg. value)
Damping Factor
0
NA
NA
NA
NA
1
NA
0.89
0.63
0.44 0.31 0.22 0.16 0.11
2
NA
1.77
1.25
0.89 0.63 0.44 0.31 0.22
3
NA
3.54
2.51
1.77 1.25 0.89 0.63 0.44
4
NA
7.09
5.01
3.54 2.51 1.77 1.25 0.89
5
NA 14.18 10.03 7.09 5.01 3.54 2.51 1.77
VI - CARRIER OFFSET EVALUATOR
An 8 bit register may be read at any time; it gives
a signed value proportionnal to the carrier frequency offset according to the expression :
∆f = 1.8 . 10-6 . m2 . N . Fs (in mode A)
where Fs is the symbol frequency, m the symbol
module (AGC reference), N the read value.
The maximum value for N is reached in nominal
conditions for a carrier offset of 16% of Fs ; if
greater, N remains saturated, giving a reliable sign
indication over more than ±50% Fs range.
Carrier Offset Register
Internal Address : Hex10
Signed number
This 8 bit R/W register may be written at any time
to force the central frequency of the derotator to
start the carrier research, or read, when the loop is
locked,in order to know the current carrier offset
(one LSB correspond to Fs/2048).
V.2 - Loop Equations
The natural pulsation is :
ωn = 10−3 ⋅ fs ⋅ √

m ⋅ 2beta_car

and the damping factor is :
m
ξ = 0.128 ⋅ 2alpha_car ⋅ √

.
beta_car
2
where m is the reference value (see AGC registers).
The next table gives for the nominal amplitude
m = 24 the natural period (in symbols), and the
damping factor for the possible values of alpha_car.
As an example, the corresponding natural frequency
is given assuming a symbol frequency of 20MBauds.
10/23
NA
Signed number
VI.1 - Lock Indicator
This 1 bit Carrier Found flag may be read (see
Viterbi Status register) at any time ; it indicates that
a QPSK signal is found, and that the carrier loop is
closed ; This flag allows to detect false lock that can
happen if the loop bandwidth is small regarding the
frequency offset.
VII - CARRIER TO NOISE INDICATOR
Internal Address : Hex14. Read only register.
b7
b6
b5
b4
b3
b2
b1
b0
This register can be used to estimate the carrier to
noise level (Eb/No) in a range from 4 to 16dB.
The register value dependson both the AGC reference level ”m” (see paragraph VIII) and the control
bits ”SN[1..0]” (see paragraph IX). For more details
about how to use this register, please refer to the
Annexe 1.
STV0196B
FUNCTIONAL DESCRIPTION (continued)
VIII - AGC CONTROL
IX - VITERBI DECODER AND SYNCHRONIZATION
The modulusof the input is compared to a programmable threshold; the difference is scaled by the
AGC coefficient, then integrated; the result is converted into a pulse density modulation signal to
drive the AGC output ; it may be filtered by a simple
analogue filter to control the gain command of any
amplifier before the A to D converter.
The convolutives codes are generated by the
polynoms Gx = 171oct and Gy = 133oct.
The Viterbi decoder computes for each symbol the
metrics of the four possible paths, proportional to
the square of the Euclidian distance between the
received I and Q and the theoretical symbol value.
The puncture rate and phase are estimated on the
error rate basis.
Five rates are allowed and may be enabled/disabled through register programming :
1/2, 2/3, 3/4,5/6, 7/8.
In Mode B, 7/8 is replaced by 6/7.
For each enabled rate, the current error rate is
compared to a programmable threshold; if it is
greater, anotherphase (or anotherrate) is tried until
the good rate is obtained.
A programmable hysteresis is added to avoid to
loose the phase during short term perturbation.
The rate may also be imposed by the external
software, and the phase is incremented only on
micro request ; the error rate may be read at any
time in order to use other algorithm than implemented.
The decoder is accessed via a set of 9 registers :
The 8 integrator MSB’s may be read or written at
any time by the micro; when written, the LSB’s are
reset. The integrator value is the level of the AGC
output, after low pass filtering ; it gives an image of
the input signal power, whatever this signal is, and
can be used to point the antenna.
The coefficient may be reset by programmation; in
that case, the AGC reduces to a programmable
voltage synthesiser.
