1998 May 15 10
Philips Semiconductors Product specification
Enhanced Video Input Processor (EVIP) SAA7111A
8 FUNCTIONAL DESCRIPTION
8.1 Analog input processing
The SAA7111A offers four analog signal inputs, two
analog main channels with source switch, clamp circuit,
analog amplifier, anti-alias filter and video CMOS ADC
(see Fig.5).
8.2 Analog control circuits
The anti-alias filters are adapted to the line-locked clock
frequency via a filter control circuit. During the vertical
blanking time, gain and clamping control are frozen.
8.2.1 C
LAMPING
The clamp control circuit controls the correct clamping of
the analog input signals. The coupling capacitor is also
used to store and filter the clamping voltage. An internal
digital clamp comparator generates the information with
respect to clamp-up or clamp-down. The clamping levels
for the two ADC channels are fixed for luminance (60) and
chrominance (128). Clamping time in normal use is set
with the HCL pulse at the back porch of the video signal.
8.2.2 G
AIN CONTROL
Signal (white) peak control limits the gain at signal
overshoots. The flow charts (see Figs 13 and 14) show
more details of the AGC. The influence of supply voltage
variation within the specified range is automatically
eliminated by clamp and automatic gain control.
The gain control circuit receives (via the I
2
C-bus) the static
gain levels for the two analog amplifiers or controls one of
these amplifiers automatically via a built-in automatic gain
control (AGC) as part of the Analog Input Control (AICO).
Fig.3 Analog line with clamp (HCL) and gain
range (HSY).
handbook, halfpage
HCL
MGL065
HSY
analog line blanking
TV line
1
60
255
GAIN CLAMP
The AGC (automatic gain control for luminance) is used to
amplify a CVBS or Y signal to the required signal
amplitude, matched to the ADCs input voltage range.
The AGC active time is the sync bottom of the video signal.
8.3 Chrominance processing
The 8-bit chrominance signal is fed to the multiplication
inputs of a quadrature demodulator, where two subcarrier
signals from the local oscillator DTO1 are applied
(0 and 90° phase relationship to the demodulator axis).
The frequency is dependent on the present colour
standard. The output signals of the multipliers are
low-pass filtered (four programmable characteristics) to
achieve the desired bandwidth for the colour difference
signals (PAL and NTSC) or the 0 and 90° FM-signals
(SECAM).
The colour difference signals are fed to the
Brightness/Contrast/Saturation block (BCS), which
includes the following five functions:
AGC (Automatic Gain Control for chrominance
PAL and NTSC)
Chrominance amplitude matching (different gain factors
for R Y and B Y to achieve CCIR-601 levels
Cr and Cb for all standards)
Chrominance saturation control
Luminance contrast and brightness
Limiting YUV to the values 1 (min.) and 254 (max.) to
fulfil CCIR-601 requirements.
Fig.4 Automatic gain range.
handbook, halfpage
analog input level
controlled
ADC input level
maximum
minimum
range tbf0 dB
0 dB
MGG063
+4.5 dB
7.5 dB
(1 V(p-p) 27/47 )
1998 May 15 11
Philips Semiconductors Product specification
Enhanced Video Input Processor (EVIP) SAA7111A
The SECAM-processing contains the following blocks:
Baseband ‘bell’ filters to reconstruct the amplitude and
phase equalized 0 and 90° FM-signals
Phase demodulator and differentiator
(FM-demodulation)
De-emphasis filter to compensate the pre-emphasised
input signal, including frequency offset compensation
(DB or DR white carrier values are subtracted from the
signal, controlled by the SECAM-switch signal).
The burst processing block provides the feedback loop of
the chroma PLL and contains;
Burst gate accumulator
Colour identification and killer
Comparison nominal/actual burst amplitude (PAL/NTSC
standards only)
Loop filter chrominance gain control (PAL/NTSC
standards only)
Loop filter chrominance PLL (only active for PAL/NTSC
standards)
PAL/SECAM sequence detection, H/2-switch
generation
Increment generation for DTO1 with divider to generate
stable subcarrier for non-standard signals.
