FAIRCHILD TMC22091KHC

www.fairchildsemi.com
TMC22091/TMC22191
Digital Video Encoders/Layering Engine
Features
•
•
•
•
•
•
•
• All-digital video encoding
• Internal digital oscillators, no crystals required
• Multiple input formats supported
– 24-bit and 15-bit GBR/RGB
– YCBCR422 or 444
– Color indexed
• 30 overlay colors (TMC22191)
• Fully programmable timing
• Supports input pixel rates of 10 to 15 Mpps
• 256 x 8 x 3 color look-up tables (bypassable on
TMC22191)
• 8-bit mask register
• 8-bit composite digital video input
• Hardware and 24-bit data keying
• Synchronizes with TMC22071 Genlocking Video
Digitizer
• 8:8:8 video reconstruction
• SMPTE 170M NTSC or CCIR Report 624 PAL
compatible
• Supports PAL-M and NTSC without pedestal
• Simultaneous S-VIDEO (Y/C) NTSC/PAL output
• 10-bit D/A conversion (three channels)
Description
The TMC22x91 digital video encoders convert digital computer image or graphics data (in RGB, YCBCR, or color
indexed format) or a CCIR-601 signal into a standard analog
baseband television (NTSC or PAL) signal with a modulated
color subcarrier.
Both composite (single lead) and S-VIDEO (separate
chroma and luma) formats are active simultaneously at the
three analog output pins, each of which generates a standard
video-level signal into doubly-terminated 75Ω load.
The TMC22x91 accepts digitized video from the companion
TMC22071 Genlocking Video Digitizer. Soft switching
between video sources is done under either hardware or
programmable data control.
The TMC22191 offers 4-layer keying capability, bypassable
CLUT, and 30 Overlay colors.
Logic Symbol
BYPASS and OL4:0 on TMC22191 only.
GENLOCK
INTERFACE
CVBS7-0
8
GHSYNC
GVSYNC
RESET
D7-0
A1-0
TMC22x91
DIGITAL
VIDEO
ENCODER
VREF
COMP
RREF
TDI
TMS
TCK
TDO
The TMC22x91 is fabricated in a submicron CMOS process
and packaged in an 84 Lead Plastic Leadless Chip Carrier, or
in a 100 Lead Metric Quad Flat Pack. Performance is guaranteed from 0°C to 70°C.
8
2
LDV
PXCK
CS
R/W
CLOCKS
MICROPROCESSOR
INTERFACE
PDC
VHSYNC
VVSYNC
KEY
BYPASS
CHROMA
LUNA
COMPOSITE
JTAG TEST
INTERFACE
6
ANALOG
INTERFACE
OL4-0
24
VIDEO
OUTPUT
FRAME BUFFER
INTERFACE
PD23-0
Controlled edge rates
3 power-down modes
Built-in color bars and modulated ramp test signals
JTAG (IEEE Std 1149.1-1990) test interface
Single +5V power supply
84 lead PLCC package
100 lead MQFP package
27008A
Rev. 1.1.0
TMC22091/TMC22191
PRODUCT SPECIFICATION
Block Diagram
BYPASS and OL4:0 on TMC22191 only.
PD23-0
R/R-Y
256 x 8 x 3
COLOR
B/B-Y
LOOK-UP
TABLE
G/Y
FORMATTER B/G/CB
MASK, KEY
COMPARATOR R/B/CR
MATRIX
INT
CHROMA
MODULATOR
B-Y
LPF
Data Key
DIGITAL
SYNC.
GEN.
GVSYNC
10-bit
D/A
CHROMA
10-bit
D/A
LUMA
10-bit
D/A
COMPOSITE
INT
SYNC,
BLANK
INSERT
KEY
PDC
VVSYNC
VHSYNC
GHSYNC
GENLOCK
INTERFACE
LPF
R-Y
INTERPOLATOR
4:2:2/4:4:4
ENCODED VIDEO OUTPUT
G/R/Y MAP
VIDEO
SWITCH
SUBCARRIER
SYNTHESIZER
INT
CVBS7-0
CLOCK
CONTROL
JTAG
INTERPOLATION
FILTERS
D/A
REF.
TDI
TMS
TCK
TDO
CLOCKS
RESET
D7:0
A1:0
CS
R/W
LDV
PXCK
VREF
COMP
RREF
MICROPROCESSOR
INTERFACE
JTAG TEST
INTERFACE
ANALOG
INTERFACE
FRAME BUFFER INTERFACE
OL4-0
BYPASS
27006A
Functional Description
Timing
The TMC22091 and TMC22191 are totally integrated, fullyprogrammable digital video encoders with simultaneous
composite and Y/C (S-VIDEO) outputs. The TMC22x91
video outputs are compatible with SMPTE 170M NTSC,
CCIR Report 624 PAL, PAL-M, and NTSC without pedestal
television standards. No external component selection or
tuning is required.
The encoder operates from a single clock at twice the system
pixel rate. This frequency may be set between 20 MHz and
36 MHz (pixel rates of 10 Mpps to 18 Mpps). Within this
range are included CCIR-601, D2, and square-pixel formats,
as well as a variety of computer-specific pixel rates. An array
of programmable timing registers allows the software selection of all pertinent signal parameters to produce NTSC
(with or without 7.5 IRE pedestal) and PAL, and PAL-M
outputs.
The encoders accept digital image data at the PD port in one
of several formats, which are matrixed into luminance and
chrominance components. The chrominance signals are
modulated onto a digitally synthesized subcarrier. The luminance and chrominance signals are separately interpolated to
twice the pixel rate, and converted to analog levels by 10-bit
D/A converters. They are also digitally combined and the
resulting composite signal is output by a third 10-bit D/A
converter. This composite signal may be keyed (pixel rate
switching) with a second composite digital video signal presented to the encoder.
The output video frames may be internally timed by the
TMC22x91, synchronized with the external frame buffer, or
slaved to the companion Genlocking Video Digitizer
(TMC22071). All operational parameters are fully programmable over a standard microprocessor port.
Table 1 shows the key features that distinguish between the
TMC22091 and TMC22191. All of the information presented in this data sheet applies to both products unless otherwise noted. Statements, paragraphs, tables, and figures that
apply to only one or two of the encoders have notation specifying the applicable part number.
2
Table 1. Comparing the TMC22x91 Encoders
Feature
TMC22091 TMC22191
OL4-0 pixel inputs for 30
overlay colors
No
Yes
Number of video layers
supported
2
4
No
Yes
BYPASS input for
bypassing CLUTs
Input Formatting
The input section accepts a variety of video and graphics formats, including 24-bit GBR and RGB, 15-bit GBR and
RGB, YCBCR422, YCBCR444, and 8-bit color-indexed data
(Figure 1a and 1b).
The input section of the TMC22x91 includes a key comparator which monitors the pixel data port with three independent
8-bit comparators, and invokes a video key when the selected
registers match the incoming data.
PRODUCT SPECIFICATION
TMC22091/TMC22191
Mask Register
Colorspace Conversion Matrix and
Interpolator
A Mask Register is provided which is logically ANDed with
incoming color-index data to facilitate pixel animation and
other special graphics effects. The Mask Register is ahead of
the Data Key comparators and is enabled only when colorindex input is selected. Mask Register programming and
operation are similar to that of the 171/176 family of graphics RAMDACS.
The matrix converts RGB data (whether from RGB inputs or
color-indexed CLUT data) into Y, B-Y, R-Y format for
encoding. In input configurations where the pixel input is
already in Y, B-Y, R-Y format, the matrix is bypassed. When
pixel data is input in YCBCR422 format, the interpolation filters produce YCBCR444 for encoding.
Color Lookup Table
Sync Generator
The Color Lookup Table (CLUT) is a 256 x 8 x 3 randomaccess memory. It provides means for offset, gain, gamma,
and color correction in RGB and YCBCR operating modes. It
provides a full 24-bit color lookup function for color-index
mode. It can be loaded in the same manner as a standard
VGA RAMDAC.
The TMC22x91 can operate in Master, Genlock, or Slave
modes. In Master and Genlock modes, the encoder internally
generates all timing and sync signals, and provides Horizontal Sync, Vertical Sync, and Pixel Data Control (PDC) to the
external frame buffer circuitry. PDC is independently selectable to function as an input or an output. In Genlock mode,
the TMC22x91 timing is controlled by the TMC22071 Genlocking Video Digitizer over the CVBS7-0 bus, GVSYNC,
and GHSYNC. The encoder, in turn, produces VHSYNC,
VVSYNC, and PDC for the frame buffer interface.
MODE
GBR444
RGB444
YCBCR444
YCBCR422
MSB
23
16 15
8
G7
G
G0 B7
B
B 0 R7
R
R0
R7
R
R0 G7
G
G0 B 7
B
B0
Y7
Y
Y0 CB7
CB
CB0 RR7
CR
RR0
Y7
Y
CB7
Y0 CR7
CB/CR
CB0
CR0
COLOR INDEX
GBR15
RGB15
LSB
0
7
Format Control Register
MSB
LSB
00011000
00010000
00011100
00011101
Pixel
P7
P0
G4
G
G0 B 4
B
B0 R4
R
R0
R4
R
R 0 G4
G
G 0 B4
B
B0
0001X011
00011010
00010010
24300A
Figure 1a. Pixel Data Format
3
TMC22091/TMC22191
MODE
GBR444
RGB444
YCBCR444
YCBCR422
COLOR INDEX
RGB15
GBR15
PRODUCT SPECIFICATION
MSB
23
16 15
8
LSB
0
7
G7
G
G0 B7
B
B 0 R7
R
R0
R7
R
R0 G 7
G
G0 B7
B
B0
Y7
Y
Y0 CB7
CB
CB0 CR7
CR
CR0
Y7
Y
Y0 CB7
CB*
CB0 CR7
CR*
CR0
P7
Pixel
P0 P7
Pixel
P0 P7
Pixel
P0
R4
R
R0
G4
G
G0
B4
B
B0
G4
G
G0
B4
B
B0
R4
R
R0
Format Control Register
MSB
LSB
01011000
01010000
0101X000
0101X001
0101X011
01010010
01011010
24393A
*CB and CR are loaded on alternate LDV cycles
Figure 1b. Pixel Data Format (TMC22191 when CLUTs are Bypassed)
In Slave mode, VHSYNC, VVSYNC, and PDC (optional)
are inputs to the TMC22x91. These inputs determine when
new lines, frames, and active picture areas begin. The external controlling circuitry needs to establish the correct timing
for these signals.
Horizontal and vertical synchronization signals are digitally
generated by the TMC22x91 with controlled rise and fall
times on all sync edges, the beginning and end of active
video, and the burst envelope. All elements of horizontal
sync timing are programmable, as are the frequency, phase,
and duration of color burst.
Video Input
The TMC22x91 accepts genlocked synchronization data and
digital composite video signals from the TMC22071 Genlocking Video Digitizer over the 8-bit CVBS bus. The
encoder synchronizes its digital subcarrier oscillator to the
video input from the TMC22071 with this data. The composite video data output from the TMC22071 is passed to the
internal video switch for keying with the encoded pixel data.