The AGC reference level ”m” value impacts the
value of the following functions :
- carrier to noise indicator (see paragraph VII)
- the carrier loop (see paragraph V.2)
- the timing loop (paragraph IV.2)
- carrier offset evaluator (paragraph VI)
Control Registers
Internal Addresses : Hex11
Iagc
0
0
Invert
signal
Reserved
1
1
0
0
0
AGC reference
level (”m”)
Threshold
Value
Internal Addresses : Hex12
AGC integrator value (signed)
(Read/write register)
Internal Addresses : Hex13
0
0
0
Reserved
Threshold Registers (VTH0 to VTH4)
Internal Address : Hex1 (VTH0) to 5 (VTH4)
Reset Value : Hex20
0
0
0
1
0
G[2..0] :
AGC coefficient
The 8 bit signed value in the integrator is the image
of the AGC output; reading this value gives an
image of the RF signal power.
A constant error on the modulus leads to a ramp at
the output of the integrator with value :
AGC_Int = 2AGC_Coeff-16 . error
As a consequence, for the reset conditions, a constant signal of null value (error = 24) should cause
the output AGC duty cycle to go from 100% to 0%
in 222 symbol periods, or 8.7ms at 20MBauds.
If Iagc is set, the sign of the integrator is inverted.
VTH0 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0
rate 1/2
VTH1 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0
rate 2/3
VTH2 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0
rate 3/4
VTH3 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0
rate 5/6
VTH4 0 Th6 Th5 Th4 Th3 Th2 Th1 Th0
rate 7/8
or 6/7
For each register, bits 6 to 0 represent an error rate
threshold : the average number of errors happening during 256 bit periods; the maximum programmable value is 127/256 (higher error rates are of
no practical use).
Puncture Rate Enable register
Internal Address : Hex09
Reset Value : Hex10 (Mode A)
0
E4 :
E3 :
E2 :
E1 :
E0 :
0
0
E4
E3
E2
E1
E0
enablePuncturedRate 7/8(Mode A)or6/7(Mode B)
enable Punctured Rate 5/6
enable Punctured Rate 3/4
enable Punctured Rate 2/3
enable Basic Rate 1/2
11/23
STV0196B
FUNCTIONAL DESCRIPTION (continued)
IX - VITERBI DECODER AND SYNCHRONIZATION (continued)
Other Registers
VSEARCH
Internal Address : Hex06
A/M
F
SN [1..0]
TO [1..0]
H [1..0]
A/M
: Automatic/manual
F
: Freeze
SN [1..0] : Averaging period. It gives the number of
bits required to calculate the rate error :
SN [1..0]
Number of bits
00
1.024
01
4.096
10
16.384
11
65.536
Reset Value : SN=01 (4096 bits)
The SN[1..0] bits also inpacts the C/N
indicator (see paragraph VII).
TO [1..0] : Time out value. It programs the
maximum durationof the synchro word
research in automatic mode; if no sync
is found within this duration, the phase
is incremented.
Time out
(in 1024 bit periods)
00
16
01
32
10
64
11
128
Reset Value : TO=10 (64K bit periods).
synchro found ; this is the default (reset) mode.
- if A/M=0 and F=1, the current puncture rate is
frozen, if no sync is found, the phase is incremented, but not the rate number; this mode allows to shorten the recovery time in case of noisy
conditions: the puncture rate is not supposed to
change in a given channel.
In a typical computer aided implementation, the research begins in automatic mode; the micro reads
the error rate or the PRF flag in order to detect the
captureof a signal; thenit switches F to 1, until a new
channelis requested by the remote control.
- if AM=1 : manual mode; in this case, only one
puncture rate should be validated, the system is
forced to this rate, on the current phase, ignoring the
time-out registerandtheerror rate;in this mode, each
0 to 1 transition of the bit F leads to a phase incrementation, allowing full control of the operation by an
external micro by choosing the lowest error rate :
ResetValue: A/M=0, andF=0;automaticsearchmode
VERROR (Read only register)
Internal Address : Hex07
ERROR RATE
TO [1..0]
H [1..0]
: Hysteresis value. It programs the
maximum value of the Sync counter.
Th e un it is t h e b loc k du rat io n
(204 bytes in Mode A).
Sync Counter max value
(in blocks periods)
00
forbidden value
01
32
10
64
11
128
Reset Value : H=01 (32 blocks).
H [1..0]
In Mode A, the sync word is 47hex and it is complemented to B8hex for every 8th block.
An Up/Down Sync counter counts whenever a sync
word is recognized with the good timing, and
counts down for each missing sync word ; this
counter is bounded by a programmable maximum
value; when this value is reached, the LK bit
(”locked”) is set in VSTATUS register; when the
event counter counts down until 0, this flag is reset.
