TW2802/4 Multiple Video Decoder: Preliminary Data Sheet From Techwell, Inc
TW2802/4 Multiple Video Decoder: Preliminary Data Sheet From Techwell, Inc
TW2802/4 Multiple Video Decoder: Preliminary Data Sheet From Techwell, Inc
Disclaimer
This document provides technical information for the user. Techwell Inc. reserves the right to
modify the information in this document as necessary. The customer should make sure that
they have the most recent data sheet version. Techwell Inc. holds no responsibility for any
errors that may appear in this document. Customers should take appropriate action to ensure
their use of the products does not infringe upon any patents. Techwell Inc. respects valid
patent rights of third parties and does not infringe upon or assist others to infringe upon such
rights.
Table of Contents
TMPSENS (Temporal Sensitivity) ____ 20
Introduction and Features ______________3
Velocity Control _________________ 21
Features ___________________________3 Mask Detection Region____________ 22
Applications _______________________3 Output Format ___________________ 23
Block Diagram _____________________4 ITU-R BT.656 Format ____________ 23
8-bit ITU-R BT.601 Format ________ 24
Pin Diagram _______________________5
Dual ITU-R BT.656 Format in 54MHz 25
Pin Description _____________________5
Host Interface ______________________ 26
Analog Interface Pins _______________5
Digital Data Interface Pins ___________6 Serial Interface ___________________ 26
System Control Pins ________________7 Parallel Interface _________________ 27
Power/Ground Pins_________________7
Interrupt Interface ________________ 28
Functional Description _________________8
Control Register __________________ 29
Video Input Formats ________________8 Register Map____________________ 29
Analog-to-Digital Converter __________8 Recommended Value _____________ 31
Register Description ______________ 33
Sync Processing ____________________9
Video Level Adjustment ____________9 Parametric Information_______________ 71
Horizontal Sync Processing __________9 DC Electrical Parameters___________ 71
Vertical Sync Processing ____________9
AC Electrical Parameters___________ 73
Color Decoding ____________________10
Package Dimension __________________ 75
Decimation Filter _________________10
Y/C Separation ___________________11 Application Information ______________ 77
The TW280X includes four high quality − Four built-in motion detectors for
NTSC/ PAL video decoders, which convert security system
analog composite to digital component − Supports the standard ITU-R BT.656 /
YCbCr for security application. The TW280X 8bit ITU-R BT.601 format
contains four 10-bit A/D and proprietary − Supports two differently scaled output
digital gain/clamp controllers and utilizes mode with 54MHz ITU-R BT.656 format
proprietary techniques for separating lumin- − Supports a two-wire serial or parallel
ance & chrominance to reduce both cross- interface
luminance and cross-chrominance artifacts. − Low power consumption
The high performance dual scalers in each − 128 PQFP package
channel offer two differently scaled video
outputs with 54MHz ITU-R BT.656 format Applications
for security system design. Four built-in Security systems
motion detectors can also increase the
feature of security system.
Device Options
Device Name Features
Features TW2802 2 Channel Video Decoder
TW2804 4 Channel Video Decoder
− Accepts all NTSC (M/N/4.43) / PAL
(B/D/G/H/I/K/L/M/N/60) standard
formats with auto detection VIN1A
H/V Scaler
Color Decoder
VD1[7:0]
− Four 10-bit video CMOS analog to VIN1B
MUX ADC with
Comb Filter
H/V Scaler
VALID1
HS1
digital converters VS1
FLD1
Motion Detector H/V Sync Processor
ACTIV1
− Adjust video level with proprietary NVMD1
HS2
of color demodulation
NVMD3
−
Host Interface Clock Generator CLK27O
Dual high quality horizontal and vertical HRDB
HWRH
HDAT
Block Diagram
H/V Scaler
VIN1A Color Decoder
VD1[7:0]
MUX ADC with
VALID1
VIN1B Comb Filter
H/V Scaler
HS1
VS1
FLD1
Motion Detector H/V Sync Processor
ACTIV1
NVMD1
H/V Scaler
VIN2A Color Decoder
VD2[7:0]
MUX ADC with
VALID2
VIN2B Comb Filter
H/V Scaler
HS2
VS2
FLD2
Motion Detector H/V Sync Processor