Chroma Modulator
A 32-bit digital subcarrier synthesizer feeds a quadrature
modulator, producing a digital chrominance signal. The relative phases of the burst and active video portions of the subcarrier can be individually adjusted to compensate for
external phase errors and to effect a hue control.
Interpolation Filters
Interpolation filters on the luminance and chrominance signals double the pixel rate in preparation for D/A conversion.
This band-limited process greatly simplifies the output filtering required following the D/A converters and dramatically
reduces sin(x)/x distortion.
4
An interpolation filter on the CVBS data similarly raises the
sample rate of the video signal, for mixing with the encoded
pixel data.
Composite Video Switch
The Composite Video Switch selects between the composite
video input (CVBS) and the composite encoded pixel data
on a pixel-by-pixel basis, under the control of a key function.
Keying may be managed by hardware or software. The hardware key input (KEY pin) directly controls the video switch.
The encoder may be programmed to operate with a data key,
represented by three 8-bit registers that compare with the 24
input bits. They operate in all input modes and may be individually enabled or disabled.
D/A Converters
The analog outputs of the TMC22x91 are the outputs of
three 10-bit D/A converters, operating at twice the pixel
clock rate. The outputs are capable of driving standard video
levels into a doubly-terminated 75Ω coaxial video cable
(37.5Ω total load). An internal voltage reference is provided
which can be used to provide reference current for the three
D/A converters. For accurate video levels, an external fixed
or variable voltage reference source is recommended. The
video signal levels from the TMC22x91 may be adjusted to
overcome the insertion loss of analog low-pass output filters.
The D/A converters on the TMC22x91 may be powereddown via Control Register 0E bits 5 and 6. The
COMPOSITE D/A is controlled by bit 6 and the LUMA and
CHROMA D/A converters are controlled by bit 5.
PRODUCT SPECIFICATION
Microprocessor Interface
The microprocessor interface employs a 13 line format. The
RESET pin sets all internal state machines to their initialized
conditions, disables the analog outputs, sets the internal
SRESET bit LOW (reset condition), and places the encoder
in a power-down mode. All register and CLUT data are
maintained in power-down mode. If the HRESET bit is set
HIGH, line 1 field 1 is started when RESET goes HIGH, and
SRESET is ignored. If HRESET is LOW, the encoder
remains idle after RESET goes HIGH until Control Register
bit SRESET is set HIGH, which initiates line 1 field 1.
Two address lines are provided and decoded for access to the
internal Control Registers and CLUT. Control Registers and
CLUT are accessed by loading a desired address through the
8-bit D7-0 port, followed by the desired data read or write for
that address. Both the CLUT and the Control Registers are
self-indexing, allowing continuous reads or writes to successive addresses.
TMC22091/TMC22191
bars are useful as an idle system output signal. The test signals may be used to verify proper operation of the analog
video signal chain.
TMC22090/TMC22190 Compatibility
The TMC22090 and TMC22190 are earlier versions of the
TMC22091 and TMC22191, respectively. They lack the following features of the newer versions:
1.
Selectable Setup (to support NTSC EIA-J video output
for Japan)
2.
PAL-M format (for South American applications)
3.
Extended EH and SL intervals (to support pixel rates
above 15 Mpps)
4.
Individual D/A power-down (to reduce total dissipation
when some outputs are not required)
5.
Luminance I/O processing (to reduce flicker in graphics
applications)
JTAG Test Interface
The TMC22x91 includes a standard 4-line JTAG (IEEE Std
1149.1-1990) test interface port, providing access to all digital input/output data pins. This is provided to facilitate component and board-level testing.
Test/Validation Mode
The TMC22x91 may be configured to produce standard
color bars or a 40 IRE modulated (or unmodulated) video
ramp, independent of any pixel or video data input. Color
These features are controlled by registers 0E and 0F, and
enabled by setting Register OE bit 7 to ONE. If an application of the TMC22x90 is programmed with this bit set to
ZERO (as recommended in the product documentation) then
the corresponding TMC22x91 will perform identically.
Though the earlier parts continue to be available, it is recommended that the newer devices be used in new designs for
the additional flexibility. Older designs may be readily
converted to the newer versions to take advantage of the
added features and lower cost of the later technology.
5
TMC22091/TMC22191
PRODUCT SPECIFICATION
Pin Assignments
84 Lead PLCC
1 84
65-3751-01
Pin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Name
CVBS2
CVBS1
CVBS0
KEY
RESET
CS
R/W
A1
A0
DGND
PDC
VHSYNC
VVSYNC
D7
D6
D5
D4
D3
D2
D1
D0
Pin
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Name
TDO
TCK
TMS
TDI
DGND
VDD
BYPASS (TEST)
OL4 (TEST)
VREF
RREF
AGND
COMPOSITE
AGND
LUMA
AGND
CHROMA
AGND
COMP
VDDA
VDDA
VDDA
Pin
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Name
VDDA
CVBS7
CVBS6
CVBS5
CVBS4
OL3 (TEST)
OL2 (TEST)
OL1 (TEST)
OL0 (TEST)
PD23
PD22
PD21
PD20
PD19
PD18
PD17
PD16
PD15
PD14
PD13
PD12
Pin
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
Name
VDD
DGND
PD11
PD10
PD9
PD8
PD7
PD6
PD5
PD4
PD3
PD2
PD1
PD0
LDV
PXCK
DGND
VDD
GVSYNC
GHSYNC
CVBS3
Note: Pin names in parentheses apply to TMC22091.
100 Lead MQFP
100
1
65-3751-02
Pin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Name
NC
COMPOSITE
NC
AGND
LUMA
AGND
NC
CHROMA
AGND
COMP
NC
NC
VDDA
VDDA
VDDA
VDDA
VDDA
CVBS7
CVBS6
CVBS5
CVBS4
OL3 (TEST)
OL2 (TEST)
OL1 (TEST)
OL0 (TEST)
Pin
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Name
PD23
PD22
NC
NC
NC
PD21
PD20
PD19
PD18
PD17
PD16
PD15
PD14
PD13
PD12
VDD
DGND
PD11
PD10
PD9
PD8
PD7
PD6
PD5
PD4
Pin
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Name
PD3
NC
NC
NC
NC
PD2
PD1
PD0
LDV
PXCK
DGND
VDD
GVSYNC
GHSYNC
CVBS3
CVBS2
CVBS1
CVBS0
NC
KEY
RESET
CS
R/W
A1
A0
Pin
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Note: Pin names in parentheses apply to TMC22091.
6
Name
DGND
PDC
NC
NC
VHSYNC
VVSYNC
D7
D6
D5
D4
D3
D2
D1
D0
TDO
TCK
TMS
TDI
DGND
VDD
BYPASS (TEST)
OL4 (TEST)
VREF
RREF
AGND
PRODUCT SPECIFICATION
TMC22091/TMC22191
Pin Descriptions
Pin Number
Pin Name
84-Lead 100-Lead
PLCC
MQFP
Value
Pin Function Description
Clocks
PXCK
79
60
TTL
Master Clock Input. This 20 to 30 MHz clock is internally
divided by 2 to generate the internal pixel clock, PCK, which a
LOW on RESET forces LOW. PXCK drives the entire
TMC22x91, except the asynchronous microprocessor interface
and the semi-synchronous LDV data input clock. All internal
registers are strobed on the rising edge of PXCK.
LDV
78
59
TTL
Pixel Data Load Clock. On each rising edge of LDV, data on
PD23-0 are latched into the input preload register, for transfer
into the input demultiplexer on the next rising edge of PCK.
52-63,
66-77
26, 27,
31-40,
43-51,
56-58
TTL
Pixel Data Inputs. In YCBCR, GBR, RGB, and color-indexed
mode, pixel data enter the TMC22x91 on PD23-0. The specific
format is found in Figures 1a and 1b. LDV is the clock that
controls the loading of pixel data.
VHSYNC
12
80
TTL
Horizontal Sync I/O. In Master and Genlock modes, the
TMC22x91 outputs horizontal sync on this pin. In Slave modes,
the TMC22x91 accepts and locks to horizontal sync input on
this pin (with vertical sync on VVSYNC). VHSYNC and
VVSYNC must be coincident since they are clocked into the
TMC22x91 on the same rising edge of PXCK.
VVSYNC
13
81
TTL
Vertical Sync I/O. In separate V and H sync Master and
Genlock modes, the TMC22x91 outputs vertical block sync
(VVSYNC LOW for the 2.5 (PAL) or 3 (NTSC) lines on which
vertical sync pulses occur). In composite sync (H and V sync
on same signal) Master and Genlock modes, the TMC22x91
outputs horizontal sync, vertical sync, and equalization over
this pin. In Slave mode, the TMC22x91 accepts and locks to
vertical sync input on this pin (with horizontal sync on
VHSYNC). VHSYNC and VVSYNC must be coincident such
that they are clocked into the TMC22x91 on the same rising
edge of PXCK.
PDC
11
77
TTL
Pixel Data Control. In Master mode, the TMC22x91 forces
PDC HIGH when and only when it wants active video from the
frame buffer. During blanking (syncs, equalization, burst, and
porches), it forces PDC LOW, signaling that it will ignore any
data presented over PD23-0. When PDC is used as an input,
forcing it HIGH allows the TMC22x91 to receive PD during the
active video state.
KEY
4
70
TTL
Hardware Key Input. When the HKEN control bit is set HIGH
and hardware key pin, KEY, is HIGH, video data entering on
CVBS7-0 are routed to the COMPOSITE output. This control
signal is pipelined so the pixel that is presented to the PD port
when the KEY signal is invoked is at the midpoint of the soft
key transition. When HKEN is LOW, KEY is ignored. Like PD
data, KEY is clocked into the TMC22x91 on the rising edge of
LDV.
Frame Buffer Interface
PD23-0
7
TMC22091/TMC22191
PRODUCT SPECIFICATION
Pin Descriptions (continued)
Pin Number
Pin Name
84-Lead 100-Lead
PLCC
MQFP
Value
OL4-0
29, 48-51 97, 22-25
TTL
Overlay Data Inputs (TMC22191 only). 30 of the 256
locations of the CLUT may be reserved for overlay operation.
These CLUT locations are directly accessed by five input pins,
OL4-0. OL4-0 are entered into the TMC22191 on a pixel-bypixel basis and select which of the 30 overlay colors is to be
encoded. When all five OL4-0 inputs are LOW, no overlay
occurs.
Pin Function Description
28
96
TTL
CLUT Bypass Control (TMC22191 only). When BYPASS is
HIGH, the CLUT is in the pixel data path within the TMC22191.
When BYPASS is LOW, pixel data bypasses the CLUT.
BYPASS is active only for certain modes of the Layering
Control Register (LCR) when the Format Control Register bit 6
is HIGH.
GHSYNC
83
64
CMOS
Genlock Horizontal Sync. In Genlock mode, the TMC22x91
will start a new horizontal line (blank-to-sync-edge transition)
with each falling edge of GHSYNC. In non-genlock modes, the
TMC22x91 ignores GHSYNC. The internal pixel clock, PCK, is
aligned with the falling edge of VHSYNC or GHSYNC (Genlock
mode).