VSEARCH bit 7 (A/M) and bit 6 (F) programs the
automatic/manual(or computer aided) search mode :
- if A/M =0 and F=0 : automatic mode; successive
enabled punctured rates are tried with all possible
phases, until the system is locked and the block
12/23
At any time, the last value of the error rate may be
read in this register (unlike VTH, the possible range
is 0 to 255/256).
VSTATUS (Read only register)
Internal Address : Hex08
CF
0
0
PRF
LK
PR [2..0]
CF
: Carrier Found flag (see carrier recovery)
CF when set, indicates that a QPSK
signal is present at the input of the
Viterbi decoder.
PRF
: Puncture Rate Found
P RF indicat es t he sta te of th e
pu n ct ure rat e re search : 0 fo r
searching, 1 when found ; this bit is
irrelevant in manual mode.
LK
: Locked/searchingthe sync word
LK indicates the state of the sync word
research: 0 forsearching, 1 when found.
PR [2..0] : Current Puncture Rate
It hold the current puncture rate indice
with the correspondance :
Punctured Rate
Basic 1/2
Punctured 2/3
Punctured 3/4
Punctured 5/6
Punctured 7/8 (Mode A)
or 6/7 (Mode B)
Regiter Value PR[2..0]
100
000
001
010
011
STV0196B
FUNCTIONAL DESCRIPTION (continued)
X - CONVOLUTIONAL DE-INTERLEAVER
This is a 204 x 12 convolutional interleaver in
Mode A ; the periodicity of 204 bytes for sync byte
is preserved.
The de-interleaver may beskipped (see RS register).
XI - REED-SOLOMON DECODER
AND DESCRAMBLER
The input blocks are 204 byte long with 16 parity
bytes in Mode A; the synchro byte is the first byte
of the block. Up to 8 byte errors may be fixed.
Code Generator polynom:
g(x) = (x - ω0) (x - ω1) (...) (x - ω15)
over the Galois Field generated by :
X8 + X4 + X3 + X2 + 1 = 0
Energy dispersal descrambler :
Output energy dispersal descrambler generator :
X15 + X14 + 1
The polynom is initialised every eight blocks with
the sequence 100101010000000. The synchro
words are unscrambled.
Control register : RS register
Internal Address : Hex0A
The reset value is written in each register cell
RS7
RS6
RS5
RS4
RS3
RS2
RS1
RS0
1
0
1
1
1
0
0
0
RS7 : De-interleaver Enable
If 1, the input flow is deinterleaved.
If 0, the flow is not affected.
RS6 : If 0, Output data are corrected bytes
(normal operating mode).
If1, OutputdataareReed-Solomoncorrection
bytes (error count mode) (see Note 1).
RS5 : Reed-Solomon Enable
If 1, the input code is corrected.
If 0, no correction happens; all the data are
fed to the descrambler.
The error signal remains inactive.
RS4 : Descrambler Enable
If 1, the output flow from Reed-Solomon
decoder is descrambled.
If 0, the descrambler is desactived.
RS3 : Write Error Bit
If RS3=1, and uncorrectible error happens,
the MSB of the first byte following the sync
byte is forced to 1after descrambling.
RS2 : Super Synchro Suppression
If RS2=1, all synchro bytes are Hex47 in
mode A.
If RS2=0, the synchro is complemented
every 8th packet. It allows, when scrambler
is off, to provide RS coded signals for use
in low-cost SMATV interface.
RS1 : Output Clock Polarity
If RS1=0, data and control signals change
during high to low transition of CK_OUT.
If RS1=1, they change during the low to
high transition.
RS0 : Output Clock Configuration
If RS0=0, CK_OUT is continuous.
If RS0=1, CK_OUT remains low during the
parity bits.
Note 1 :
When RS6 = 1, the output data are the correction bytes
applied to data incoming the Reed-Solomon block.
The number of bits at 1 in these output data represent
therefore the number of errors remaining at the output of
VITERBI decoder.
All null output data mean no error left after VITERBI
decoding.
Remark : Output datas are meaningless when error flag (Pin 34) is
set to high level.
Figure 3
No Error
Da ta
CK_OUT
No Error
Uncorre cte d P a cke t
P a rity
Da ta
P a rity
Da ta
P a rity
RS 0=0, RS 1=0
RS 0=1, RS 1=0
RS 0=0, RS 1=1
RS 0=1, RS 1 =1
0196B-05.EPS
D/P
S TR_OUT
ERROR
13/23
STV0196B
Symbol
VDD
VI
Vo
Tstg
Toper
PD
Parameter
Value
-0.3 to 4
-0.3 to VDD + 0.3
-0.3 to VDD +0.3
-40 to +150
-10 to +85
1.5
Power Supply (1)
Voltage on Input pins (2)
Voltage on Output pins
Storage Temperature
Operating Ambient Temperature
Power Dissipation
Unit
V
V
V
o
C
o
C
W
0196B-02.TBL
ABSOLUTE MAXIMUM RATINGS
Maximum limits indicate where permanent device damages occur, continuous operation at these limits is
not intended and should be limited to those conditions specified in section ”DC Electrical Specifications”.
