Product specification
File under Integrated Circuits, IC02
1999 March 01
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
FEATURES
16/32/64/128/256 QAM demodulator
•
(DVB-C compatible : ETS 300-429).
On chip 9-bit ADC.
•
On chip PLL for crystal frequency
•
multiplication.
Digital down conversion.
•
Half Nyquist filters (roll off = 15 %).
•
Automatic gain control PWM output
•
(AGC).
Symbol timing recovery, with
•
programmable second order loop filter.
Variable symbol rate capability from
•
SACLK/64 to SACLK/4
(SACLK max = 36 MHz)
• Programmable anti-aliasing filters.
• Full digital carrier recovery loop.
• Carrier acquisition range up to 8 % of
symbol rate.
Integrated adaptative equalizer (Linear
•
Transversal Equalizer or Decision
Feedback Equalizer).
• On chip FEC decoder (Deinterleaver &
RS decoder), full DVB-C compliant.
• DVB compatible differential decoding
and mapping.
• Parallel or serial transport stream
interface.
• I2C bus interface, for easy control.
• CMOS 0.35µm technology.
APPLICATIONS
DVB-C fully compatible.
•
Digital data transmission using QAM modulations.
•
Cable demodulation.
•
Cable modems
•
MMDS (ETS 300-429).
•
DESCRIPTION
The VES1820X is a single chip channel receiver for 16, 32, 64, 128
and 256-QAM modulated signals. The device interfaces directly to
the IF signal, which is sampled by a 9-bit AD converter.
The VES1820X performs the clock and the carrier recovery
functions. The digital loop filters for both clock and carrier recovery
are programmable in order to optimize their characteristics
according to the current application.
After base band conversion, equalization filters are used for echo
cancellation in cable applications. These filters are configured as Tspaced transversal equalizer or DFE equalizer, so that the system
performance can be optimized according to the network
characteristics. A proprietary equalization algorithm, independent of
carrier offset, is achieved in order to assist carrier recovery. Then a
decision directed algorithm takes place, to achieve final
equalization convergence.
The VES1820X implements a FORNEY convolutional deinterleaver
of depth 12 blocks and a Reed-Solomon decoder which corrects up
to 8 erroneous bytes. The deinterleaver and the RS decoder are
automatically synchronized thanks to the frame synchronization
algorithm which uses the MPEG2 sync byte. Finally descrambling
according to DVB-C standard, is achieved at the Reed Solomon
output. This device is controlled via an I2C bus.
Designed in 0.35 µm CMOS technology and housed in a 100 pin
MQFP package, the VES1820X operates over the commercial
temperature range.
1999 March 012
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
FIGURE 1 : FUNCTIONAL BLOCK DIAGRAM
IF
SACLK
XIN
ADC
PLL
9
NCO
BASE-BAND
CONVERSION
DE-INTERLEAVER
CLOCK
RECOVERY
FILTERS
BANK
SDA
SCL
EQUALIZER
R. S.
DECODER
INTERFACE
AGCTO AGC
2
IC
8
CARRIER
RECOVERY
DE-SCRAMBLER
PWM
OUTPUT
INTERFACE
DO
OCLK
DEN
1999 March 013
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
TABLE 1 : ABSOLUTE MAXIMUM RATINGS
ParameterMinMaxUnit
Ambient operating temperature : Ta070°C
DC supply voltage- 0.5+ 4.1V
DC Input voltage- 0.5VDD + 0.5V
DC Input Current± 20mA
Lead Temperature+300°C
Junction Temperature+150°C
Stresses above the absolute maximum ratings may cause permanent damage to the device. Exposure to absolute
maximum ratings conditions for extended periods may affect device reliability.
VIPPositive analog input0.5V
VIMNegative analog input-0.5V
temperature
@ IOH = + 2mA
@ IOL = + 2mA
Symbol Rate =6Mbd
1
All inputs are 5V tolerant
2
IOH, IOL = ± 4mA only for pins SACLK, OCLK, SDA, CTRL1, CTRL2, IT
1999 March 014
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
FUNCTIONAL DESCRIPTION
ADC
½
The VES1820X implements a 9-bit analog to digital converter. No external voltage references are required to use
the ADC.
PLL
½
The VES1820X implements a PLL used as clock multiplier by 1, 2, 3, 4, 5, 6, 7 or 8, so that the crystal can be
low frequency (fundamental tone).