GVSYNC
82
63
CMOS
Genlock Vertical Sync. In Genlock mode, the TMC22x91 will
start a new vertical sync sequence at line 1 field 1 whenever
GVSYNC and GHSYNC are coincident such that they are
clocked into the TMC22x91 on the same rising edge of PXCK.
If GVSYNC falls at any other time, the TMC22x91 will assume
that this marks the start of field 2, and will ignore it (in odd-field
sync mode) or (in all-field sync mode) respond by generating a
single vertical sync pulse, followed by 2 (PAL) or 2.5 (NTSC)
lines of vertical sync, keyed to the next falling edge on
GHSYNC. See Interface Control Register bit 0 for odd-field and
all-field operation.
CVBS7-0
44-47, 84,
1-3
18-21,
65-68
TTL
Composite Video Inputs. The encoder receives digitized
video, subcarrier phase, and subcarrier frequency over this 8bit bus at the PCK rate. This data may be provided by the
companion TMC22071 Genlocking Video Digitizer. In Genlock
mode, the TMC22x91 expects subcarrier phase and frequency
data during each line’s horizontal sync interval, as well as video
data when keying is engaged, transferred at the PCK rate.
BYPASS
Genlock Interface
Microprocessor Interface
8
D7-0
14-21
82-89
TTL
Data I/O Port. All control parameters are loaded into and read
back over this 8-bit port. For digital testing, the five lower bits
can also serve as a two-cycle 10-bit data output port. For D/A
converter testing, it can be used as a 10-bit two-cycle input
port, facilitating, for example, ramp-based D/A converter
linearity tests.
A1-0
8-9
74-75
TTL
µProc Port Controls. As in a RAMDAC, this control governs
whether the microprocessor interface selects a table address
or reads/writes table contents. It also governs setting and
verification of the TMC22x91’s internal operating modes, also
over port D7-0.
PRODUCT SPECIFICATION
TMC22091/TMC22191
Pin Descriptions (continued)
Pin Number
Pin Name
84-Lead 100-Lead
PLCC
MQFP
Value
Pin Function Description
CS
6
72
TTL
Chip Select. When CS is HIGH, the microprocessor interface
port, D7-0, is set to HIGH impedance and ignored. When CS is
LOW, the microprocessor can read or write parameters over
D7-0. One additional falling edge of CS is needed to move input
data to its assigned working registers.
R/W
7
73
TTL
Bus Read/Write Control. When R/W and CS are LOW, the
microprocessor can write to the control registers or CLUT over
D7-0. When R/W is HIGH and CS is LOW, it can read the
contents of any CLUT address or control register over D7-0.
RESET
5
71
TTL
Master Reset Input. Bringing RESET LOW sets the software
reset control bit, SRESET, LOW, forcing the internal state
machines to their starting states and disabling all outputs.
Bringing RESET HIGH synchronizes the internal pixel clock
(PCK = PXCK / 2) to maintain a defined pipeline delay through
the TMC22x91. If HRESET is set HIGH, the encoder is enabled
when RESET goes HIGH. If HRESET is LOW, the host restarts
the TMC22x91 by setting SRESET HIGH. RESET does not
affect the CLUT or the control registers, except SRESET.
COMPOSITE
33
2
1 V P-P
NTSC/PAL Video. Analog output of composite D/A converter,
nominally 1.35 volt peak-to-peak into a 37.5Ω load.
LUMA
35
5
1 V P-P
Luminance-only Video. Analog output of luminance D/A
converter, nominally 1.35 volt peak-to-peak into a 37.5Ω load.
CHROMA
37
8
1 V P-P
Chrominance-only Video. Analog output of chrominance D/A
converter, nominally 1.35 volt peak-to-peak into a 37.5Ω load.
VREF
30
98
+1.23 V
Voltage Reference Input. External voltage reference input,
internal voltage reference output, nominally 1.235 V.
COMP
39
10
0.1 µF
Compensation Capacitor. Connection point for 0.1µf
decoupling capacitor.
RREF
31
99
392Ω
Current-setting Resistor. Connection point for external
current-setting resistor for D/A converters. The resistor (392Ω)
is connected between RREF and AGND. Output video levels
are inversely proportional to the value of RREF.
Video Output
Analog Interface
JTAG Test Interface
TDI
25
93
TTL
Data Input Port. Boundary scan data input port.
TMS
24
92
TTL
Scan Select Input. Boundary scan (HIGH)/normal operation
(LOW) selector.
TCK
23
91
TTL
Scan Clock Input. Boundary scan clock.
TDO
22
90
TTL
Data Output Port. Boundary scan data output port.
+5 V
Positive digital power supply.
+5 V
Positive analog power supply.
Power Supply
VDD
VDDA
27, 64, 81 41, 62, 95
40-43
13-17
DGND
10, 26, 65, 42, 61, 76,
80
94
0.0 V
Digital Ground.
AGND
32, 34, 36,
38
0.0 V
Analog Ground.
4, 6, 9,
100
9
PRODUCT SPECIFICATION
TMC22091/TMC22191
Pin Descriptions (continued)
Pin Number
Pin Name
84-Lead 100-Lead
PLCC
MQFP
Value
Pin Function Description
0.0 V
Factory testing (TMC22091 only). Reserved for factory
testing. These pins have no effect on the operation but do
function as JTAG registers. They should be grounded directly
or pulled down to ground with 1kΩ or smaller resistors.
Test
TEST
28, 29,
48-51
22-25,
96-97
N/A
1, 3, 7,
11-12,
28-30,
52-55,
69, 78-79
NC
No Connect
Control Registers
The TMC22x91 is initialized and controlled by a set of registers. The registers are organized into 13 categories:
1.
Global Control
2.
Format Control
3.
Interface Control
4.
Test Control
5.
Key Control
6.
Misc. Control
7.
Standards Control
8.
Layering Control (TMC22191)
9.
Key Value
10. Timing
11. Subcarrier
12. Test I/O
13. Mask Register
An external controller loads the Control Registers through a
standard interface port. It also loads the CLUT and reads its
contents or those of the Control Registers. The port is
governed by pins CS, R/W, and A1-0.
The Address Register for the CLUT and the Control Register
pointer automatically increment to allow successive writes to
sequential addresses. In the CLUT, the Address Register has
two additional bits which increment in modulo-three to
sequentially access the red, green, and blue portions. All
three colors must be written when any CLUT address is
changed.
The control register autoincrement follows the sequence
indicated in the Control Register Map. When it reaches
address 40, it stops incrementing, allowing multiple reads or
writes of test data from/to the TESTDAT register. To exit the
test mode, reset the Control Register pointer by setting A1-0,
D7-0, and R/W LOW and then bring CS LOW. Address 1F is
a read-only status register. It is addressed by the autoincrement sequencer. Any data may be written into this port at
that time but it will not be stored. When address 50 is
accessed, no autoincrement takes place, allowing multiple
writes to the Mask Register.
10
PRODUCT SPECIFICATION
TMC22091/TMC22191
Table 2. Microprocessor Port Control
A1-0
R/W
Action
00
0
Write D7-0 into Control Register
pointer
00
1
01
01
Table 3. Control Register Map (continued)
Reg
Bit
Name
Function
02
2
FBDIS
Frame buffer signals
disable
Read Control Register pointer
on D7-0
02
1
PDCDIR
PDC master, slave
select
0
Write D7-0 into CLUT Address
Pointer
02
0
FLDLK
Field lock select
1
Read CLUT Address Pointer on
D7-0
Test Control Register
03
7
Reserved
10
0
Write D7-0 to addressed Control
Register
03
6
LIMEN
Luminance limiter
enable
10
1
Read addressed Control
Register on D7-0
03
5
TESTEN
Test enable
03
4
HOLDEN
MSBs/LSBs hold select
11
0
Write D7-0 to addressed CLUT
location
03
3
TSTMSB
LSBs, MSBs in/out
select
11
1
Read addressed CLUT location
on D7-0
03
2
LUMTST
LUMA channel test
03
1
8FSUBR
8-field subcarrier reset
enable
03
0
CHRTST
CHROMA channel test
Table 3. Control Register Map
Reg
Bit
Name
Key Control Register
04
7
04
6
HKEN
Hardware key enable
Reserved
04
5
BUKEN
Burst key enable
Function
Global Control Register
Reserved
00
7-5
00
4
SRESET
Software reset
04
4
SKEXT
00
3
PAL
Standard select, NTSC
or PAL
Data key operation
select
04
3
DKDIS
00
2
LUMDIS
Luminance input disable
Green/red/Y data key
disable
00
1
CHRDIS
Chrominance input
disable
04
2
EKDIS
Blue/green/CB data key
disable
00
0
HRESET
Software reset disable
04
1
FKDIS
Red/blue/CR data key
disable
04
0
SKEN
Data key enable
Format Control Register
01
7
Reserved
01
6
LCREN
Layering Control
Register enable
(TMC22191)
01
5
RAMPEN
Modulated ramp test
01
4
CB
Color bar test
01
3-2
FORMAT
PD23-0 input format
select
01
1-0
INMODE
PD23-0 input mode
select
Interface Control Register
02
7
VITSEN
VITS lines enable
02
6
SHCY
Short-cycle test mode
02
5-4
TBASE
Time-base source select
02
3
SOUT
Sync output mode select
Layering Control Register (TMC22191)
04
7
LAYMODE
MSB of Layer
Assignments select
04
6
HKEN
Hardware key enable
04
5
BUKEN
Burst key enable
04
4
SKEXT
Data key operation
select
04
3-1
LAYMODE
LSBs of Layer
Assignments select
04
0
SKEN
Data key enable
Key Value Registers
05
7-0
DKEY
Green/red/Y data key
value
06
7-0
EKEY
Blue/green/CB data key
value
11
TMC22091/TMC22191
PRODUCT SPECIFICATION
Table 3. Control Register Map (continued)
Reg
Bit
Name
Function
07
7-0
FKEY
Red/blue/CR data key
value
08-0D
Table 3. Control Register Map (continued)
Reg
Bit
Name
Function
19
7-0
FP
Front porch length
1A
7-0
EL
Equalization pulse LOW
length
1B
7-0
EH
Equalization pulse HIGH
length
Reserved
Misc. Control Register
0E
7
EFEN
Register 0E and 0F
enable
1C
7-0
SL
0E
6
COMPD/A
COMPOSITE D/A
disable
Vertical sync LOW
length
1D
7-0
SH
Vertical sync HIGH
length
1E
7-0
CBL
Color bar length
1F
7-5
FIELD
Field identification
1F
4-0
LTYPE
Line type identification
0E
5
SVIDD/A
LUMA/CHROMA D/A
disable
0E
4
FKREN
Luminance processing
enabled
0E
3
RATIO
Luminance ratio select
0E
2
TFLK
Luminance pass
threshold select
20
7-0
FREQL
Subcarrier frequency 4th
byte (LSBs)
0E
1
T512
EH/SL offset select
21
7-0
FREQ3
0E
0
CB100
NTSC/PAL Color Bars
Subcarrier frequency 3rd
byte
22
7-0
FREQ2
Same as Reg 0E bit 7
but read-only
Subcarrier frequency
2nd byte
23
7-0
FREQM
625/525 line per frame
select
Subcarrier frequency 1st
byte (MSBs)
24
7-0
SYSPHL
Video phase offset LSBs
Standards Control Register
0F
0F
7
6
EFEN
SIX25
Subcarrier Registers
0F
5
PALID
Phase alternate line
select
25
7-0
SYSPHM
Video phase offset
MSBs
0F
4
SETUP
7.5 IRE Pedestal Enable
26
7-0
BURPHL
Burst phase offset LSBs
0F
3-2
YGAIN
Luminance gain settings
27
7-0
BURPHM
Burst phase offset MSBs
0F
1-0
CGAIN
Chrominance gain
settings
Timing Registers
7-0
SY
Horizontal sync tip
length
11
7-0
BR
Breezeway length
12
7-0
BU
Burst length
13
7-0
CBP
Color back porch length
14
7-0
XBP
Extended color back
porch 8 LSB
15
7-0
VA
Active video 8 LSB
16
7-0
VC
Active video start 8 LSB
17
7-0
VB
Active video end 8 LSB
18
7-6
XBP
Extended color back
porch 2 MSB
18
5-4
VA
Active video 2 MSB
18
3-2
VC
Active video start 2 MSB
18
1-0
VB
Active video end 2 MSB
Reserved
Test I/O Register
40
10
12
28-3F
7-0
TESTDAT
Test data input/output
Mask Register
50
7-0
MASK
Mask register
Y-Component Register
60
7-0
Y
Y-component input/
output
Notes:
1. Functions are listed in the order used for reading and
writing.