Notes : 1. All VDD to be tied together
2. SCL, SDA, NRES Pins can be tied to 5V ± 10% with an impedance ≥ 2kΩ (remark in these conditions the input leakage current
becomes higher than 10µA).
DC ELECTRICAL CHARACTERISTICS (VDD = 3.3V, Tamb = 25oC unless otherwise specified)
Parameter
Operating Voltage
IDD
Average Power Supply Current
VIL
VIH
VIL
VIH
ILK
C IN
VOL
VOH
Input Logic Low Voltage except M_CLK
Input Logic High Voltage except M_CLK
Input Logic Low Voltage for M_CLK
Input Logic High Voltage for M_CLK
Input Leakage Current
Input Capacitance
Output Logic Low Voltage
Output Logic High Voltage
Test conditions
o
o
0 C ≤ Toper ≤ 70 C
o
o
0 C < Toper < 85 C, M_CLK ≤ 55MHz
CLOAD = 20pF on all outputs,
M_CLK = 60MHz
M_CLK = 60MHz
M_CLK = 60MHz
Min.
3.0
3.15
Typ. Max.
3.3
3.6
3.3 3.45
300 480
-0.3
2.0
-0.3
2.2
0.8
3.6
0.8
3.6
10
VIN = 0V and VDD
3.5
CLOAD = 20pF, ILOAD = 2mA,
M_CLK = 60MHz
0.5
2.4
Unit
V
V
mA
V
V
V
V
µA
pF
V
V
0196B-03.TBL
Symbol
VDD
Note :This product doesn’t withstand the MIL 883C Norm at 2kV, but only at 1.5kV (all VDD tied together).
TIMING CHARACTERISTICS
Symbol
Parameter
Min.
Typ.
Max.
Unit
4
4
ns
ns
ns
ns
ns
ns
PRIMARY CLOCK (see Figure 4)
tM_CLK
tHIGH
tLOW
tR
tF
Master Clock Period
Clock
Clock
Clock
Clock
o
o
0 C ≤ Toper ≤ 70 C
0oC < Toper < 85oC
High Time
Low Time
Rising Edge
Falling Edge
16.6
18.2
6
6
I[5:0],Q[5:0] INPUT SPECIFICATION S (see Figure 5)
tSU
tH
I,Q stable before M_CLK
I,Q stable after M_CLK
4
4
ns
ns
D60 OUTPUT CHARACTERISTICS (see Figure 6)
t60
D60 period
(Tm_clk * 60)
- 10
(Tm_clk*60)
+10
ns
Bit RS1 = 1 in register RS ( adr = 0x0A) (see Figure 7)
D[7:0],D/P,STR_OUT,ERROR stable before CK_OUT Falling Edge
tCKSU
tCKH
D[7:0],D/P,STR_OUT,ERROR stable after CK_OUT Falling Edge
Bit RS1 = 0 in register RS ( adr = 0x0A) (see Figure 8)
tCKSU
D[7:0],D/P,STR_OUT,ERROR stable before CK_OUT Rising Edge
tCKH
D[7:0],D/P,STR_OUT,ERROR stable after CK_OUT Rising Edge
14/23
32
32
ns
ns
32
32
ns
ns
0196B-04.TBL
D[7:0],D/P,CK_OUT,STR_OUT,ERROR OUTPUT CHARACTERISTICS
STV0196B
I2C BUS CHARACTERISTICS (see Figure 9)
Parameter
Test Conditions
Min.
Typ. Max.
Unit
VIL
VIH
Input Logic Low Voltage
Input Logic High Voltage
See Note 1
-0.3
2.0
0.8
5.5
V
V
VOL
VOH
Output Logic Low Voltage
Output Logic High Voltage
C LOAD = 20pF, ILOAD = 2mA,
M_CLK = 60MHz, see Note 1
2.4
0.5
5.5
V
V
VIN = 0V to VDD, see Note 2
-10
10
µA
ILK
Input Leakage Current
C IN
Input Capacitance
IOL
Output Sink Current
tSP
Pulse W idth of Spikes which must be
suppressed by the Input filter
0
50
ns
fSCL
SCL Clock Frequency
0
400
kHz
tBUF
Bus Free Time between a STOP and START
Condition
1.3
µs
tHD,STA
Hold Time (repeated) START Condition. After
this period, the first clock pulse is generated.