ACTIV2
NVMD2
H/V Scaler
VIN3A Color Decoder
VD3[7:0]
MUX ADC with
VALID3
VIN3B Comb Filter
H/V Scaler
HS3
VS3
FLD3
Motion Detector H/V Sync Processor
ACTIV3
NVMD3
H/V Scaler
VIN4A Color Decoder
VD4[7:0]
MUX ADC with
VALID4
VIN4B Comb Filter
H/V Scaler
HS4
VS4
FLD4
Motion Detector H/V Sync Processor
ACTIV4
NVMD4
HSPB
HCSB
HALE
Host Interface Clock Generator CLK27O
HRDB
HWRH
HDAT
IRQ CLK54I
Pin Diagram
102
101
100
77
75
73
71
68
66
99
98
97
96
95
94
93
92
91
90
89
88
87
86
85
84
83
82
81
80
79
78
76
74
72
70
69
67
65
VDDO
VDDO
HS4
FLD4
CLK27O
VSS
CLK54I
VD4[0]
VD4[5]
VD4[6]
VSS
VALID4
ACTIVE4
VS4
VDD
VDD
VDD
HDAT[0]
VSS
HDAT[3]
VSS
HDAT[5]
VSS
HDAT[6]
VSS
VD4[1]
VD4[2]
VD4[3]
VD4[4]
VSS
VD4[7]
VSS
VD3[0]
HDAT[1]
HDAT[2]
HDAT[4]
HDAT[7]
NVMD4
103 IRQ VD3[1] 64
104 HWRB VSS 63
105 VDDO VD3[2] 62
106 HRDB VD3[3] 61
107 HALE VDD 60
108 VSS VD3[4] 59
109 HCSB VD3[5] 58
110 HSPB VSS 57
111 VDDAD VD3[6] 56
112 VDDA VD3[7] 55
113 VIN1A VSS 54
114 VIN1B VALID3 53
115
116
VSSA
VSSA
TW280X ACTIVE3
VSS
52
51
117 VIN2A (128QFP) HS3 50
118 VIN2B VS3 49
119 VDDA Analog Pin VDD 48
120 VDDA (4 ADC) FLD3 47
121 VIN3A NVMD3 46
122 VIN3B VDDO 45
123 VSSA VD2[0] 44
124 VSSA VD2[1] 43
125 VIN4A VSS 42
126 VIN4B VD2[2] 41
127 VDDA VD2[3] 40
128 VSSAD VSS 39
ACTIVE1
ACTIVE2
NVMD1
NVMD2
VALID2
VALID1
VD1[7]
VD1[4]
VD1[3]
VD1[1]
VD2[6]
VD2[5]
VD1[6]
VD1[5]
VD1[2]
VD1[0]
VD2[7]
VD2[4]
VDDO
VDDO
RSTB
TEST
FLD2
FLD1
VDD
VDD
VDD
VSS
VSS
VSS
VSS
VSS
VSS
VSS
HS1
HS2
VS2
VS1
26
28
30
32
35
37
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
27
29
31
33
34
36
38
2
8
9
1
Pin Description
Power/Ground Pins
Name Number Type Description
9,21,33,48,60,
VDD P Digital power for internal logic. 2.5V.
72,90,102
6,24,45,
VDDO P Digital power for output driver. 3.3V.
66,84,105
3,12,15,18, 27,
30,36,39,42,51,
VSS 54,57,63,69,75, G Digital ground.
78,81,87,93,96,
99,108
VDDA 112,119,120,127 P Analog power. 2.5V.
VSSA 115,116,123,124 G Analog ground.
VDDAD 111 P Analog digital power. 2.5V.
VSSAD 128 G Analog digital ground.
Functional Description
Analog-to-Digital Converter
The TW280X contains four 10-bit Analog to Digital converters that digitizes the analog video
inputs. As the inputs are digitized at greater than two times that of the Nyquist sampling rate,
only simple external anti-aliasing LPF are needed to prevent out-of-band frequencies. Each
ADC has two analog switches that are controlled by ANA_SW (0x22, 0x62, 0xA2, 0xE2)
registers. The A/D converters can also be put into power-down mode by the ADC_PWDN
(0x78) registers.
Sync Processing
The sync processor of TW280X detects horizontal synchronization and vertical synchronization
signals in the composite. The TW280X utilizes proprietary technology for locking to weak, noisy,
or unstable signals such as those from on air signal and fast forward or backward of VCR
system.
Color Decoding
Decimation Filter
The digitized composite video data at 2X pixel clock rate first passes through decimation filter.
The decimation filter is required to achieve optimum performance and prevent high frequency
components from being aliased back into the video image. Fig 1 shows the characteristic of the
decimation filter.
-10
Magnitude Response (dB)
-20
-30
-40
-50
-60
0 2 4 6 8 10 12
Frequency (Hertz) 6
x 10
-10
Magnitude Response (dB)
-20
-30
-40
-50
-60
0 1 2 3 4 5 6
Frequency (Hertz) 6
x 10
-10
Magnitude Response (dB)
-20
-30
-40
-50
-60
0 1 2 3 4 5 6
Frequency (Hertz) 6
x 10
Luminance Processing
The luminance signal is separated by adaptive comb or trap filter is then fed to a peaking circuit.
The peaking filter enhances the high frequency components of the luminance signal. Fig. 4
shows the characteristics of the peaking filter for four different gain modes. The picture contrast
and brightness adjustment is provided through CONT (0x11, 0x51, 0x91, 0xD1) and BRT (0x12,
0x52, 0x92, 0xD2) registers. The contrast adjustment range is from approximately 0 to 200
percent, and the brightness adjustment is in the range of ±25 IRE. Moreover, a high frequency
coring function is also embedded in TW280X to minimize a high frequency noise. The coring
level is adjustable through the Y_H_CORE (0xF8) register.