2. For each register listed above, all bits not listed are
reserved and should be set to zero to ensure proper
operation.
3. The meaning of Register 04 (Key Control Register/Layering Control Register) is determined by Format Control
Register bit 6 (TMC22191).
PRODUCT SPECIFICATION
TMC22091/TMC22191
Control Register Definitions
Global Control Register (00)
7
6
5
Reserved
Name
4
3
2
1
0
SRESET
PAL
LUMDIS
CHRDIS
HRESET
Reg
Bit
Function
00
7-5
00
4
SRESET
Software reset. When LOW, resets and holds internal state machines and
disables outputs. When HIGH (normal), starts and runs state machines and
enables outputs.
00
3
PAL
Video standard select. When LOW, the NTSC standard is generated with 7.5
IRE pedestal. When HIGH, PAL standard video is generated. This bit is ignored
if Register 0E bit 7 is HIGH, enabling the 0E and 0F registers.
00
2
LUMDIS
Luminance input disable. When LOW (normal), luminance (Y) data from
external frame buffer is enabled. When HIGH, luminance (Y) data into the
TMC22x91 is forced to 0 IRE but sync pulses continue from the LUMA output.
00
1
CHRDIS
Chrominance input disable. When LOW (normal), burst and frame buffer data
into the TMC22x91 are enabled. when HIGH, burst and frame buffer data are
suppressed, enabling monochrome operation.
00
0
HRESET
Software reset enable. SRESET is forced LOW when the RESET pin is taken
LOW. State machines are reset and held. When HRESET is LOW, RESET may
be taken HIGH at any time. The TMC22x91 is enabled and a new frame is
begun with line 1, field 1 on the next PXCK after SRESET is set HIGH. The D/A
converters are powered down while RESET is LOW. When HRESET is HIGH,
a new frame is begun with line 1, field 1 on the next PXCK after RESET is taken
HIGH. SRESET is ignored. The D/A converters remain active during the reset
sequence.
Reserved.
13
TMC22091/TMC22191
PRODUCT SPECIFICATION
Control Register Definitions (continued)
Format Control Register (01)
7
6
5
4
Reserved
LCREN
RAMPEN
CB
Name
3
2
FORMAT
1
0
INMODE
Reg
Bit
Function
01
7
01
6
LCREN
(TMC22191) Layering Control Register enable. When LOW, the Layering
Control Register is not available and Key Control Register functions are
enabled. In this mode, the TMC22191 functions like the TMC22091. When
HIGH, the Layering Control Register takes the place of the Key Control
Register and enables the layering functions. Data loaded into the Key or
Layering Control Registers will remain but have a different meaning if this bit is
changed.
01
5
RAMPEN
Modulated ramp test. When LOW (normal), the TMC22x91 encodes and
outputs video corresponding to input data. When RAMPEN and CB are both
HIGH, an internally generated 40 IRE modulated ramp is produced, preempting
input data.
01
4
CB
Color bar test. When HIGH (normal), the TMC22x91 encodes and outputs
video corresponding to input data. When CB, RAMPEN, and Format Control
Register bit 0 are LOW, internally generated color bars are produced,
preempting input data.
01
3-2
FORMAT
PD23-0 input format select. Two bits select RGB, GBR, or YCBCR input data.
When bits 3 and 2 are:
Reserved.
0 0 the CLUT output is interpreted as RGB and is converted to YCBCR.
0 1 is reserved. Bits 3 and 2 must be 00 or 10 when the Layering Control
Register is enabled (TMC22191).
1 0 the CLUT output is interpreted as GBR, and is converted to YCBCR.
1 1 the CLUT output is interpreted as YCBCR.
01
1-0
INMODE
PD23-0 input mode select. These two bits set up the TMC22x91 for either 444,
422, 15-bit, or 8-bit input modes.
00
01
10
11
24-bit/pixel GBR, RGB, or YCBCR444 data enters from PD23-0
YCBCR422 data enters from PD23-8; CR and CB alternate from PD15-8
15-bit/pixel GBR or RGB data from PD14-0
8-bit/pixel color indexed data enters from PD7-0.
Bits 1 and 0 must be 00, 01, or 11 when the Layering Control Register is
enabled (TMC22191).
14
PRODUCT SPECIFICATION
TMC22091/TMC22191
Control Register Definitions (continued)
Interface Control Register (02)
7
6
VITSEN
SHCY
5
4
TBASE
3
2
1
0
SOUT
FBDIS
PDCDIR
FLDLK
Reg
Bit
Name
Function
02
7
VITSEN
VITS lines enable. When LOW, all UBB lines in the vertical interval are black
burst regardless of input data. When HIGH, all UBB lines in the vertical interval
become UVV active video and are dependent upon input data.
02
6
SHCY
Short-cycle test mode. When LOW, normal operation is enabled. when HIGH,
EH (equalization pulse HIGH length) and SL (vertical sync LOW length) are
shortened by 256.
02
5-4
TBASE
Time-base source select. These two bits set up the TMC22x91 for either
genlock or frame buffer control of timing. When bits 5 and 4 are:
0 0 the encoder counts out its own time-base from input clock PXCK.
0 1 the encoder locks to synchronizing signals from external genlock.
1 0 the encoder locks to synchronizing signals from frame buffer controller.
02
3
SOUT
Sync output mode select. When LOW, VHSYNC and VVSYNC output separate
horizontal and vertical sync pulses. When HIGH, composite sync (H and V) is
output on VVSYNC while horizontal sync is output on VHSYNC.
02
2
FBDIS
Frame buffer signals enable. When LOW, VVSYNC and VHSYNC outputs to
frame buffer are enabled. When HIGH, VVSYNC and VHSYNC outputs to
frame buffer are disabled.
02
1
PDCDIR
PDC master/slave select. When LOW, PDC is an output where the encoder is
requesting data from the frame buffer. When HIGH, PDC is an input, and
directs the encoder to accept data from the frame buffer.
02
0
FLDLK
Field lock select. When LOW, (in Slave mode) the encoder locks to each new
field. When HIGH, the encoder locks to field 1 only.
15
TMC22091/TMC22191
PRODUCT SPECIFICATION
Control Register Definitions (continued)
Test Control Register (03)
7
6
5
4
3
2
1
0
Reserved
LIMEN
TESTEN
HOLDEN
TSTMSB
LUMTST
8FSUBR
CHRTST
Reg
Bit
03
7
03
6
LIMEN
Luminance limiter enable. When LOW, all luminance values are passed to
modulator. when HIGH, luminance values are limited to 101 IRE.
03
5
TESTEN
Test enable. When LOW, normal operation is enabled. When HIGH,
TESTDAT7-0 (Register 40) is connected to the composite output (READ) and
D/A converters (WRITE) for test.
03
4
HOLDEN
MSBs/LSBs hold select. When LOW, alternates MSBs and LSBs in test, at
PXCK rate. When HIGH, reads/writes only MSBS or LSBS in test (per
TSTMSB, bit 3)
03
3
TSTMSB
LSBS,MSBS hold select. When LOW, connects 2 LSBs to TESTDAT1-0
for testing when TESTEN is HIGH. When HIGH, connects 8 MSBs to
TESTDAT7-0 for testing when TESTEN is HIGH.
03
2
LUMTST
LUMA channel test. When LOW (normal), the luminance D/A converter is
driven from luminance channel. When HIGH, the luminance D/A converter is
driven from TESTDAT for testing when TESTEN is HIGH.
03
1
8FSUBR
8-field subcarrier reset enable. When LOW, the internal subcarrier generator is
reset with frequency and phase data from FREQ, SYSPH, and BURPH
registers every eight fields. When HIGH, the internal subcarrier generator freeruns on the basis of frequency and phase data from the last time it was reset.
When RESET goes LOW, the subcarrier frequency and phase will be reset
from FREQ, SYSPH, and BURPH after field 8.
03
0
CHRTST
CHROMA channel test. When LOW (normal), the chrominance D/A converter is
driven from chrominance channel. When HIGH, the chrominance D/A converter
is driven from TESTDAT when TESTEN is HIGH.
16
Name
Function
Reserved.
PRODUCT SPECIFICATION
TMC22091/TMC22191
Control Register Definitions (continued)
Key Control Register (04)
7
6
5
4
3
2
1
0
Reserved
HKEN
BUKEN
SKEXT
DKDIS
EKDIS
FKDIS
SKEN
Reg
Bit
Name
Function
04
7
04
6
HKEN
Hardware key enable. When LOW, the KEY input pin ignored. When HIGH, the
KEY input pin is enabled.
04
5
BUKEN
Burst key enable. When LOW, output video burst is generated on TMC22x91.
When HIGH, output burst is taken from genlock input data.
04
4
SKEXT
Data key operation select. When LOW, data keying is allowed only during
active video. When HIGH, keying is allowed during active video and blanking.
04
3
DKDIS
Green/red/Y data key disable. When LOW, green/red/Y input data is enabled
for data keying. When HIGH, green/red/Y input data is ignored for data keying.
This function is enabled when Layering Control Register is enabled
(TMC22191).
04
2
EKDIS
Blue/green/CB data key disable. When LOW, Blue/green/CB input data is
enabled for data keying. When HIGH, Blue/green/CB input data is ignored for
data keying. This function is enabled when Layering Control Register is
enabled (TMC22191).
04
1
FKDIS
Red/blue/CR data key disable. When LOW, red/blue/CR input data is enabled
for data keying. When HIGH, red/blue/CR input data is ignored for data keying.
This function is enabled when Layering Control Register is enabled
(TMC22191).
04
0
SKEN
Data key enable. When LOW, data keying is disabled. When HIGH, data keying
is enabled.
Reserved.