The chrominance comb filter block eliminates crosstalk
between the chrominance channels in accordance with the
PAL standard requirements. For NTSC colour standards
the chrominance comb filter can be used to eliminate
crosstalk from luminance to chrominance (cross-colour)
for vertical structures. The comb filter can be switched off
if desired. The embedded line delay is also used for
SECAM recombination (cross-over switches).
The resulting signals are fed to the variable Y-delay
compensation, RGB matrix, dithering circuit and output
interface, which contains the VPO output formatter and the
output control logic (see Fig.6).
8.4 Luminance processing
The 8-bit luminance signal, a digital CVBS format or a
luminance format (S-VHS, HI8), is fed through a
switchable prefilter. High frequency components are
emphasized to compensate for loss. The following
chrominance trap filter (f
0
= 4.43 or 3.58 MHz centre
frequency selectable) eliminates most of the colour carrier
signal, therefore, it must be bypassed for S-video
(S-VHS and HI8) signals.
The high frequency components of the luminance signal
can be peaked (control for sharpness improvement via
I
2
C-bus) in two band-pass filters with selectable transfer
characteristic. This signal is then added to the original
(unpeaked) signal. A switchable amplifier achieves
common DC amplification, because the DC gains are
different in both chrominance trap modes. The improved
luminance signal is fed to the BCS control located in the
chrominance processing block (see Fig.7).
8.5 RGB matrix
Y, Cr and Cb data are converted after interpolation into
RGB data in accordance with CCIR-601
recommendations. The realized matrix equations consider
the digital quantization:
R = Y + 1.371 Cr
G=Y0.336 Cb 0.698 Cr
B = Y + 1.732 Cb.
After dithering (noise shaping) the RGB data is fed to the
output interface within the VPO-bus output formatter.
8.6 VBI-data bypass
For a 27 MHz VBI-data bypass the offset binary CVBS
signal is upsampled behind the ADCs. Upsampling of the
CVBS signal from 13.5 to 27 MHz is possible, because the
ADCs deliver high performance at 13.5 MHz sample clock.
Suppressing of the back folded CVBS frequency
components after upsampling is achieved by an
interpolation filter (see Fig.42).
The TUF block on the digital top level performs the
upsampling and interpolation for the bypassed CVBS
signal (see Fig.6).
For bypass details see Figs 8 to 10.
8.7 VPO-bus (digital outputs)
The 16-bit VPO-bus transfers digital data from the output
interfaces to a feature box or a field memory, a digital
colour space converter (SAA7192 DCSC), a video
enhancement and digital-to-analog processor
(SAA7165 VEDA2) or a colour graphics board
(Targa-format) as a graphical user interface.
1998 May 15 12
Philips Semiconductors Product specification
Enhanced Video Input Processor (EVIP) SAA7111A
The output data formats are controlled via the I
2
C-bus bits
OFTS0, OFTS1 and RGB888. Timing for the data stream
formats, YUV (4 : 1 : 1) (12-bit), YUV (4 : 2 : 2) (16-bit),
RGB (5, 6 and 5) (16-bit) and RGB (8, 8 and 8) (24-bit)
with an LLC2 data rate, is achieved by marking each
second positive rising edge of the clock LLC in conjunction
with CREF (clock reference) (except RGB (8, 8 and 8),
see special application in Fig.32). The higher output
signals VPO15 to VPO8 in the YUV format perform the
digital luminance signal. The lower output signals
VPO7 to VPO0 in the YUV format are the bits of the
multiplexed colour difference signals (B Y) and (R Y).
The arrangement of the RGB (5, 6 and 5) and
RGB (8, 8 and 8) data stream bits on the VPO-bus is given
in Table 6.