5
Manitude Response (dB)
0
0 1 2 3 4 5 6
Frequency (Hertz) 6
x 10
Chrominance Processing
Chrominance Demodulation
The chrominance demodulation is done by first quadrature mixing for NTSC and PAL. The
mixing frequency is equal to the sub-carrier frequency of NTSC and PAL. After the mixing, a
LPF is used to remove 2X carrier signal and yield chrominance components. The LPF
characteristic can be selected for optimized transient color performance. In case of a mistuned
IF source, IF compensation filter makes up for any attenuation at higher frequencies or
asymmetry around the color sub-carrier. The gain for the upper chrominance side band is
controlled by IFCMP_MD (0x13, 0x53, 0x93, 0xD3) register. Fig. 5 and Fig. 6 show the
frequency response of IF-compensation filter and chrominance LPF.
10
5
Magnitude Response (dB)
-5
-10
-15
1.5 2 2.5 3 3.5 4 4.5 5 5.5
Frequency (Hertz) 6
x 10
-5
-10
-20
-25
-30
-35
-40
-45
0 0.5 1 1.5 2 2.5 3 3.5 4
Frequency (Hertz) 6
x 10
Video Scaling
The TW280X includes a high quality horizontal and vertical down scaler. The video images can
be downscaled in both horizontal and vertical direction to an arbitrary size. The luminance
horizontal scaler includes an anti-aliasing filter to reduce image artifacts in the resized image
and a 32 poly-phase filter to accurately interpolate the value of a pixel. This results in more
aesthetically pleasing video as well as higher compression ratios in bandwidth-limited
applications. Fig 7 shows the frequency response of anti-aliasing filter for horizontal scaling and
Fig 8 shows the 32 poly-phase filter characteristics. Similarly, the vertical scaler also contains
an anti-aliasing filter and 16 poly-phase filter for down scaling. The filter characteristics are
shown in Fig. 9.
-5
-10
Magnitude Response (dB)
-15
-20
-25
-30
-35
-40
-45
0 1 2 3 4 5 6
Frequency (Hertz) 6
x 10
0.5
0.4
0.3
0.2
-0.1
-0.2
-0.3
-0.4
-0.5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Frequency (Hertz) 6
x 10
-5
-10
Magnitude Response (dB)
-15
-20
-25
-30
-35
-40
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Vertical Frequency/Line Rate
Following equation is used to determine the horizontal scaling ratio to be written into the 16bit
HSCALE register.
HSCALE = [Npixel_desired/ HACTIVE] * (2^16 – 1)
Where Npixel_desired is the desired number of active pixels per line
For example, to scale full picture (HACTIVE is 720) to CIF (360 pixels), the HSCALE value can
be found as:
HSCALE = [320/720] * (2^16 – 1) = 0x7FFF
Following equation is used to determine the vertical scaling ratio to be written into the 16bit
VSCALE register.
VSCALE = [Nline_desired / VACTIVE] * (2^16 - 1)
Where Nline_desired is the desired number of active lines per field
For example, to scale full picture (VACTIVE is 240or288) to CIF (120/144 lines), the VSCALE
value can be found as:
VSCALE = [120 / 240] * (2^16 – 1) = 0x7FFF for 60Hz
VSCALE = [144 / 288] * (2^16 – 1) = 0x7FFF for 50Hz
Table 2 HSCALE and VSCALE value for some popular video formats
Scaling Output
Format HSCALE VSCALE
Ratio Resolution
NTSC 720x480 0xFFFF 0xFFFF
1
PAL 720x576 0xFFFF 0xFFFF
NTSC 360x240 0x7FFF 0x7FFF
1/2 (CIF)
PAL 360x288 0x7FFF 0x7FFF
NTSC 180x120 0x3FFF 0x3FFF
1/4 (QCIF)
PAL 180x144 0x3FFF 0x3FFF
The horizontal delay register HDELAY determines the number of pixel delays between the
horizontal reference and the leading edge of the active region. The horizontal active register
HACTIVE determines the number of active pixels to be processed. Note that these values are
referenced to the pixel number before scaling. Therefore, even if the scaling ratio is changed,
the active video region used for scaling remains unchanged as set by the HDEALY and
HACTIVE register. In order for the cropping to work properly, the following equation should be
satisfied.
To process full size region, the HDELAY should be set to 32 and HACTIVE set to 720 for both
60Hz and 50Hz system.
The vertical delay register (VDELAY) determines the number of line delays from the vertical
reference to the start of the active video lines. The vertical active register (VACTIVE)
determines the number of lines to be processed. These values are referenced to the incoming
scan lines before the vertical scaling. In order for the vertical cropping to work properly, the
following equation should be satisfied.
To process full size region, the VDELAY should be set to 7 and VACTIVE set to 240 for 60Hz
and the VDELAY should be also set to 4 and VACTIVE set to 288 for 50Hz.