17
TMC22091/TMC22191
PRODUCT SPECIFICATION
Control Register Definitions (continued)
Layering Control Register (04) (TMC22191)
7
6
5
4
LAYMODE
HKEN
BUKEN
SKEXT
Reg
Bit
Name
3
2
LAYMODE
1
0
SKEN
Function
04
7
LAYMODE
MSB of Layer Assignments select.
04
6
HKEN
Hardware key enable. When LOW, the KEY input pin ignored. When HIGH, the
KEY input pin is enabled.
04
5
BUKEN
Burst key enable. When LOW, output video burst is generated on TMC22191.
When HIGH, output burst is taken from genlock input data.
04
4
SKEXT
Data key operation select. When LOW, data keying is allowed only during
active video. When HIGH, data keying is allowed during active video and
blanking.
04
3-1
LAYMODE
Three LSBs of Layer Assignments select.
04
0
SKEN
Data key enable. When LOW, data keying is disabled. When HIGH, data keying
is enabled.
Key Value Registers (05-07)
Reg
Bit
Name
Function
05
7-0
DKEY
Green/red/Y data key value. Eight bits hold the match value which triggers
keying on red/Y.
06
7-0
EKEY
Blue/green/U data key value. Eight bits hold the match value which triggers
keying on green/U.
07
7-0
FKEY
Red/blue/V key value. Eight bits hold the match value which triggers keying on
blue/V.
18
PRODUCT SPECIFICATION
TMC22091/TMC22191
Control Register Definitions (continued)
Miscellaneous Control Register (0E)
7
6
5
4
3
2
1
0
EFEN
COMPD/A
SVIDD/A
FKREN
RATIO
TFLK
T512
CB100
Reg
Bit
Name
Function
0E
7
EFEN
Register 0E and 0F enable. When LOW, the functions of Register 0E and 0F
are disabled. When HIGH, Registers 0E and 0F are active. When Registers 0E
and 0F are enabled, Register 00 bit 3 is ignored. Register 0E bit 7 will read back
whatever value was written.
0E
6
COMPD/A
COMPOSITE D/A disable. When HIGH, the COMPOSITE D/A converter is
powered-down. When LOW, the D/A is enabled.
0E
5
SVIDD/A
LUMA/CHROMA D/A disable. When HIGH, the LUMA and CHROMA D/A
converters are powered-down. When LOW, they are enabled.
0E
4
FKREN
Luminance processing enable. When FKREN is HIGH, the KEY input defines
the function of CVBS input data. When the KEY input is HIGH, CVBS data is
keyed over PD input data. When KEY is LOW, CVBS data is assumed to be
luminance data delayed by one When FKREN is LOW, the KEY input operates
normally, switching between CVBS and PD data.
0E
3
RATIO
Luminance ratio control bit. When LOW, 1/2 of current luminance and 1/2 of
field delayed luminance from the CVBS input are added to yield a new
combined luminance value. When RATIO is HIGH, 3/4 of current luminance is
added to 1/4 of the delayed luminance to produce a new luminance value.
0E
2
TFLK
Luminance-pass threshold. The difference between current luminance and
delayed luminance (from the CVBS inputs) is compared against a preset
threshold set by TFLK. When TFLK is LOW, the high threshold must be
exceeded to trigger the combining of current and delayed luminance (according
to RATIO). If the higher threshold is not exceeded, current luminance is passed
without modification. When TFLK is HIGH, a lower threshold is used to trigger
the combining of current and delayed luminance.
0E
1
T512
EH/SL offset control bit. When LOW, the true value of EH and SL is offset by
256. When HIGH, the true value for EH and SL is offset by 512.
0E
0
CB100
NTSC/PAL color bars select. When HIGH, color bars with 100% white level are
selected. When LOW, color bars will have 75% white level.
19
TMC22091/TMC22191
PRODUCT SPECIFICATION
Control Register Definitions (continued)
Standards Control Register (0F)
7
6
5
4
EFEN
SIX25
PALID
SETUP
3
2
YGAIN
1
0
CGAIN
Reg
Bit
Name
Function
0F
7
EFEN
Same as Register 0E bit 7, but read-only.
0F
6
SIX25
Select 625 lines per frame. When HIGH, the encoder assumes 625 line per
frame. When LOW, 525 lines per frame are assumed.
0F
5
PALID
PAL select. When HIGH, Phase alternate line (PAL) operation is selected.
When LOW, operation conforms to NTSC standards.
0F
4
SETUP
Setup enable. When HIGH, a 7.5 IRE Pedestal is added to the output video.
when LOW, no pedestal is added.
0F
3-2
YGAIN
Luminance gain settings are adjusted to conform to the following NTSC and
PAL standards:
00
01
10
11
0F
1-0
CGAIN
NTSC without SETUP
NTSC-A and PAL-M
PAL-I and PAL-N
Reserved
Chrominance gain settings are adjusted to conform to the following NTSC and
PAL standards:
00
01
10
11
NTSC without SETUP
NTSC-A and PAL-M
PAL-I and PAL-N
Reserved
Timing Registers (10-17)
Reg
Bit
Name
Function
10
7-0
SY
Horizontal sync tip length. This 8-bit register holds a value extending from 0 to
255 PCK cycles.
11
7-0
BR
Breezeway length. This 8-bit register holds a value extending from 0 to 255
PCK cycles.
12
7-0
BU
Burst length. This 8-bit register holds a value extending from 0 to 255 PCK
cycles.
13
7-0
CBP
Color back porch length. This 8-bit register holds a value extending from 0 to
255 PCK cycles.
14
7-0
XBP
Extended color back porch 8 LSBs. This 8-bit register holds the LSBs of a
10-bit value extending from 0 to 1023 PCK cycles. The two MSBs are located in
control register 18.
15
7-0
VA
Active video 8 LSBs. This 8-bit register holds the LSBs of a 10-bit value
extending from 0 to 1023 PCK cycles. The two MSBs are located in control
register 18.
16
7-0
VC
Active video start 8 LSBs. This 8-bit register holds the LSBs of a 10-bit value
which is the initial half active video length extending from 0 to 1023 PCK cycles.
The two MSBs are located in control register 18.
17
7-0
VB
Active video end 8 LSBs This 8-bit register holds the LSBs of a 10-bit value
which is the end half active video length extending from 0 to 1023 PCK cycles.
The two MSBs are located in control register 18.
20
PRODUCT SPECIFICATION
TMC22091/TMC22191
Control Register Definitions (continued)
Timing Register (18)
7
6
5
XBP
4
3
VA
2
1
VC
0
VB
Reg
Bit
Name
Function
18
7-6
XBP
Extended color back porch 2 MSBs. These two bits hold the MSBs of a 10-bit
value extending from 0 to 1023 PCK cycles. The LSBs are located in control
register 14.
18
5-4
VA
Active video 2 MSB. These two bits hold the MSBs of a 10-bit value extending
from 0 to 1023 PCK cycles. The LSBs are located in control register 15.
18
3-2
VC
Active video start 2 MSBs. These two bits hold the MSBs of a 10-bit value which
is the initial half active video length extending from 0 to 1023 PCK cycles. The
LSBs are located in control register 16.
18
1-0
VB
Active video end 2 MSBs. These two bits hold the MSBs of a 10-bit value which
is the end half active video length extending from 0 to 1023 PCK cycles. The
LSBs are located in control register 17.
Timing Registers (19-1E)
Reg
Bit
Name
Function
19
7-0
FP
Front porch length. This 8-bit register holds a value extending from 0 to 255
PCK cycles.
1A
7-0
EL
Equalization pulse LOW length. This 8-bit register holds a value from 0 to 255
PCK cycles.
1B
7-0
EH
Equalization pulse HIGH length. This 8-bit register holds a value extending from
0 to 255 PCK cycles. This value, when added to 256 (or 512), determines the
final pulse length in the range of 256 to 511 (or 767) PCK cycles.
1C
7-0
SL
Vertical sync LOW length. This 8-bit register holds a value from 0 to 255 PCK
cycles. This value, when added to 256 (or 512), determines the final pulse
length in the range of 256 to 511 (or 767) PCK cycles.
1D
7-0
SH
Vertical sync HIGH length. This 8-bit register holds a value extending from 0 to
255 PCK cycles.
1E
7-0
CBL
Color bar length. This 8-bit register holds a value which is the length of each
color bar displayed extending from 0 to 255 PCK cycles.
Timing Register (1F)
7
6
5
FIELD
4
3
2
1
0
LTYPE
Reg
Bit
Name
Function
1F
7-5
FIELD
Field identification (read only). These three bits are updated 12 PXCK periods
after each VHSYNC. They allow the user to determine field type on a
continuous basis.
1F
4-0
LTYPE
Line type identification (read only). These five bits are updates 5 PXCK periods
after each VHSYNC. They allow the user to determine line type on a continuous
basis.
21
TMC22091/TMC22191
PRODUCT SPECIFICATION
Control Register Definitions (continued)
Subcarrier Registers (20-27)
Reg
Bit
Name
Function
20
7-0
FREQL
Subcarrier frequency 4th byte (LSBs). This 8-bit register holds the LSB (bits
7-0) of the 32-bit subcarrier frequency value (non-genlock modes). The next
eight most significant bits are held in Register 21.
21
7-0
FREQ3
Subcarrier frequency 3rd byte. This 8-bit register holds bits 15:8 of the
subcarrier frequency value (non-genlock modes). The next eight most
significant bits are held in Register 22.
22
7-0
FREQ2
Subcarrier frequency 2nd byte. This 8-bit register holds bits 23-16 of the
subcarrier frequency value (non-genlock modes). The eight MSBs are held in
Register 23.
23
7-0
FREQM
Subcarrier frequency 1st byte (MSBs). This 8-bit register holds the MSBs (bits
31-24) of the 32-bit subcarrier frequency value (non-genlock modes).
24
7-0
SYSPHL
Video phase offset LSBs. This 8-bit register holds the 8 LSBs of color subcarrier
phase offset during active video.
25
7-0
SYSPHM
Video phase offset MSBs. This 8-bit register holds the 8 MSBs of color
subcarrier phase offset during active video.
26
7-0
BURPHL
Burst phase offset LSBs. This 8-bit register holds the 8 LSBs of burst phase
offset for color adjustment.
27
7-0
BURPHM
Burst phase offset MSBs. This 8-bit register holds the 8 MSBs of burst phase
for color adjustment.
Test I/O Register (40)
Reg
Bit
Name
Function
40
7-0
TESTDAT
Test data input/output. This 8-bit register holds MSBs or LSBs, as determined
by the Test Control Register. This control address does not auto-increment
during read or write operations. To exit the test mode, reset the Control
Register pointer by setting A1-0 and R/W LOW and then bring CS LOW.
Mask Register (50)
Reg
Bit
Name
Function
50
7-0
MASK
Mask register. This 8-bit register holds an 8-bit word that is logically ANDed with
the incoming data presented to the three CLUTs in color-index mode. This
register is a write-only register.
Y-Component Register (60)
Reg
Bit
Name
Function
60
7-0
Y
Y-component register. This register holds the contents of the luminance value
before the Sync and Blank Insert circuitry of the encoder. Loading the Control
Register pointer with 60h brings 8-bit Y values out on the D7-0 port.