DOWN CONVERTER AND NYQUIST FILTERS
½
The digital down converter performs the down conversion of the bandpass input signal into the 2 classical
quadrature I & Q channels. Then these two signals are passed through anti-alias filters and through a half
Nyquist filter having a fixed roll-off of 0.15. The digital filter gives a stop band attenuation of more than 40 dB.
EQUALIZER
½
After Nyquist filtering, the signal is fed to an equalization filter, for echo cancellation. This equalizer can be
configured as either a transversal Equalizer or a decision feedback equalizer. The following table shows some
echos configuration that the VES1820X corrects with an equivalent degradation of less than 1dB @ BER = 10
-4
.
DELAY
(nS)
50-10worst
150
and
800
1600-20worst
CARRIER RECOVERY
½
The carrier synchronizer implements a fully digital algorithm allowing to recover carrier frequency offsets up to
± 8 % symbol rate. A phase error detector followed by a progr ammable second order loop filter provides an
estimation of the carrier phase, to compensate the input carrier frequency offset.
½ CLOCK RECOVERY
A timing error detector implements an application of Gardner algorithm for digital clock recovery.
The resulting error is fed to a programmable second order loop filter, which provides a 8-bit command to the
NCO block. This one allows to determine the right sampling time instant of the input signal.
½ AUTOMATIC GAIN CONTROL
An estimation of input signal magnitude is performed and compared to a threshold value which is programmable
via the microcontroller interface. The resulting error is then filtered to produce an 10-bit command which is then
PWM encoded and provided on pin VAGC. The PWM signal can be passed through a single RC filter to control
the input gain amplifier.
½ OUTPUT INTERFACE
After carrier recovery, the demodulated output symbol must be decoded according to the constellation diagram
given by DVB standard for 16, 32, 64, 128 and 256 QAM. The r esulting symbols are then differentially decoded
(DVB compliant) and serially provided to the FEC part.
AMPLITUDE
(dB)
-12
and
-20
PHASE
worst
½BLOCK SYNCHRONIZATION
At demodulator output, the length of some error bursts may exceed that which can be reliably corrected by the
Reed-Solomon decoder. The implemented de-interleaving is a convolutional one (Forney) of depth 12. The first
operation consists in synchronizing the de-interleaver. This is accomplished by detecting α consecutive MPEG2
sync words (or
sync) which are present as the first byte of each packet.
1999 March 015
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
Next, the RAM memory associated with the deinterleaver fills up and the first deinterleaved bytes are pr ovided to
the input of the Reed-Solomon decoder. The state machine of the de-interleaver goes to the control phase which
counts β consecutive missed MPEG2 sync words (or
going back to the synchronization phase. α and β are programmable through the I2C interface.
When the inverted sync word is detected at the input of the de-interleaver, the bytes provided to the ReedSolomon decoder are inverted at the output of the deinterleaver.
REED-SOLOMON DECODER
½
The Reed-Solomon decoder decodes the symbol stream from the de-interleaver according to the (204, 188)
shortened Reed-Solomon code. Synchronization to Reed-Solomon code is defined over the finite Galois field GF
8
). The field generator polynomial is given by :
(2
sync ) before declaring the system desynchronized and
G(x) =
15
p
i=0
(x
+a
i
)
This Reed-Solomon decoder corrects up to eight erroneous symbols in each block. When the correction
capability of the decoder is exceeded, the block is not changed and is provided as it has been entered. In this
case the flag UNCOR is set and the MSB of the second byte in the MPEG2 frame is forced to one (error
indicator). The correction capability of the RS decoder can be inhibited.
DESCRAMBLER
½
In order to comply with energy dispersal requirements of radio transmission regulations and to ensure adequate
binary transitions, the MPEG2 frames are scrambled at the encoder side. Dual operation is achieved at the
output of the Reed-Solomon decoder using the same scrambler/descrambler. The polynomial for the pseudo
random binary sequence (PRBS generator is 1 + x
14
+ x15. The PRBS registers are initialized at the start of ever y
eight transport packets. To provide an initialization signal for the descrambler, the MPEG2 sync byte of the first
transport packet is inverted from 47
to B816. When detected, the descrambler is loaded with the initial sequence
16
"100101010000000". The descrambler can be inhibited.
INTERFACE
½
The VES1820X integrates an I2C interface in slave mode. This I2C interface fulfills the Philips component I2C
bus specification.