V reference
VDELAY
VACTIVE
HDELAY HACTIVE
H reference
V reference
VDELAY
VACTIVE * VSCALE
VACTIVE
HDELAY HACTIVE
H reference
Motion Detector
The TW280X supports hardware motion detector for 4 channels individually. The motion
detection algorithm built in the TW280X uses difference between two luminance levels of the
adjacent two fields. Motion is detected for full screen image and each channel has 144(12x12)
mask regions, which enable or disable motion detection for that region. The motion detection
has several attributes, sensitivity and velocity of motion detector controlled by programming the
register. The Host takes the result of motion detection via IRQ or NVMD pin. Refer to the host
Interface for the detail.
Sensitivity Control
The motion detector has three sensitivity control parameters. One is level sensitivity control
parameter (LVLSENS), another is spatial sensitivity control parameter (SPTSENS), and a third
is temporal sensitivity control parameter (TMPSENS). The recommended values of sensitivity
control parameters for a proper operation are listed in Table 3
Velocity Control
Motion has various velocities. That is, in a fast motion an object appears and disappears rapidly
between the adjacent fields while in a slow motion it is to the contrary. As the built-in motion
detection algorithm uses the luminance level difference between two adjacent fields, a slow
motion is inferior in detection rate to a fast motion. To compensate this weakness, the
MDPERIOD parameter is used. MDPERIOD parameter adjusts the field interval in which the
luminance level is compared. Thus, for detection of a fast motion a small value is needed and
for a slow motion a large value is required. The parameter MDPERIOD value should be greater
than TMPSENS value.
720 P ixels
M ask1[0] M ask1[1] M ask1[2] M ask1[3] M ask1[4] M ask1[5] M ask1[6] M ask1[7] M ask1[8] M ask1[9] M ask1[10] M ask1[11]
M ask2[0] M ask2[1] M ask2[2] M ask2[3] M ask2[4] M ask2[5] M ask2[6] M ask2[7] M ask2[8] M ask2[9] M ask2[10] M ask2[11]
M ask3[0] M ask3[1] M ask3[2] M ask3[3] M ask3[4] M ask3[5] M ask3[6] M ask3[7] M ask3[8] M ask3[9] M ask3[10] M ask3[11]
240 Lines for 60H z, 288 Lines for 50H z
M ask4[0] M ask4[1] M ask4[2] M ask4[3] M ask4[4] M ask4[5] M ask4[6] M ask4[7] M ask4[8] M ask4[9] M ask4[10] M ask4[11]
M ask5[0] M ask5[1] M ask5[2] M ask5[3] M ask5[4] M ask5[5] M ask5[6] M ask5[7] M ask5[8] M ask5[9] M ask5[10] M ask5[11]
M ask6[0] M ask6[1] M ask6[2] M ask6[3] M ask6[4] M ask6[5] M ask6[6] M ask6[7] M ask6[8] M ask6[9] M ask6[10] M ask6[11]
M ask7[0] M ask7[1] M ask7[2] M ask7[3] M ask7[4] M ask7[5] M ask7[6] M ask7[7] M ask7[8] M ask7[9] M ask7[10] M ask7[11]
M ask8[0] M ask8[1] M ask8[2] M ask8[3] M ask8[4] M ask8[5] M ask8[6] M ask8[7] M ask8[8] M ask8[9] M ask8[10] M ask8[11]
M ask9[0] M ask9[1] M ask9[2] M ask9[3] M ask9[4] M ask9[5] M ask9[6] M ask9[7] M ask9[8] M ask9[9] M ask9[10] M ask9[11]
M ask10[0] M ask10[1] M ask10[2] M ask10[3] M ask10[4] M ask10[5] M ask10[6] M ask10[7] M ask10[8] M ask10[9] M ask10[10]M ask10[11]
M ask11[0] M ask11[1] M ask11[2] M ask11[3] M ask11[4] M ask11[5] M ask11[6] M ask11[7] M ask11[8] M ask11[9] M ask11[10]M ask11[11]
M ask12[0] M ask12[1] M ask12[2] M ask12[3] M ask12[4] M ask12[5] M ask12[6] M ask12[7] M ask12[8] M ask12[9] M ask12[10]M ask12[11]
Output Format
The TW280X supports three 8bit output formats, ITU-R BT.656, 8bit ITU-R BT.601 and Dual
ITU-R BT.656 with 54MHz data format. The output data is synchronous with rising or falling
edge of CLK27O for ITU-R BT.656 and 8bit ITU-R BT.601 format and with rising edge of
CLK54I for Dual ITU-R BT.656 with 54MHz format. The polarity of CLK27O is controlled by the
CK27O_POL register (0x3B). For Dual ITU-R BT.656 with 54MHz format, two kinds of scaled
image are time-multiplexed with 54MHz. The output formats are selected by the OUT_FMT
register (0x22, 0x62, 0xA2, 0xE2).