22
PRODUCT SPECIFICATION
TMC22091/TMC22191
Color Lookup Table
gamma correction, are easily loaded. The color data is
loaded into the tables in G-B-R sequence in GBR mode, and
R-G-B sequence in RGB mode.
The CLUT can be used in a variety of ways, depending on
the data format and source presented to the PD port.
The CLUT is loaded like a RAMDAC, sequentially writing
one byte to each of the three locations associated with the
selected CLUT address. These three locations are referred to
as Tables D, E, and F as shown in table 16 (not R,G, and B
because they may or may not contain RGB information). and
are loaded in that sequence. The address will increment automatically after the three values at one address are written or
read.
Luminance/Color Difference Modes
The TMC22x91 expects Y, B-Y, and R-Y signals at the input
to its modulator section. When presenting CCIR-601 YCBCR
or digitized Y, B-Y, R-Y data to the CLUT, gain and offset factors are needed. Table 4 specifies the recommended transfer
functions. The CLUT is loaded in Y-CB-CR sequence.
Overlay Operation
Color-index Modes
In color-index (CI) mode, the CLUT is used to store the
color look-up data, translating the 8-bit source pixel data into
24-bit RGB colors. Table D holds red data, Table E is green
data, and Table F holds blue Data. The incoming data are
presented to the three tables in parallel, and a 24-bit output is
produced.
When the encoder is connected in parallel with a RAMDAC
in a VGA system, the CLUT can be loaded simultaneously
with the CLUT in the output RAMDAC. If a 6-bit RAMDAC
is employed, 6 bits can be loaded via data pins D7-2 (MSB
justified). The two LSBs should be set to 00 for optimal
black level representation, but the largest error introduced by
extraneous data in the LSBs is 3/4 LSB (at 6 bits). The
encoder will produce the closest possible translation of the
VGA colors in the encoded video environment.
GBR/RGB Modes
The nominal configuration for GBR/RGB modes is unity
gain (CLUT data = CLUT address) for PAL and NTSC.
Other transfer functions, such as gain adjustment, offset, and
For the TMC22091 and TMC22191 (when Format Control
Register Bit 6 = LOW), the OL4-0 inputs are inactive. In
CCIR-601 operation, the nominal data range for Y is from 16
to 235 and for CB and CR is from 16 to 240. This means that
CLUT locations 0 to 15 and 241 to 255 are available for
overlay colors. When the overlay locations are addressed (by
forcing CLUT addresses outside the normal CCIR-601 data
range), the addressed CLUT data is encoded resulting in the
specific color found in that CLUT location. Overlay colors
information stored in the unused CLUT locations must be Y,
B-Y, R-Y values. Y, B-Y, and R-Y values are found from
RGB values by:
Y = 0.299 R + 0.587 G + 0.114 B
B-Y = -0.299 R – 0.587 G + 0.886 B
R-Y = 0.701 R – 0.587 G – 0.114 B
For the TMC22191, when the Format Control Register
Bit 6 = HIGH, Overlay is controlled by the OL4-0 inputs
which directly access CLUT locations, 01 thru 0F and F1
thru FF, as shown in Table 5. The values stored in these
CLUT locations must be in RGB format.
Table 4. CLUT Transfer Functions for NTSC and PAL
Input Format (CLUT Address)
Output Format (CLUT Data)
Component
Data Range
Transfer Equations
Component
Data Range
R
0 to 255
R0=R
RO
0 to 255
G
0 to 255
G0=G
GO
0 to 255
B
0 to 255
B0=B
BO
0 to 255
Y
16 to 235
YO = Y * 1.1644 – 18.63
YO
0 to 255
CB
±112
(B-Y)O = CB * 1.0126
(B-Y)O
±113
CR
±112
(R-Y)O = CR * 0.8011
(R-Y)O
±90
Y
0 to 255
Y0=Y
YO
0 to 255
B-Y
±127
(B-Y)O = (B-Y) * 0.893
(B-Y)O
±113
R-Y
±127
(R-Y)O = (R-Y) * 0.7065
(R-Y)O
±90
23
TMC22091/TMC22191
PRODUCT SPECIFICATION
Table 5. CLUT Locations Addressed by
Overlay Inputs (TMC22191)
Table 5. CLUT Locations Addressed by
Overlay Inputs (TMC22191) (continued)
OL4-0
CLUT location
OL4-0
CLUT location
00
No Overlay
1E
FE
01
01
1F
FF
02
02
•
•
•
•
0E
0E
Color-space Conversion in the Matrix
0F
0F
10
No Overlay
11
F1
12
F2
•
•
•
•
When the input pixels are in RGB, GBR, or color-index format and the CLUT is bypassed (TMC22191), the Matrix
remains enabled, converting RGB data to color-difference
format. When the input pixels are in 444 format
(YCBCR444, RGB, GBR, CI), the Interpolator (which converts 422 to 444) is not active. When the input pixels are in
YCBCR format, the CLUT is enabled to scale the data to
color-difference values and the Matrix is inactive. In colorindex mode, the Matrix is active, converting the RGB CLUT
output data to color-difference values.
Table 6. Pixel Input Operation for Format Control Register bit 6 = HIGH (TMC22191)
Format Control Register
Pixel Data Format
FORMAT
Bit 3,2
INMODE
Bit 1,0
BYPASS = LOW
CLUT bypassed
BYPASS = HIGH
CLUT enabled
00 (RGB)
00 (444)
RGB
YCBCR444
00
01 (422)
RGB
YCBCR422
00
10 (15-bit)
RGB
RGB15
00
11 (CI)
RGB
CI
01
xx
reserved
reserved
10 (GBR)
00 (444)
GBR
YCBCR444
10
01 (422)
GBR
YCBCR422
10
10 (15-bit)
GBR
GBR15
10
11 (CI)
GBR
CI
11
xx
not allowed
not allowed
Format Control Register Bit 6 = HIGH
MSB
23
Pixel 1
Pixel 2
Pixel 3
Pixel 4
Pixel n
Pixel n+1
16 15
Format Control Register Bit 6 = LOW
LSB
0
87
Y1
Y2
Y3
Y4
CB1
CB1
CB3
CB3
CR1
CR1
CR3
CR3
Yn
Yn+1
CBn
CBn
CRn
CRn
MSB
23
Pixel 1
Pixel 2
Pixel 3
Pixel 4
Pixel n
Pixel n+1
16 15
87
Y1
Y2
Y3
Y4
CB1
CR1
CB3
CR3
Yn
Yn+1
CBn
CRn
LSB
0
27003A
Note: The pixel input sequence begins on the first LDV pulse after PDC goes HIGH. n = Odd number
Figure 2. Pixel Data (PD23-0) Sequence for YCBCR422
24
PRODUCT SPECIFICATION
Gamma Correction
Gamma is built into broadcast television systems as a correction factor for nonlinearity in image acquisition (nonlinear
conversion of light into current in a vidicon) and at the display (phosphor nonlinearity in converting current into light).
A Gamma corrector transfer function takes the form of
Output = k ( Input )1/γ
where a typical Gamma is 2.2 for NTSC, 2.8 for PAL.
Computer systems usually ignore Gamma in driving a display monitor. Each R, G, and B channel is treated as linear.
When encoding a computer display output to video, the user
must decide whether to apply a gamma correction factor and,
if so, what value. It is a good assumption that, since the digital video input over the CVBS bus is in composite form, it
has been Gamma corrected.
Gamma correction is applied in the RGB domain. When
operating in YCBCR, for example when encoding a CCIR601 signal, Gamma should have already been applied.
Gamma correction is readily added to the RGB transfer
equations shown in Table 4.
Video Timing
The TMC22x91 can be programmed to accommodate a wide
range of system timing requirements. With a line locked
pixel rate of 10 to 15 Mpps, the digitally synthesized horizontal waveforms and subcarrier frequency and phase are
determined from 24 registers that are loaded by a controller.
TMC22091/TMC22191
Table 7. Horizontal Timing Specifications
Parameter
NTSC-M
(µs)
PAL-I
(µs)
PAL-M
(µs)
FP
SY
BR
BU
CBP
VA
H
1.5
4.7
0.6
2.5
1.6
52.6556
63.5556
1.65
4.7
0.9
2.25
2.55
51.95
64.0
1.9
4.95
0.9
2.25
1.8
51.692
63.492
Vertical Programming
Vertical interval timing is also fully programmable, and is
established by loading the timing registers with the durations
of each vertical timing element, the duration expressed in
PCK clock cycles. In this way as with horizontal programming, any pixel rate between 10 and 15 Mpps can be accommodated, and any desired standard or non-standard vertical
video timing may be produced.
Like horizontal timing parameters, vertical timing parameters are calculated as follows:
t = N x (PCK period)
= N x (2 x PXCK period)
where N is the value loaded into the appropriate timing register, and PCK is the pixel clock period.
The Vertical Interval comprises several different line types
based upon H, the Horizontal line time.
Horizontal Programming
Horizontal interval timing is fully programmable, and is
established by loading the timing registers with the durations
of each horizontal element. The duration is expressed in
PCK clock cycles. In this way, any pixel clock rate between
10 MHz and 15 MHz can be accommodated, and any desired
standard or non-standard horizontal video timing may be
produced. Figure 3 illustrates the horizontal blanking interval with timing register identification.
Horizontal timing parameters can be calculated as follows:
t = N x (PCK period)
= N x (2 x PXCK period)
where N is the value loaded into the appropriate timing register, and PCK is the pixel clock period.
Horizontal timing resolution is two PXCK periods. PXCK
must be chosen such that it is an even integer multiple of the
horizontal line frequency. This ensures that the horizontal
line period, H, contains an integer number of pixels. The horizontal line comprises the sum of appropriate elements.
H = FP + SY + BR + BU + CBP + VA
When programming horizontal timing, subtract 5 PCK
periods from the calculated values of CBP and add 5 PCK
periods to the calculated value for VA.
H = (2 x SL) + (2 x SH) [Vertical sync pulses]
= (2 x EL) + (2 x EH) [Equalization pulses]
The VB and VC lines are added to produce the half-lines
needed in the vertical interval at the beginning and end of
some fields. These must properly mate with components of
the normal lines.
VB = CBP + VA - XBP = H/2 – CBPVC
= VA – (EL + EH) = VA – H/2
where Equalization HIGH and LOW pulses (EL + EH) = H/2
and the Extended Color Back Porch, XBP = VA + CBP –VB.