The data stream format YUV 4:2:2 (the 8 higher output
signals VPO15 to VPO8) in LLC data rate fulfils the
CCIR-656 standard with its own timing reference code at
the start and end of each video data block.
A pixel in the format tables is the time required to transfer
a full set of samples. If 16-bit 4 : 2 : 2 format is selected
two luminance samples are transmitted in comparison to
one (B Y) and one (R Y) sample within a pixel.
The time frames are controlled by the HREF signal.
Fast enable is achieved by setting input FEI to LOW.
The signal is used to control fast switching on the digital
VPO-bus. HIGH on this pin forces the VPO outputs to a
high-impedance state (see Figs 18 and 19). The I
2
C-bus
bit OEYC has to be set HIGH to use this function.
The digitized PAL, SECAM or NTSC signals AD1 (7 to 0)
and AD2 (7 to 0) are connected directly to the VPO-bus
via I
2
C-bus bit VIPB = 1 and MODE = 4, 5, 6 or 7.
AD1 (7 to 0) VPO (15 to 8) and
AD2 (7 to 0) VPO (7 to 0).
The selection of the analog input channels is controlled via
I
2
C-bus subaddress 02 MODE select.
The upsampled 8-bit offset binary CVBS signal (VBI-data
bypass) is multiplexed under control of the I
2
C-bus to the
digital VPO-bus (see Fig.8).
8.8 Reference signals HREF, VREF and CREF
HREF: The positive slope of the HREF output signal
indicates the beginning of a new active video line.
The high period is 720 luminance samples long and is
also present during the vertical blanking.
The description of timing and position from HREF is
illustrated in Figs 15, 16, 21 and 23.
VREF: The VREF output delivers a vertical reference
signal or an inverse composite blank signal controlled
via the I
2
C-bus [subaddress 11, inverse composite
blank (COMPO)]. Furthermore four different modes of
vertical reference signals are selectable via the I
2
C-bus
[subaddress 13, vertical reference output control
(VCTR1 and VCTR0)]. The description of VREF timing
and position is illustrated in Figs 15, 16, 24 and 25.
CREF: The CREF output delivers a clock/pixel qualifier
signal for external interfaces to synchronize to the
VPO-bus data stream.
Four different modes for the clock qualifier signal are
selectable via the I
2
C-bus [subaddress 13, clock
reference output control (CCTR1 and CCTR0)].
The description of CREF timing and position is
illustrated in Figs 16, 18, 20 and 21.
8.9 Synchronization
The prefiltered luminance signal is fed to the
synchronization stage. Its bandwidth is reduced to 1 MHz
in a low-pass filter. The sync pulses are sliced and fed to
the phase detectors where they are compared with the
sub-divided clock frequency. The resulting output signal is
applied to the loop filter to accumulate all phase
deviations. Internal signals (e. g. HCL and HSY) are
generated in accordance with analog front-end
requirements. The output signals HS, VS, and PLIN are
locked to the timing reference, guaranteed between the
input signal and the HREF signal, as further improvements
to the circuit may change the total processing delay. It is
therefore not recommended to use them for applications
which require absolute timing accuracy on the input
signals. The loop filter signal drives an oscillator to
generate the line frequency control signal LFCO
(see Fig.7).
8.10 Clock generation circuit
The internal CGC generates all clock signals required for
the video input processor. The internal signal LFCO is a
digital-to-analog converted signal provided by the
horizontal PLL. It is the multiple of the line frequency
Internally the LFCO signal is multiplied by a factor of 2 or 4
in the PLL circuit (including phase detector, loop filtering,
VCO and frequency divider) to obtain the LLC and LLC2
output clock signals. The rectangular output clocks have
a 50% duty factor (see Fig.26).
6.75MHz
429
432
----------
f
H
×=

SAA7111AHZ/V4,557

Mfr. #:
Manufacturer:
NXP Semiconductors
Description:
IC VIDEO INPUT PROCESSOR 64-LQFP
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union

Products related to this Datasheet