0.6
µs
Low Period of the SCL Clock
High Period of the SCL Clock
1.3
0.6
µs
µs
tSU,STA
Set-up Time for a repeated START Condition
0.6
µs
tSU,STO
Set-up Time for STOP Condition
tHD,DAT
Data Hold Time
See Note 3
0
tSU,DAT
Data Set-up Time
See Note 4
100
tLOW
tHIGH
tR , tF
CB
4.
5.
pF
10
mA
µs
0.6
Rise and Fall Time of both SDA and SCL See Note 5
signals
Capacitive Load for each Bus Line
Notes : 1.
2.
3.
VOL = 0.5V
3.5
20 +
0.1 C B
0.9
µs
ns
300
ns
400
pF
An impedance higher than 2kΩ is required when SDA and SCL are tied to a 5V ± 10% voltage line.
Leakage current exceeds ± 10µA when SDA and SCL are tied to a 5V ± 10% line.
A device must internally provide a hold time of at least 300ns for the SDA signal (refered to the VIH Min. of the SCL signal) in order
to bridge the undefined region of the falling edge of SCL.
The maximum tHD,DAT has only to be met if the device does not stretch the low period (tLOW) of the SCL signal.
A fast-mode I2C bus device can be used in a standard-mode I2C bus system, but the requirement tSU,DAT ≥ 250ns must then be
met. This will automatically be the case if the device does not stretch the low period of the SCL signal. If such a device does
stretch the low period of the SCL signal, it must output the next data bit to the SDA line tR Max. + tSU,DAT = 1000 +250 = 1250ns
(according to the standard-mode I2C bus specification) before the SCL line is released.
CB = total capacitance of one bus line in pF.
15/23
0196B-03.BTBL
Symbol
STV0196B
Figure 4
tF
tR
2.0V
M_CLK
0.8V
tLOW
0196B-06.EPS
tHIGH
tM_CLK
Figure 5
M_CLK
(VIL + VIH) / 2
tS U
0196B-07.EPS
I,Q
tH
Figure 6
t60
Figure 8
CK_OUT
CK_OUT
D[7:0], D/P .
S TR_OUT,
ERROR
D[7:0], D/P .
STR_OUT,
ERROR
tCKSU
0196B-09.EPS
Figure 7
tCKH
tCKS U
tCKH
0196B-10.EPS
0196B-08.EPS
D60
Figure 9
SDA
tLOW
tR
tHD,S TA
tF
tSP
SC L
tHD,S TA
16/23
tHD,DAT
tHIGH
tS U,DAT
tS U,S TA
tS U,S TO
0196B-11.EPS
tBUF
3.3V
5V
12V
LNB
Supply
& Control
22µF
22µH
22µH
22µF
35V
C3
100nF
0196B-12.EPS
C1
100nF
100nF
C2
VDDA
VDDL
5VA
5V
12V
20V to 28V
C4
100nF
+5VA
C8
100nF
C5
100nF
+12V
C7
220µF
+5VA
BSFR68G15
TUNER
+5VA
C6
100nF
R1
1kΩ
C11
100nF
I
VDDL
VDDA
100nF
C16
100nF
100nF
C18
C17
100nF
C15
100nF
C14
R3 68Ω
C20
100nF
C19
22µF
R4 68Ω
C10
100nF
VDDA
Q
R2
1kΩ
C21
100nF
C22
22µF
R31
8.2kΩ
R30
22kΩ
VDDL
9
VDDL
16
15
13
14
17
VDDL
64
63
62
61
18
11
12
60
59
58
57
56
55
54
53
52
19
20
21
22
23
24
25
26
27
10
9
8
7
6
5
4
3
2
51
4
C45
1
48
C36
1nF
11
3
C42
2
47
3
1
VDDL
4
45
Pins 38-43-50
C46
5
44
C35
1nF
6
43
42
C37
1nF
41
40
39
C48
8
VDDL
9
C49
VDDL
10
Pins 27-31
C43
VDDL
VDDL
7
12
37
VDDL
VDDL
C44
11
38
VDDL
STV0196B
VDDL
R37
82kΩ
R34
82kΩ
R32
39kΩ
VDDL
14
C47
46
13
2
VDDL
R36
8.2kΩ
12
LM324