XBP begins after the end of burst, BU, taking the place of
CBP in vertical interval UBV lines. Figure 5 shows the vertical sync and equalization pulse detail
Table 8. Vertical Timing Specifications
Parameter
(µs)
NTSC-M
(µs)
PAL-I
(µs)
PAL-M
(µs)
H
EH
EL
SH
SL
63.5556
29.4778
2.3
4.7
27.1
64
29.65
2.35
4.7
27.3
63.492
29.45
2.3
4.65
27.1
25
TMC22091/TMC22191
PRODUCT SPECIFICATION
H
H/2
Burst
EH
VA
FP
SY
BU
FP
FP
VA
SL
EL
SH
24318B
24319A
Figure 3. Horizontal Blanking Interval Timing
Figure 4. Sync and Equalization Pulse Detail Timing
FIELDS 1 AND 3
524
523
UVV
UVV
PRE-EQUALIZATION
3H
1
2
3
VERTICAL BLANKING
VERTICAL SYNC
3H
4
5
6
POST-EQUALIZATION
3H
7
8
9
10
19
20
EE
SS
EE
UBB
UBB
UBB
EE
EE
SS
SS
EE
EE
21
22
UVV
UVV
VHSYNC
VVSYNC
COMPOSITE
SYNC
262
UVV
FIELDS 2 AND 4
263
UVE
264
265
266
267
268
269
270
271
272
273
282
EE
EE
ES
SS
SS
SE
EE
EE
EB
UBB
UBB
283
284
285
UBV
UVV
UVV
VHSYNC
VVSYNC
COMPOSITE
SYNC
27000A
Figure 5. NTSC Vertical Interval
26
PRODUCT SPECIFICATION
TMC22091/TMC22191
Table 9. NTSC Field/Line Sequence and Identification
Field 1
FIELD ID = x00
Field 2
FIELD ID = x01
Field 3
FIELD ID = x10
Field 4
FIELD ID = x11
Line
ID
LTYPE
Line
ID
LTYPE
Line
ID
LTYPE
Line
ID
LTYPE
1
EE
00
264
EE
00
1
EE
00
264
EE
00
2
EE
00
265
EE
00
2
EE
00
265
EE
00
3
EE
00
266
ES
01
3
EE
00
266
ES
01
4
SS
03
267
SS
03
4
SS
03
267
SS
03
5
SS
03
268
SS
03
5
SS
03
268
SS
03
6
SS
03
269
SE
02
6
SS
03
269
SE
02
7
EE
00
270
EE
00
7
EE
00
270
EE
00
8
EE
00
271
EE
00
8
EE
00
271
EE
00
9
EE
00
272
EB
10
9
EE
00
272
EB
10
10
UBB
0D
273
UBB
0D
10
UBB
0D
273
UBB
0D
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
20
UBB
0D
282
UBB
0D
20
UBB
0D
282
UBB
0D
21
UVV
0F
283
UBV
0E
21
UVV
0F
283
UBV
0E
22
UVV
0F
284
UVV
0F
22
UVV
0F
284
UVV
0F
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
262
UVV
0F
524
UVV
0F.
262
UVV
0F
524
UVV
0F
263
UVE
0C
525
UVV
0F
263
UVE
0C
525
UVV
0F
EE
SS
EB
UVV
UBV
Equalization pulse
Vertical sync pulse
Equalization broad pulse
Active video
Half-line black, half-line video
SE
ES
UBB
UVE
Master and Genlock mode details of VHSYNC, VVSYNC,
and composite VVSYNC (SOUT = HIGH) outputs are
shown in Figures 5 and 6. When VHSYNC and VVSYNC
Half-line vertical sync pulse, half-line equalization pulse
Half-line equalization pulse, half-line vertical sync pulse
Black burst
Half-line video, half-line equalization pulse
are used as inputs (Slave mode), their falling edges mark the
beginning of the sync interval and the width of the input
pulse is specified under Operating Conditions.
27
TMC22091/TMC22191
1247
PRODUCT SPECIFICATION
1248
UVV
VE
309
310
FIELDS 1 AND 5
1249
1250
1
2
3
4
5
6
7
22
23
EE
EE
SS
SS
SE
EE
EE
BB
UBB
UBB
UBB
24
25
26
UBV
UVV
UVV
336
337
VHSYNC
VVSYNC
COMPOSITE
SYNC
UVV
VV
622
623
FIELDS 2 AND 6
311
312
313
314
315
316
317
318
319
320
334
335
EE
EE
ES
SS
SS
EE
EE
EB
UBB
UBB
UBB
UBV
UVV
UVV
649
650
651
UBV
UVV
UVV
961
962
UVV
UVV
VHSYNC
VVSYNC
COMPOSITE
SYNC
UVV
VE
934
935
FIELDS 3 AND 7
624
625
626
627
628
629
630
631
632
647
648
EE
EE
SS
SS
SE
EE
EE
UBB
UBB
UBB
UBB
VHSYNC
VVSYNC
COMPOSITE
SYNC
UVV
UV
FIELDS 4 AND 8
936
937
938
939
940
941
942
943
944
945
959
960
EE
EE
ES
SS
SS
EE
EE
EB
BB
UBB
UBB
UBV
VHSYNC
VVSYNC
COMPOSITE
SYNC
27001A
Figure 6. PAL Vertical Interval
28
PRODUCT SPECIFICATION
TMC22091/TMC22191
Table 10. PAL Field/Line Sequence and Identification
Field 1
FIELD ID = 000, 100
Field 2
FIELD ID = 001, 101
Field 3
FIELD ID = 010, 110
Field 4
FIELD ID = 011, 111
Line
ID
LTYPE
Line
ID
LTYPE
Line
ID
LTYPE
Line
ID
LTYPE
1
SS
03
313
ES
01
626
SS
03
938
ES
01
2
SS
03
314
SS
03
627
SS
03
939
SS
03
3
SE
02
315
SS
03
628
SE
02
940
SS
03
4
EE
00
316
EE
00
629
EE
00
941
EE
00
5
EE
00
317
EE
00
630
EE
00
942
EE
00
6
-BB
05
318
EB
10
631
UBB
0D
943
EB
10
7
UBB
0D
319
UBB
0D
632
UBB
0D
944
-BB
05
8
UBB
0D
320
UBB
0D
633
UBB
0D
945
UBB
0D
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
22
UBB
0D
335
UBB
0D
647
UBB
0D
960
UBB
0D
23
UBV
0E
336
UVV
0F
648
UBV
0E
961
UVV
0F.
24
UVV
0F
337
UVV
0F
649
UVV
0F
962
UVV
0F
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
•••
308
UVV
0F
621
UVV
0F
933
UVV
0F
1246
UVV
0F
309
UVV
0F
622
-VV
07
934
UVV
0F
1247
UVV
0F
310
-VV
07
623
-VE
04
935
UVV
0F
1248
-VE
04
311
EE
00
624
EE
00
936
EE
00
1249
EE
00
312
EE
00
625
EE
00
937
EE
00
1250
EE
00
Equalization pulse
Vertical sync pulse
Equalization broad pulse
Black burst with color burst suppressed
Active video with color burst suppressed
Half-line video, half-line equalization
pulse, color burst suppressed.
SE
ES
UBB
UVV
UVE
UBV
EE
SS
EB
-BB
-VV
-VE
Half-line vertical sync pulse, half-line equalization pulse
Half-line equalization pulse, half-line vertical sync pulse
Black burst
Active video
Half-line video, half-line equalization pulse
half-line black, half-line video
29
TMC22091/TMC22191
521
UVV
PRODUCT SPECIFICATION
522
UVV
18
FIELDS 1 AND 5
523
524
525
1
2
3
4
5
6
7
8
9
17
EE
EE
EE
SS
SE
SS
EE
EE
EE
BB
BB
UBB
UBB
UVV
VHSYNC
VVSYNC
COMPOSITE
SYNC
259
260
FIELDS 2 AND 6
UVV
261
262
263
264
265
266
267
268
269
270
271
279
EE
EE
ES
SS
SS
SE
EE
EE
EB
BB
UBB
UBV
VE
280
281
UVV
UVV
VHSYNC
VVSYNC
COMPOSITE
SYNC
521
18
522
UVV
VV
FIELDS 3 AND 7
523
524
525
1
2
3
4
5
6
7
8
9
17
EE
EE
EE
SS
SS
SS
EE
EE
EE
BB
UBB
UBB
UBB
UVV
VHSYNC
VVSYNC
COMPOSITE
SYNC
258
UVV
259
UV
260
VE
280
281
UBV
UVV
FIELDS 4 AND 8
261
262
263
264
265
266
267
268
269
270
271
279
EE
EE
ES
SS
SS
SE
EE
EE
EB
BB
UBB
UBB
VHSYNC
VVSYNC
COMPOSITE
SYNC
27082A
Figure 7. PAL-M Vertical Interval
30
PRODUCT SPECIFICATION
TMC22091/TMC22191
Table 11. PAL-M Field/Line Sequence and Identification
Field 1
FIELD ID = 000, 100
Field 2
FIELD ID = 001, 111
Field 3
FIELD ID = 010, 110
Field 4
FIELD ID = 011, 111
Line
ID
LTYPE
Line
ID
LTYPE
Line
ID
LTYPE
Line
ID
LTYPE
1
SS
03
263
ES
01
1
SS
03
263
ES
01
2
SS
03
264
SS
03
2
SS
03
264
SS
03
3
SS
03
265
SS
03
3
SS
03
265
SS
03
4
EE
00
266
SE
02
4
EE
00
266
SE
02
5
EE
00
267
EE
00
5
EE
00
267
EE
00
6
EE
00
268
EE
00
6
EE
00
268
EE
00
7
-BB
05
269
EB
10
7
-BB
05
269
EB
10
8
-BB
05
270
-BB
05
8
UBB
05
270
-BB
05
9
UBB
0D
271
UBB
1D
•••
•••
•••
271
UBB
1D
•••
•••
•••
•••
•••
•••
17
UBB
0D.
•••
•••
•••
17
UBB
0D
279
UBB
0D
18
UVV
0F
279
UBB
0D
18
UVV
0F
280
UBV
0E.
•••
•••
•••
280
UBV
0E.
···
···
···
281
UVV
0F
258
UVV
0F
281
UVV
0F
259
UVV
0F
•••
•••
•••
259
-VV
07
•••
•••
•••
260
-VE
04
521
UVV
0F
260
-VE
04
521
UVV
0F
261
EE
00
522
-VV
07
261
EE
00
522
UVV
0F
262
EE
00
262
EE
00
EE
SS
EB
-BB
-VV
-VE
523
EE
00.
523
EE
00
524
EE
00
524
EE
00
525
EE
00
525
EE
00
Equalization pulse
Vertical sync pulse
Equalization broad pulse
Black burst with color burst suppressed
Active video with color burst suppressed
Half-line video, half-line equalization
pulse, color burst suppressed.
SE
ES
UBB
UVV
UVE
UBV
Half-line vertical sync pulse, half-line equalization pulse
Half-line equalization pulse, half-line vertical sync pulse
Black burst
Active video
Half-line video, half-line equalization pulse
half-line black, half-line video
31
TMC22091/TMC22191
PRODUCT SPECIFICATION
Table 12. Standard Timing Parameters
Timing Register (hex)
Field Horizontal Pixel PXCK
Rate
Freq.
Rate
Freq.