1999 March 016
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
INPUT - OUTPUT SIGNAL DESCRIPTION
SYMBOLPIN NUMBERTYPEDESCRIPTION
CLR#27IThe CLR# input is asynchronous and active low, and clears the
VES1820X. When CLR# goes low, the circuit immediately enters its
RESET mode and normal operation will resume 4 XIN falling edges
later after CLR# returned high. The I2C register contents are all
initialized to their default values. The minimum width of CLR# at low
level is 4 XIN clock periods.
XIN2IXTAL oscillator input pin. Typically a fundamental XTAL oscillator is
connected between the XIN and XOUT pins (see typical application
on FIGURE 10 page 15). The XTAL frequency MUST be chosen so
that the system frequency SYSCLK (= XIN * multiplying factor of the
PLL) equals to 1.6 times the tuner output Intermediate Frequency :
SYSCLK = 1.6 x IF.
XOUT3OXTAL oscillator output pin. Typically a fundamental XTAL oscillator is
connected between the XIN and XOUT pins (see typical application
FIGURE 10 page 15).
SACLK18O
FI[8:0]5,6,7,8,12,
13,14,15,16
VAGC20O
DO[7:0]46,49,50,51
52,53,54,55
OCLK44O
DEN45O
UNCOR42O
PSYNC43O
TESTO[16:0]78,77,76,75,74
71,70,69,68,67
64,63,62,61,60
57,56
(5V)
(5V)
(3.3V)
(3.3V)
(3.3V)
(3.3V)
(3.3V)
(3.3V)
Sampling CLocK. This output clock can be fed to an external 9-bit
ADC as the sampling clock.SACLK = SYSCLK/2.
IFI [8:0] is the 9-bit input of the IF signal. FI[8:0] is the output of an
external A/D converter. FI[8] is the MSB. When not used, must be
tied to ground.
PWM encoded output signal for AGC. This signal is typically fed to
the AGC amplifier through a single RC network (see typical
application FIGURE 11 page 16). The maximum signal frequency on
VAGC output is XIN/16. AGC information is refreshed every 1024
symbols.
O
Data Output bus . These 8-bit parallel data are the outputs of the
VES1820X after demodulation, de-interleaving, RS decoding and descrambling.
When one of the two possible parallel interfaces is selected
(Parameter SERINT=0, index 20
) then DO[7:0] is the transport
stream output.
When the serial interface is selected (Parameter SERINT=1, index
) then the serial output is on pin DO[0] (pin 55).
20
16
Output CLock. OCLK is the output clock for the parallel DO[7:0]
outputs. OCLK is internally generated depending on which interface
is selected.
Data ENable : this output signal is high when there is a valid data on
output bus DO[7 :0].
UNCORrectable packet. This output signal is high when the provided
packet is uncorrectable (during the 188 bytes of the packet). The
uncorrectable packet is not affected by the Reed Solomon decoder,
but the MSB of the byte following the sync. byte is forced « 1 » for
the MPEG2 process : Error Flag Indicator (if RSI and IEI are set low
in the I2C table).
Pulse SYNChro. This output signal goes high when the sync byte
) is provided, then it goes low until the next sync byte. If the
(47
serial interface is selected, then PSYNC is high only during the first
bit of the sync byte (47
O
TESTO [16:0] is 17-bit Test output bus.
). See FIGURE 8 page 14.
16
1999 March 017
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
SYMBOLPIN NUMBERTYPEDESCRIPTION
IICDIV[1:0]21,22IIICDIV[1:0] allow to select the frequency of the I2C internal system
clock, depending on the crystal frequency. Internal I2C clock is a
division of XIN by 2
SADDR[1:0]23,24ISADDR[1:0] are the 2 LSBs of the I2C address of the VES1820X.
The MSBs are internally set to 00010. Therefore the complete I2C
address of the VES1820X is (MSB to LSB) : 0, 0, 0, 1, 0, SADDR[1],
SADDR[0].
SDA26I/O
(5V)
SDA is a bidirectional signal. It is the serial input/output of the I2C
internal block. A pull-up resistor (typically 4.7 kΩ) must be connected
between SDA and VDD for proper operation (Open Drain output).
SCL25II2C clock input. SCL should nominally be a square wave with a
maximum frequency of 400KHz. SCL is generated by the system I2C
master.
TEST19ITest input pin. For normal operation of the VES1820X, TEST must
be grounded.
TRST35ITest ReSeT. This active low input signal is used to reset the TAP
controller when in boundary scan mode. In normal mode of operation
TRST must be set low.
TDO37O
(5V)
Test Data Out. This is the serial Test output pin used in boundary
scan mode. Serial Data are provided on the falling edge of TCK.