10
50
S
T
V
0
1
9
0
C33
1nF
8
5
C12
22µF
R19
560Ω
28
D1
BB909
R18 47Ω
5V
R22
22kΩ
R27
220kΩ
6
49
1
C23
100nF
C25
39pF
C13
100nF
74F04
R21
330Ω
C26
82pF
X1
40MHz L1 1µH
7
VDDL
R20
10kΩ
C24
220pF
Q1
BF959
R23
10kΩ
+5V
C34
100nF
20 to 28V
C50
14
35
15
34
Pins 4-3-10-12
C52
13
36
3.3kΩ
C51
16
33
+5V
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
3.3kΩ
VDDL
VDDL
13
12
11
10
9
8
7
6
5
4
3
2
1
D0
D1
D2
D3
D4
D5
D6
D7
CK_OUT
STR_OUT
D/P
ERROR
I2C BUS
DATA
OUTPUTS
RESET
D/60
GND
1
SCL
SDA
STV0196B
APPLICATION DIAGRAM : STV0196B/STV0190Fixed 20 MBauds Application
17/23
18/23
3.3V
5V
12V
LNB
Supply
& Control
22µH
22µH
22µF
35V
C3
100nF
0196B-13.EPS
C1
100nF
100nF
C2
22µF
V DDA
V DDL
5VA
5V
12V
20V to 28V
C4
100nF
+5VA
C8
100nF
C5
100nF
+12V
C7
220µF
+5VA
BSFR68G15
TUNER
+5VA
C6
100nF
R1
1kΩ
R2
1kΩ
C11
I
VDDL
VDDA
74F04
C17
100nF
100nF
C18
C20
100nF
6
5
4
3
2
16
15
14
17
13
12
18
11
20
21
22
23
19
S
T
V
0
1
9
0
24
25
26
27
28
10
9
8
C16
7
100nF
C15
100nF
C14
100nF
R3 68Ω
1
R19
560Ω
C31
33nF
C32
1nF
R26
10kΩ
C21
100nF
C33
10nF
R27 220kΩ
R25
10kΩ
D2
2 x BB909A
C30
220pF
R18 47Ω
5V
D1
R22
22kΩ
L2
0.33µH
C29
15pF
C28
100nF
C27
22µF
C13
100nF
C23
100nF
C25
39pF
C26
82pF
R23
10kΩ
R23
10Ω
C19
22µF
R4 68Ω
100nF
C10
100nF
VDDA
C12
22µF
V DDL
Q
R20
10kΩ
C24
220pF
R21
330Ω
Q1
BF959
+5V
C22
22µF
R31
8.2kΩ
R30
1kΩ
R29
100kΩ
R28
22kΩ
8
7
VDDL
VDDL
V DDL
9
6
64
63
62
61
60
59
58
57
56
55
54
53
52
51
50
49
4
C45
1
48
11
3
C42
2
47
14
1
3
4
45
V DDL
5
6
43
VDDL
42
C37
1nF
41
R33
1kΩ
40
39
C49
8
VDDL
9
C48
V DDL
10
Pins 27-31
C43
VDDL
V DDL
7
12
37
V DDL
V DDL
C44
11
38
V DDL
D3
5.1V
R43
1kΩ
+12V
C50
14
35
15
34
C40
100nF
Pins 4-3-10-12
C52
13
36
C41
22µF
C39
22µF
R40 22kΩ
R41 68Ω
STV0196B
C35
1nF
R37
82kΩ
C46
Pins 38-43-50
C47
46
V DDL
R36
8.2kΩ
44
R32
39kΩ
R35 1kΩ
13
2
VDDL
C36
1nF
12
LM324
10
5
C34
100nF
20 to 28V
R34
82kΩ
C30
2.2nF
C51
16
33
3.3kΩ
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
VDDL
VDDL
TC74SU04
3.3kΩ
+5V
D0
D1
D2
D3
D4
D5
D6
D7
D/60
CKOUT
STROUT
D/P
ERROR
RESET
SCL
SDA
VCO ADJ
(PWM)
STV0196B
APPLICATION DIAGRAM : STV0196B/STV0190Multirate Application
STV0196B
ANNEXE 1 : C/N ESTIMATION
The C/N indicator register permanently reports a
value S which depends on the C/N level at the input
of the STV0196B.
The C/Nindicator offers a programmable sensitivity
which allows a reliable C/N estimation over a wide
Eb/No range (4dB to 16dB typically) ; this is particularly useful to optimize the dish positioning.
Remark : In this note, we have assumed that :
C Eb
=
⋅ 2 (PR), PR : Puncture Rate
N No
The sensitivity of the C/N indicator is dependant on
the SN bits of the register VSEARCH (Hex06) and
on the AGC function reference level ”m”.