CLK27O
VD[7:0] FFh 00h 00h XY 00h 00h 00h 00h FFh 00h 00h XY 00h 00h Cb0 Y0 Cr0 Y1 00h 00h
HACIVE
VALID
Analog
Input
Digital
Output
HS
VS
FLD
60Hz ODD Field
Analog
Input
Digital
Output
HS
VS VSMODE = 0
VSMODE = 1
FLD VSMODE = 0
VSMODE = 1
60Hz EVEN Field
Analog
Input
Digital
Output
HS
VS
FLD
50Hz ODD Field
Analog
Input
Digital
Output
HS
VS VSMODE = 0
VSMODE = 1
FLD VSMODE = 0
VSMODE = 1
50Hz EVEN Field
HS
Tim ing 2
Tim ing 1 : 40 system clo ck(54M H z) fo r the E ven field w ith VS M O D E = 1 o r O d d field
Tim ing 2 : 1760 system clo ck(54M H z) fo r the E ven field w ith VS M O D E = 0
CLK27O
CLK54I
VD[7:0] FFh FFh 00h 00h 00h 00h XY XY 00h 00h 00h 00h FFh FFh 00h 00h 00h 00h XY XY 00h Cb0 00h Y0 Cb0 Cr0 Y0 Y1 Cr0 Cb2 Y1 Y2 00h Cr2 00h Y3
VALID
Host Interface
The TW280X provides I2C serial and parallel interfaces that can be selected by HSPB pin.
When HSPB is low, the parallel interface is selected, the serial interface for high. Some of the
interface pins serve a dual purpose depending on the working mode. The pins HALE and
HDAT[7] in parallel mode become SCLK and SDAT pins in serial mode respectively. Each
interface protocol is shown in the following figure.
Serial Interface
HDAT[6:1] and HCSB pins define slave address. Therefore, any slave address can be assigned
for full flexibility. TW2804 also supports auto index increments in write/read mode if the data are
in sequential order.
Start Slave address R/WB Ack Index Ack Data Ack Stop
SCLK
Start Slave address R/WB Ack Index Ack Stop Start Slave address R/WB Ack Data NoAck Stop
“0” “1”
SCLK
Parallel Interface
The following figures show the write/read timing chart of parallel interface. The parallel interface
supports auto index increment after each byte of data is sent with WENB. Therefore, the host
can write multiple bytes to the slave without additional address if they are in sequential order.
The host completes the transfer cycle with CSB which is Low to High transition. Auto index
increment is also supported in read mode.
CSB
tsu(1) th(1)
WENB
tw
RENB
tw
AEN
PDATA
CSB
tsu(1) th(1)
WENB
tw
RENB
tw
AEN
PDATA
Interrupt Interface
The TW280X provides the interrupt request function via an IRQ pin. Any video loss detection or
motion detection will make the IRQ pin high until cleared via register IRQCLR (0x39) by the
host. The host processor will read the interrupt status register DET_NVMD (0x38) to find out
which channel has sensed motion or video loss. Writing high to the corresponding bit of the
interrupt clear register IRQCLR (0x39) will clear the interrupt request. Each interrupt status bit
also has its mask bit (0x3A) to disable the interrupt for that function. This sequence is described
in Fig 20.
The TW280X also provides the video loss detection or motion detection flag of individual
channel via NVMD pins. Four NVMD pins have respective channel information of motion or
video loss so that host takes status information directly by reading these pins. Its mode is
controlled by NVMDB (0x3B) that is set “1” for video loss flag and “0” for motion detection flag.
N o V id eo D etectio n
M o tio n D etectio n
M o tio n D etectio n
o n C hannel 4
o n C hannel 3
o n C hannel 2
Control Register
Register Map
Address
Mnemonic BIT7 BIT6 BIT5 BIT4 BIT3 BIT2 BIT1 BIT0
CH1 CH2 CH3 CH4
0x00 0x40 0x80 0xC0 VIDSTAT * DET_FORMAT DET_COLOR LOCK_COLOR LOCK_GAIN LOCK_OFFSET LOCK_HPLL
0x01 0x41 0x81 0xC1 FORMAT IFMTMAN IFORMAT 0 1 DET_NONSTD * DET_FLD60 *
0x02 0x42 0x82 0xC2 AGC_PLL AGC PEDEST 0 GNTIME OSTIME
0x03 0x43 0x83 0xC3 HDELAY_X HDELAY_X [7:0]
0x04 0x44 0x84 0xC4 HACTIVE_X HACTIVE_X [7:0]
0x05 0x45 0x85 0xC5 HDELAY_Y HDELAY_Y [7:0]
0x06 0x46 0x86 0xC6 HACTIVE_Y HACTIVE_Y [7:0]