TCK33ITest ClocK : an independant clock used to drive the TAP contr oller
when in boundary scan mode. In normal mode of operation, TCK
must be grounded.
TDI34ITest Data In. The ser ial input for Test data and instruction when in
boundary scan mode. In normal mode of operation, TDI must be set
to GND.
TMS36ITest Mode Select. This input signal provides the logic levels needed
to change the TAP controller from state to state. In normal mode of
operation, TMS must be set to VDD.
CTRL131I/O
(5V)
CTRL1 is equivalent to SDA I/O of VES1820X but can be tri-stated
by I2C programmation. It is actually the output of a switch controlled
by parameter BYPIIC of register TEST (index 0F
drain output, and therefore requires an external pull up resistor.
CTRL232O
(5V)
CTRL2 can be configured to be a control line output or to output SCL
input. This is controlled by parameter BYPIIC of register TEST (index
). CTRL2 is an open drain output and therefore requires an
0F
external pull up resistor.
IT38O
(5V)
InTerrupt line. This active low output interrupt line can be configured
by the I2C interface. See registers ITsel (index 32
(index 33
). IT is an open drain output and therefore requires an
external pull up resistor.
FEL39O
(5V)
By default FEL is a front-end lock indicator. In this case FEL is an
open drain output and therefore requires an external pull up resistor.
But FEL can also be configured to output a PWM signal, which value
can be programmed through the I2C interface (see register
PWMREF, index 34
VIP92IPositive input to the A/D converter. This pin is DC biased to half-
supply through an internal resistor divider (2 x 10kΩ resistors). In
order to remain in the range of the ADC, the voltage difference
between pins VIP and VIM should be between -0.5 and 0.5 volts.
VIM91INegative input to the A/D conver ter. This pin is DC biased to half-
supply through an internal resistor divider (2 x 10kΩ resistors). In
order to remain in the range of the ADC, the voltage difference
between pins VIP and VIM should be between -0.5 and 0.5 volts.
IICDIV
and must be between 6 and 20 MHz.
). CTRL1 is open
).
16
) and ITstat
1999 March 018
Philips SemiconductorsProduct specification
Single chip DVB-C channel receiverVES1820X
SYMBOLPIN NUMBERTYPEDESCRIPTION
CMCAP85IThis pin is connected to a tap point on an internal resistor divider
used to create CMO and CMI. An external capacitor of value 0.1µf
should be connected between this point and ground to provide good
power supply rejection from the positive supply at higher
frequencies.
RBIAS82IAn external resistor of value 3.3k should be connected between
this pin and ground to provide good accurate bias currents for the
analog circuits on the ADC.
VREF88OThis is the output of an on-chip resistor divider. An external capacitor
of value 0.1µf should be connected between this point and ground to
provide good power supply rejection from the positive supply at
higher frequencies. Reference voltages VREFP and VREFM are
derived from the voltage on VREF.
VREFP87OThis is a positive voltage reference for the A/D converter. It is derived
from the voltage on pin VREF through an on-chip fully-differential
amplifier. The voltage on this pin is nominally equal to CMO + 0.25
volts.
VREFM86OThis is the negative voltage reference for the A/D converter. It is
derived from the voltage on pin VREF through an on-chip fullydifferential amplifier. The voltage on this pin is nominally equal to
CMO- 0.25 volts.
CMO84OThis pin provides the common-mode out voltage for the analog
circuits on the ADC. It is the buffered version of a voltage derived
from an on-chip resistor devider, and has a nominal value of 0.5 x
VD3.
CMI83OThis pin provides the common-mode in voltage for the analog circuits
on the ADC. It is the buffered version of a voltage derived from an
on-chip resistor devider, and has a nominal value of 0.75 x VD3.
VD181IPower supply input for the digital switching circuitry (3.3 typ).
VS180IGround return for the digital switching circuitry.
VD294IPower supply input for the analog clock drivers (3.3V typ).
VS293IGround return for the analog clock drivers.
VD389IPower supply input for the analog circuits (3.3V typ).
VS390IGround return for analog circuits.
VD495IPower supply input that connects to an n-well guard ring that
surrounds the ADC (3.3V typ).
DVCC96I3.3V supply for the digital section of the PLL.
DGND97IGround connection for the digital section of the PLL.
PLLGND98IGround connection for the analog section of the PLL.
PLLVCC99I3.3V supply for the analog section of the PLL.
PPLUS100IP-well bias for the analog section of the PLL. Must be tied to 0V.