A - SUGGESTED PROCEDURE TO RELIABLY
ESTIMATE THE ACTUAL C/N
As no simple mathematical low ensumes a good
matching between the C/N indicator and the actual
C/N, the method relies on a comparaison of the
value S (reported by the C/N indicator) with a
reference look-up table which has been realized
under well controlled conditions.
Basically there are 3 steps in the C/N estimation
software.
1. To collect C/N indication (under adapted
conditions).
2. Indication scaling and correction versus the
puncture rate
3. Comparaison with the look-up table
A.1 - To collect C/N Indication
The purpose of this first step is to collect the C/N
indicator with the appropriate sensitivity (SN bits
and AGC reference level m).
Basically :
- The value reported by the C/N indicator is proportional to the Number of bits (at the output of the
VITERBI decoder) selected by the SN bits.
- The AGC reference level is only changed to appreciate the high Eb/No ratios. This second parameter has to be used with some care.
Procedure : Before to make an estimation, the
VSTATUS register (internal address Hex 08) must
be checked to make sure that :
- a carrier is actually present (bit 7)
- puncture rate is found (bit 4)
- puncture rate is known (bits 0-1-2)
Remark : Optionally, it is possible to make an
estimation without informations about the puncture
rate (useful when the dish is still very far from
optimum position), in such case the puncture rate
is forced.
The C/N indicator register has no overflow detection, so it is necessary to start the measure with the
lowest sensitivity (SN = 00) and to gradually increase it (using SN bits). Due to the noise, the result
S of the measure may have a lot of dispersion,
consequently it is recommended to measure S
several times (typically 100 times) and to calculate
the average value.
Remark : The requred duration tW between two
readings of the register must be higher than :
tW (Min.) =
BC
BR
BR = 2 (Fs) x (PR)
with
BC : Bit Count (selected by SN bits)
Fs : Symbol Rate
PR : Puncture Rate
When the current average value of the measure S
is lower than 63, the measure is done again with a
higher sensitivity. With this care the new C/N
measure S does not overflow t he counter
(the counting time is multiplied by 4 at each step).
In practice some margin is given to this threshold :
a higher sensitivity is selected when the average
value of S is lower than 60.
Wh en the maximum SN value is reached
(SN = 11 ⇔ to 65 536 bits at the output of the
VITERBI decoder), the sensitivity can be further
increased by lowering the AGC reference level
(p ara met er m, in t e rnal a dd res s He x11,
bit 0 to bit 5).
Remark : There is the need to change the AGC
reference level only in case of high C/N conditions,
then to change the reference level has no important
influence on the bit error rate (BER). In other words,
a completete C/N estimation can be run during the
operation of the receiver.
When the highest possible sensitivity is found
the result S (average value) is ready for further
process.
19/23
STV0196B
ANNEXE 1 : C/n ESTIMATION (continued)
A.2 - Scaling and Correction versus Puncture Rate
Scaling
This simple operation is recommended to easily
compare data which have been recorded under
different sensitivity conditions. To do so, the result
S of the C/N indication is multiplied by a coefficient
so that the scaled value would correspond to a
measure done with the highest counting period
(SN = 11).
Remark : Scaling is not done for results which have
beenrecordedafterchangingthe AGCreferencelevel.
Scaling operation : Scaled_value = (S) x (factor)
factor = 64 when C/N estimation is done with SN = 00
factor = 16 when C/N estimation is done with SN = 01
factor = 4 when C/N estimation is done with SN = 10
factor = 1 when C/N estimation is done with SN = 11
Correction versus puncture rate
This correction is not required when a reference
look-up table have been memorized for each possible puncture rate. When required, the correction
is done with respect to the puncture rate PRref of
the reference look-up table :
PRcurrent
Scorrected = (S) ⋅
PRref
20/23
PR current : the puncture currently identified with
the bits 0,1,2 of VSTATUS register.
A.3 - Comparing with the look-up table
In the application the read value Srs (scaled and
corrected) will seldom exactly match a value of the
look-up table ; consequentlythere will be the need
for some interpolation.
To make it simple, a linear interpolation is preferred,
with such a solution a good precision can be
achieved when the look-up table is built with a small
step for the C/N (or Eb/No).
Interpolation
Generally Ssr will be between two values of the
reference look-up table : V(Min.) ≤ Ssr ≤ V(Max.),
with V(Min.) corresponding C/N(Max.) and V(Max.)
c orre sp on din g t o C/ N(Min.) (wit h t yp ica lly
(C/N(Max.)) - (C/N(Min.)) = 0.5dB).