0x07 0x47 0x87 0xC7 MSB_ACTV HACTIVE_Y [9:8] HDELAY_Y [9:8] HACTIVE_X [9:8] HDELAY_X [9:8]
0x08 0x48 0x88 0xC8 HSWIDTH 0 HSWIDTH
0x09 0x49 0x89 0xC9 VDELAY_X VDELAY_X [7:0]
0x0A 0x4A 0x8A 0xCA VACTIVE_X VACTIVE_X [7:0]
0x0B 0x4B 0x8B 0xCB VDELAY_Y VDELAY_Y [7:0]
0x0C 0x4C 0x8C 0xCC VACTIVE_Y VACTIVE_Y [7:0]
0x0D 0x4D 0x8D 0xCD HPLL HPLLMAN HPLLTIME VACTVE_Y [8] VDELAY_Y [8] VACTVE_X [8] VDELAY_X [8]
0x0E 0x4E 0x8E 0xCE SYNCPOL FLDMODE VSMODE FLDPOL HSPOL VSPOL 1 0
0x0F 0x4F 0x8F 0xCF HUE HUE
0x10 0x50 0x90 0xD0 SAT SAT
0x11 0x51 0x91 0xD1 CONT CONT
0x12 0x52 0x92 0xD2 BRT BRT
0x13 0x53 0x93 0xD3 CFILTER IFCOMP CLPF ACCMODE APCMODE
0x14 0x54 0x94 0xD4 PEAKCKIL YPEAK_Y YPEAK_X 0 CKILL
0x15 0x55 0x95 0xD5 SCLFLT VLPF_Y VLPF_X HLPF_Y HLPF_X
0x16 0x56 0x96 0xD6 TRAP_X YBWI_X COMBMD_X 0
0x17 0x57 0x97 0xD7 TRAP_Y YBWI_Y COMBMD_Y 0
0x18 0x58 0x98 0xD8 VSCLMSB_X VSCALE_X [15:8]
0x19 0x59 0x99 0xD9 VSCLLSB_X VSCALE_X [7:0]
0x1A 0x5A 0x9A 0xDA VSCLMSB_Y VSCALE_Y [15:8]
0x1B 0x5B 0x9B 0xDB VSCLLSB_Y VSCALE_Y [7:0]
0x1C 0x5C 0x9C 0xDC HSCLMSB_X HSCALE_X [15:8]
0x1D 0x5D 0x9D 0xDD HSCLLSB_X HSCALE_X [7:0]
0x1E 0x5E 0x9E 0xDE HSCLMSB_Y HSCALE_Y [15:8]
0x1F 0x5F 0x9F 0xDF HSCLLSB_Y HSCALE_Y [7:0]
0x20 0x60 0xA0 0xE0 VSCLCON_X 0 VFLT_MD_X VBW_X PALDLY_X ODD_EN_X EVEN_EN_X 1
0x21 0x61 0xA1 0xE1 VSCLCON_Y 0 VFLT_MD_Y VBW_Y PALDLY_Y ODD_EN_Y EVEN_EN_Y 1
0x22 0x62 0xA2 0xE2 OUTFMT BGND_EN BGND_COLR NOVID_656 LIM_16 SW_RESET ANA_SW OUT_FMT
0x23 0x63 0xA3 0xE3 RESERVED 1 0 0 1 0 0 0 1
0x24 0x64 0xA4 0xE4 SENSCTL LVLSENS TMPSENS SPTSENS
0x25 0x65 0xA5 0xE5 MPERIOD 0 MDPERIOD
0x26 0x66 0xA6 0xE6 MDMASK1 MDMASK1[7:0]
Register Description
IFORMAT Force the device to operate in a particular video standard when IFMTMAN
is high or to free-run in a particular video standard on no-video status when
IFMTMAN is low
0 PAL-B/D (default)
1 PAL-M
2 PAL-N
3 PAL-60
4 NTSC-M
5 NTSC-4.43
6 NTSC-N
HDELAY This 10-bit register defines the starting location of horizontal active pixel. A
unit is 1 pixel. HDELAY1 and HDELAY2 define the different starting
location of horizontal active pixel for dual scaler output. The default value is
decimal 32.
HACTIVE This 10-bit register defines the number of horizontal active pixel. A unit is 1
pixel. HACTIVE1 and HACTIVE2 define the different number of horizontal
active pixels for dual scaler output. The default value is decimal 720.
HSWIDTH This 6bit register defines the width of horizontal sync output. A unit is 1
pixel. The default value is decimal 32
VDELAY This 9bit register defines the starting location of vertical active. A unit is 1
line. VDELAY1 and VDELAY2 define the different starting location of
vertical active line for dual scaler output. The default value is decimal 6.
VACTIVE This 9bit register defines the number of vertical active lines. A unit is 1 line.
VACTIVE1 and VACTIVE2 define the different number of vertical active
lines for dual scaler output. The default value is decimal 240.
HPLLTIME Control the time constant of horizontal PLL when HPLLMAN is high
0 Slow
: :
4 Typical (default)
: :
7 Fast
Hue Control
CH Index
[7] [6] [5] [4] [3] [2] [1] [0]
1 0x0F
2 0x4F
HUE
3 0x8F
4 0xCF
Saturation Control
CH Index
[7] [6] [5] [4] [3] [2] [1] [0]
1 0x10
2 0x50
SAT
3 0x90
4 0xD0
Contrast Control
CH Index
[7] [6] [5] [4] [3] [2] [1] [0]
1 0x11
2 0x51
CONT
3 0x91
4 0xD1
Brightness Control
CH Index
[7] [6] [5] [4] [3] [2] [1] [0]
1 0x12
2 0x52
BRT
3 0x92
4 0xD2
VSCALE The 16bit register defines a vertical scaling ratio. The actual vertical scaling
ratio is VSCALE[15:0] / (2^16 – 1). VSCALE1 and VSCALE2 define the
different vertical scaling ratio for dual scaler. The default value is 16 bit
0xFFFF.