The calculated C/N corresponding to Ssr is :
V(Min.) − Ssr
C/N = C/N(Max.) − C/N(Max.) − C/N(Min.)  ⋅

 V(Min.) − V(Max.)
in above calculation C/N (or Eb/No) are given in
algebraic value (not in dB).
STV0196B
ANNEXE 1 : C/n ESTIMATION (continued)
B - FLOW CHART
Following is a simplified flow chart.
Normal Process
C/N
Estimation
A
Go to Tuning Routine
N
Y
N
Check
VSTATUS
OK
SN < ----- 00
Collect C/N data 100 times
and calculate average value
SN < ----- SN + 1
Y
S < 60
Scaling S
SN = 11
Y
N
Correction versus
Puncture Rate
S < 40
m < ----- m - 4
Collect C/N data 100 times
and calculate average value
N
S < 20
Y
m < ----- m - 4
Collect C/N data 100 times
and calculate average value
Compare with look-up table T1
Compare with look-up table T2
Compare with look-up table T3
Output C/N estimation
A
N
End of
Estimation
Y
T1 : Look-up table for normal value of m (AGC reference level)
T2 : Look-up table for m - 4
T3 : Look-up table for m - 8
0196B-17.EPS
Restore AGC Reference
Level to normal value
Return to normal process
21/23
STV0196B
ANNEXE 1 : C/n ESTIMATION (continued)
C - RESULTS
The results reported in the following table are typical values. When evaluating another application some
differences may be especially noticed when Eb/No is higher than 10dB, in these conditions the characteristics of the tuner and the A/D converter may influence the results.
Eb
Conditions : Puncture rate : 2/3, 20MBauds signal, DVB encoding (RS : 188/204), C/N =
2 ⋅ (PR)
No
Eb/No (dB)
Measurement Conditions
S
S Scaled
152
2.432
4.5
137
2.192
5
121
1.936
5.5
105
1.664
4
SN bits (hex)
AGC, m (dec)
1
20
6
92
1.474
6.5
78
1.248
7
64
1.024
7.5
205
820
8
168
672
8.5
131
524
9
98
392
9.5
73
292
10
2
3
20
212
212
10.5
20
146
146
11
95
95
61
61
11.5
12
3
16
12.5
84
13
55
13.5
35
14
22
14.5
15
22/23
122
13
3
12
128
15.5
95
16
70
STV0196B
PACKAGE MECHANICAL DATA
64 PINS - PLASTIC QUAD FLAT PACK
D
D1
A
D3
A2
A1
48
33
49
32
0. 10mm
E
E1
E3
B
B
Seating Plane
17
64
16
1
C
PMPQFP64.EPS
L
L1
e
K
PQFP64
A
A1
A2
B
C
D
D1
D3
e
E
E1
E3
K
L
L1
Min.
0.25
2.55
0.30
0.13
16.95
13.90
16.95
13.90
0.65
Millimeters
Typ.
2.80
17.20
14.00
12.00
0.80
17.20
14.00
12.00
0.80
1.60
Max.
3.40
Min.
3.05
0.45
0.23
17.45
14.10
0.010
0.100
0.0118
0.005
0.667
0.547
17.45
14.10
0.667
0.547
0o (Min.), 7o (Max.)
0.95
0.026
Inches
Typ.
0.110
0.677
0.551
0.472
0.0315
0.677
0.551
0.472
0.0315
0.063
Max.
0.134
0.120
0.0177
0.009
0.687
0.555
0.687
0.555
0.0374
Information furnished is believed to be accurate and reliable. However, SGS-THOMSON Microelectronics assumes no responsibility
for the consequences of use of such information nor for any infringement of patents or other rights of third parties which may result
from its use. No licence is granted by implication or otherwise under any patent or patent rights of SGS-THOMSON Microelectronics.
Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all
information previously supplied. SGS-THOMSON Microelectronics products are not authorized for use as critical components in life
support devices or systems without express written approval of SGS-THOMSON Microelectronics.
 1996 SGS-THOMSON Microelectronics - All Rights Reserved
Purchase of I2C Components of SGS-THOMSON Microelectronics, conveys a license under the Philips
I2C Patent. Rights to use these components in a I2C system, is granted provided that the system conforms to
the I2C Standard Specifications as defined by Philips.
SGS-THOMSON Microelectronics GROUP OF COMPANIES
Australia - Brazil - Canada - China - France - Germany - Hong Kong - Italy - Japan - Korea - Malaysia - Malta - Morocco
The Netherlands - Singapore - Spain - Sweden - Switzerland - Taiwan - Thailand - United Kingdom - U.S.A.
23/23
PQFP64.TBL
Dimensions