HSCALE The 16-bit register defines a horizontal scaling ratio. The actual horizontal
scaling ratio is HSCALE[15:0] / (2^16 – 1). HSCALE1 and HSCALE2 define
the different horizontal scaling ratio for dual scaler. The default value is 16
bit 0xFFFF.
Output Formatter
CH Index
[7] [6] [5] [4] [3] [2] [1] [0]
1 0x22
2 0x62 SW_
BGNDEN BGNDCLR NOVID_656 LIM_16 ANA_SW OUT_FMT
3 0xA2 RESET
4 0xE2
BGNDCLR Select the background color mode only if BGNDEN bit is high
0 Blue color mode (default)
1 Black color mode
NOVID_656 Select the optional set of 656 SAV/EAV code sequence for no-video status
0 Normal 656 SAV/EAV code sequence (default)
1 An optional set of 656 SAV/EAV code sequence for no-video status
Reserved
CH Index
[7] [6] [5] [4] [3] [2] [1] [0]
1 0x23
2 0x63
1 0 0 1 0 0 0 1
3 0xA3
4 0xE3
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
MDMASK1~12 Select mask area of motion detector. An active region is divided into 12x12
mask areas as illustrated in Fig. 11. If the mask bit in specific area is
programmed into high, the specific area is ignored in operation of motion
detector. But for proper operation, more than 4 mask areas should be
enabled in any case. (default : 0x00)
IRQCLR Setting high to bits clears interrupt requests of corresponding bits. This bit
is self-clearing in a few clocks after setting high (default : 0x00)
IRQENA Enable the corresponding (0x38, 0x39) interrupt register bit (default : 0x00)
U Gain
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x3C U_GAIN[7:0]
U_GAIN Adjust gain for U (or Cb) component. The resolution is 0.8% / LSB.
0 0%
: :
128 100 % (default)
: :
255 200 %
V Gain
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x3D V_GAIN[7:0]
V_GAIN Adjust gain for V (or Cr) component. The resolution is 0.8% / LSB.
0 0%
: :
128 100 % (default)
: :
255 200 %
U Offset
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x3E U_OFF[7:0]
U_OFF U (or Cb) offset adjustment register. The resolution is 0.4% / LSB.
0 -50 %
: :
128 0 % (default)
: :
255 50 %
V Offset
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x3F V_OFF[7:0]
V_OFF V (or Cr) offset adjustment register. The resolution is 0.4% / LSB.
0 -50 %
: :
128 0 % (default)
: :
255 50 %
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x79 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x7A 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
FLD_OFST_4Y Remove the field offset between ODD and EVEN for Y path of Channel 4
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_4X Remove the field offset between ODD and EVEN for X path of Channel 4
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_3Y Remove the field offset between ODD and EVEN for Y path of Channel 3
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_3X Remove the field offset between ODD and EVEN for X path of Channel 3
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_2Y Remove the field offset between ODD and EVEN for Y path of Channel 2
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_2X Remove the field offset between ODD and EVEN for X path of Channel 2
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_1Y Remove the field offset between ODD and EVEN for Y path of Channel 1
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
FLD_OFST_1X Remove the field offset between ODD and EVEN for X path of Channel 1
0 Normal operation (default)
1 Remove the field offset between ODD and EVEN field
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x7C 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0x7D 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0xB8 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
FLD_656 Control the field polarity mode in ITU-R 656 timing codes
0 Fixed field polarity according to ITU-R 656 format (default)
1 Controllable field polarity by FLDPOL register (0x0E,0x4E,0x8E,0xCE)
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0xFA 0 0 1 1 1 1 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0xFB 0 0 0 1 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0xFC 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Reserved
Index
[7] [6] [5] [4] [3] [2] [1] [0]
0xFD 0 0 0 0 0 0 0 0
This control register is reserved for putting the part into test mode. For normal operation, the
above value should be set in this register.
Parametric Information
DC Electrical Parameters
AC Electrical Parameters
Table 11 Clock Timing Parameters
Parameter Symbol Min Typ Max Units
Delay from CLK54I to CLK27O 1 5 12 ns
Hold from CLK27O to Data 2a 16 ns
Delay from CLK27O to Data 2b 19 ns
Hold from CLK54I to Data 3a 5 ns
Delay from CLK54I to Data 3b 12 ns
CLK54I
1
CLK27O
2b
2a
Data Output
(27Mhz)
3b
3a
Data Output
(54Mhz)
Package Dimension
Application Information
Clamping / AGC
TW2804 has built-in clamping and AGC circuitry. The analog inputs must be AC coupled
through an external 2.2uF capacitor. Without it, no extra external component is needed for
this operation. The clamping and AGC tracking time constant can be controlled through
register setting.
Power-Up
After power-up, TW2804 registers have unknown values. The RSTB pin must be asserted
and released to bring all registers to its default values. After reset, TW2804 data outputs are
tri-stated. The OE (0x3B) register should be written after reset to enable outputs desired.
Application Schematic
J100
RCA JACK R101 L100 C102
CH1A 1
9
21
33
48
60
72
90
102
6
24
45
66
84
105
111
112
119
120
127
10uH U100
270 2.2uF
2
R100 C100 C101 R102
VDD
VDD
VDD
VDD
VDD
VDD
VDD
VDD
VDDA
VDDA
VDDA
VDDA
VDDO
VDDO
VDDO
VDDO
VDDO
VDDO
VDDAD
75 47pF 100pF 4.7k
VD1[7:0]
VD1_7 VD1[7:0]
13
VD1_7 14 VD1_6
VD1_6 16 VD1_5
VD1_5 17 VD1_4
113 VD1_4 19 VD1_3
VIN1A VD1_3 20 VD1_2
VD1_2 22 VD1_1
114 VD1_1 23 VD1_0
J101 VIN1B VD1_0
RCA JACK R104 L101 C105 11
CH2A VALID1 VALID1
1 8
HS1 HS1
10uH C112 7
VS1 VS1
270 2.2uF 5
FLD1 FLD1
2
R103 C103 C104 R105 0.1uF 10
ACTIVE1 ACTIVE1
75 47pF 100pF 4.7k
VD2[7:0]
VD2_7 VD2[7:0]
34
VD2_7 35 VD2_6
VD2_6 37 VD2_5
117 VD2_5 38 VD2_4
VIN2A VD2_4 40 VD2_3
VD2_3 41 VD2_2
118 VD2_2 43 VD2_1
VIN2B VD2_1 44 VD2_0
J102 VD2_0
RCA JACK R107 L102 C108 32
CH3A VALID2 VALID2
1 C113 29
HS2 HS2
10uH 28
VS2 VS2
270 2.2uF 0.1uF 26
FLD2 FLD2
2
R106 C106 C107 R108 31
ACTIVE2 ACTIVE2
75 47pF 100pF 4.7k TW2804 VD3[7:0]
VD3_7 VD3[7:0]
55
VD3_7 56 VD3_6
121 128QFP VD3_6 58 VD3_5
VIN3A VD3_5 59 VD3_4
VD3_4 61 VD3_3
122 VD3_3 62 VD3_2
VIN3B VD3_2 64 VD3_1
VD3_1 65 VD3_0
J103 VD3_0
RCA JACK R110 L103 C111 C114 53
CH4A VALID3 VALID3
1 50
HS3 HS3
10uH 0.1uF 49
VS3 VS3
270 2.2uF 47
FLD3 FLD3
2
R109 C109 C110 R111 52
ACTIVE3 ACTIVE3
75 47pF 100pF 4.7k
VD4[7:0]
VD4_7 VD4[7:0]
76
125 VD4_7 77 VD4_6
VIN4A VD4_6 79 VD4_5
VD4_5 80 VD4_4
126 VD4_4 82 VD4_3
VIN4B VD4_3 83 VD4_2
VD4_2 85 VD4_1
C115 VD4_1 86 VD4_0
VD4_0 74
VALID4 VALID4
0.1uF 71
HS4 HS4
70
VS4 VS4
68
FLD4 FLD4
73
ACTIVE4 ACTIVE4
VDD2.5V VDD
L104
88
CLK27O CLK27
BEAD + C116 C117 C118 C119 C120 C121 C122 C123 C124
RESETB
47uF/16V 0.1uF 0.1uF 0.1uF 0.1uF 0.1uF 0.1uF 0.1uF 0.1uF 103
IRQ IRQ
4
NVMD1 NVMD1
25
NVMD2 NVMD2
46
NVMD3 NVMD3
VDDA 2 67
L105 RSTB NVMD4 HDAT[7:0] NVMD4
OSC100 HDAT0 HDAT[7:0]
VDDO 101
HDAT_0 100 HDAT1 VDDO
1 8 HDAT_1 98 HDAT2
BEAD + C125 NC VCC HDAT_2 HDAT3
C126 C127 C128 C129 C130 97
DIP8 HDAT_3 95 HDAT4
47uF/16V 0.1uF 0.1uF 0.1uF 0.1uF 0.1uF HDAT_4 94 HDAT5 R113
R112
4 5 89 HDTV_5 92 HDAT6 47K
GND OUT CLK54I HDAT_6 91 HDAT7
To Micom
1 HDAT_7 104
100 HWRB
TEST HWRB 106
OSC 54MHz HRDB HRDB
VDD3.3V VDDO 107
L106 HALE HALE
109
HCSB HCSB
110
HSPB HSPB
Revision History
(1) Update Control Register Map (Read only description is added) E BAHB
1.1 Jan / 29 / 2003
(P.29~30, P.33~34, P.58) (Eng RevB)