(Hz)
(kHz)
(Mpps) (MHz)
SY
10
BR
11
BU
12
NTSC sqr.
pixel
59.94 15.734266
12.27
24.54
3A
07
1F
0F
23
NTSC
CCIR-601
59.94 15.734266
13.50
27.00
40
08
22
13
NTSC 4x
FSC
59.94 15.734266
14.32
28.64
43
09
24
PAL sqr.
pixel
50.00 15.625000
14.75
29.50
45
0D
PAL
CCIR-601
50.00 15.625000
13.50
27.00
40
PAL 15
Mpps
50.00 15.625000
15.00
30.00
PAL-M
sqr.pixel
60.00 15.750000
12.50
PAL-M
CCIR-601
60.00 15,750000
PAL-M 4x
FSC
60.00 15,750000
Standard
VC
16
VB Note 1 FP
17
18
19
EL EH2 SL2 SH
1A 1B 1C 1D
8B
05
77
65
12
1C
6A
4C
3A
52
3F
CA
1D
9D
65
13
1F
8E
6E
3F
59
12
54
F7
30
B5
65
15
21
A6
84
43
5F
21
21
6D
03
2B
B7
75
19
23
B5
93
45
61
0C
1E
22
4D
BE
0E
93
65
16
20
90
71
3F
58
46
0D
22
21
73
11
31
BF
75
19
23
BD
9A
47
62
25.01
3E
0B
1C
13
26
86
FE
8B
61
18
1D
70
53
3A
52
13.50
27.00
44
0C
1E
13
26
Bf
12
99
65
1A
1F
8E
6E
3F
57
14.30
28.60
47
0D
20
15
4C
E8
22
AC
65
1B
21
A5
84
42
5D
CBP XBP VA
13
14
15
CBL
1E
Notes:
1. XBP, VA, VC, and VB are 10-bit values. The 2 MSBs for these four variables are in Timing Register 18. See Table 3.
2. EH and SL are 9-bit values. A most significant "1" is forced by the TMC22x91 since EH and SL must range from 256 to 511.
EH and SL may be extended to 767. Only the eight LSBs are stored in Timing Registers 1B and 1C.
3. Every calculated timing parameter has a minimum value of 5 except EH and SL which have minimum values of 256.
VITS Insertion
Subcarrier Programming
In both NTSC and PAL, the TMC22x91 can be set up to
allow Vertical Interval Test Signals (VITS) in the vertical
interval in place of normal black burst lines (UBB). This is
controlled by Interface Control Register bit 7. If this bit is
LOW, UBB lines are black burst and are independent of
TMC22x91 input data. If the bit is HIGH, all vertical interval
UBB lines become UVV. UVV lines are active video and
depend upon data input to the TMC22x91. VITS lines may
carry special test signals or pass captioning data through the
encoder.
The color subcarrier is produced by an internal 32-bit digital
frequency synthesizer which is completely programmable in
frequency and phase. Separate registers are provided for phase
adjustment of the color burst and of the active video, permitting external delay compensation and color adjustment.
Edge Control
SMPTE 170M NTSC and Report 624 PAL video standards
call for specific rise and fall times on critical portions of the
video waveform. The TMC22x91 does this automatically.
The TMC22x91 digitally defines slopes compatible with
SMPTE 170M NTSC or CCIR Report 624 PAL on:
1.
H and V Sync leading and trailing edges.
2.
Burst envelope.
3.
Active video leading and trailing edges.
32
In Master or Slave mode, the subcarrier is internally synchronized to establish and maintain a specified relationship
between the falling edge of horizontal sync and color burst
phase (SCH). In NTSC and PAL, SCH synchronization is
performed every eight fields, on field 1 of the eight-field
sequence. Proper subcarrier phase is maintained through the
entire eight fields, including the 25 Hz offset in PAL
systems. See the description of 8FSUBR under Test Control
Register bit 1 for the subcarrier reset function.
In Genlock mode, the phase and relative frequency of the
incoming video are transmitted by the TMC22071 Genlocking Video Digitizer over the CVBS bus at the beginning of
each line, which synchronize the digital subcarrier synthesizer. When key control register bit BUKEN is HIGH and
digitized burst from the TMC22071 is passed through to the
reconstruction D/A converter, the reference subcarrier for the
chrominance modulator is still synthesized within the
encoder.
PRODUCT SPECIFICATION
TMC22091/TMC22191
NTSC Subcarrier
PAL Subcarrier
For NTSC encoding, the subcarrier synthesizer frequency
has a simple relationship to the pixel clock period, repeating
over 2 lines: The decimal value is:
The PAL relationship is more complex, repeating only once
in 8 fields (the well-known 25 Hz offset):
( 1135 ⁄ 4 ) + ( 1 ⁄ 625 )
32
FREQ = --------------------------------------------------- × 2
( pixels/line )
( 455 ⁄ 2 )
32
FREQ = ----------------------------- × 2
( pixels/line )
This value must be converted to binary and split as described
previously for NTSC.
This value must be converted to binary and split into four
8-bit registers, FREQM, FREQ2, FREQ3, and FREQL. The
number of pixels/line is:
For PAL, the decimal value for SYSPH is found from:
FREQ
SYSPH = ---------------- = BURPH
17
2
Pixels/line = (2/PXCK frequency) (H period)
SYSPH establishes the appropriate phase relationship
between the internal synthesizer and the chroma modulator.
The nominal value for SYSPH is zero.
This value must be converted to binary and split into two
8-bit registers, SYSPHM and SYSPHL. Burst Phase in PAL
is identical to SYSPH. Therefore, the same values for
SYSPHM and SYSPHL must be used for BURPHM and
BURPHL.
Other values for SYSPH must be converted to binary and
split into two 8-bit registers, SYSPHM and SYSPHL.
PAL-M Subcarrier
Burst Phase (BURPH) sets up the correct relative NTSC
modulation angle. The value for BURPH is:
( 909 ⁄ 4 )
32
FREQ = ----------------------------- × 2
( pixels/line )
BURPH = SYSPH + 8,192
FREQ
SYSPH = ---------------- = BURPH
17
2
This value must be converted to binary and split into two
8-bit registers, BURPHM and BURPHL. The decimal number 8,192 advances the burst phase by 45°.
Table 13. Standard Subcarrier Parameters
Field
Rate
(Hz)
Horizontal
Freq.
(kHz)
Pixel
Rate
(MHz)
PXCK
Freq.
(MHz)
Subcarrier Register (hex)
Sub-carrier
Freq.
BURPHM BURPHL SYSPHM SYSPHL FREQM FREQ2
(MHz)
27
26
25
24
23
22
NTSC sqr.
pixel
59.94
15.734266
12.27
24.54
3.57954500
20
00
00
00
4A
AA
AA
AB
NTSC
CCIR-601
59.94
15.734266
13.50
27.00
3.57954500
20
00
00
00
43
E0
F8
3E
NTSC 4x
fSC
59.94
15.734266
14.32
28.64
3.57954500
20
00
00
00
40
00
00
00
PAL sq.
pixel
50.00
15.625000
14.75
29.50
4.43361875
00
00
00
00
4C
F3
18
19
PAL
CCIR-601
50.00
15.625000
13.50
27.00
4.43361875
00
00
00
00
54
13
15
96
PAL 15
Mpps
50.00
15.625000
15.00
30.00
4.43361875
00
00
00
00
4B
AA
C6
A1
PAL-M sq.
pixel
60
15.750
12.50
25.01
3.57561149
00
00
00
00
49
45
00
51
PAL-M
CCIR-601
60
15,750
13.50
27.00
3.57561149
00
00
00
00
43
DF
3F
D7
PAL-M 4x
fSC
60
15,750
14.30
28.60
3.57561149
00
00
00
00
40
10
66
F5
Standard
FREQ3 FREQL
21
20
33
TMC22091/TMC22191
PRODUCT SPECIFICATION
SCH Phase Error Correction
SCH refers to the timing relationship between the 50% point
of the leading edge of horizontal sync and the positive or
negative zero-crossing of the color burst subcarrier reference. SCH error is usually expressed in degrees of subcarrier
phase. In PAL, SCH is defined for line 1 of field 1, but since
there is no color burst on line 1, SCH is usually measured at
line 7 of field 1. The need to specify SCH relative to a particular line in PAL is due to the 25 Hz offset of PAL subcarrier
frequency. Since NTSC has no such 25 Hz offset, SCH
applies to all lines.
The SCH relationship is only important in the TMC22x91
when two video sources are being combined or if the composite video output is externally combined with another
video source. In these cases, improper SCH phasing will
result in a noticeable horizontal jump of one image with
respect to another and/or a change in hue proportional to the
SCH error between the two sources.
SCH phasing can be adjusted by modifying BURPH and
SYSPH values by equal amounts. SCH is advanced/delayed
by one degree by increasing/decreasing the value of BURPH
and SYSPH by approximately B6h. An SCH error of 15° is
corrected with SYSPH and BURPH offsets of AAAh.
Video Test Signals
The TMC22x91 has two standard video test waveforms for
evaluating video signal integrity. They are selected and
controlled by the Format Control Register.
(Figure 11), the luminance component stair-step signal at the
LUMA output, and the chrominance component on the
CHROMA output. The six colors are 100% saturated PAL
and 75% saturated for NTSC.
The percentage color saturation is selectable via Misc.
Control Register 0E, bit 0.
The color bar test pattern comprises eight equal-width bars
during VA, the active video period. The Timing Register
value for CBL is found from:
VA + 7
CBL = -----------------8
If CBL is larger than this, the color bars are truncated at the
end of VA. If CBL is smaller than VA/8, the color bar
sequence will repeat, starting with another white bar. From
left to right color bars 1 to 8 should be white, yellow, cyan,
green, magenta, red, blue, and black.
The modulated ramp waveform is enabled by setting the Format Control Register to 30h. It comprises constant-amplitude
and constant-phase subcarrier modulation superimposed on a
linear ramp which slews from black to white during the
active video portion of each horizontal line (Figure 12). This
waveform is useful in making differential gain and differential phase measurements. Differential gain is a measure of
the variation in saturation of a color as the luminance component is varied from black to white. Differential phase is a
measure of the variation in hue of a color as the luminance
component is varied from black to white.
Setting the Format Control Register bits 0, 4, and 5 LOW
generates standard color bars at the COMPOSITE output
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
24387A
Figure 8. 100% Color Bars With 100% and 75% Chrominance Saturation
34
PRODUCT SPECIFICATION
TMC22091/TMC22191
Microprocessor Interface
The microprocessor interface comprises 13-lines. Two
address bits provide four addresses for device programming
and CLUT/register management. Address bit 0 selects
between control registers and CLUT memory. Address bit 1
selects between reading/writing the register addresses and
reading/writing register or CLUT data.
When writing, the address is presented along with a LOW on
the R/W pin during the falling edge of CS. Eight bits of data
are presented on D7-0 during the subsequent rising edge of
CS.
24388A
Figure 7. Modulated Ramp Waveform
tPWLCS
One additional falling edge of CS is needed to move input
data to the assigned working registers.
tPWHCS
CS
tSA
tHA
R/W
A1-A0
tSD
tHD
D7-D0
24323A
Figure 10. Microprocessor Port – Write Timing
tPWLCS
tPWHCS
CS
tSA
tHA
R/W
A1-A0
tDOM
tHOM
D7-D0
24324A
tDOZ
Figure 11. Microprocessor Port – Read Timing
35
TMC22091/TMC22191
PRODUCT SPECIFICATION
In read mode, the address is accompanied by a HIGH on the
R/W pin during a falling edge of CS. The data output pins go
to a low-impedance state tDOZ ns after CS falls. Valid data is
present on D7-0 tDOM after the falling edge of CS. Because
this port operates asynchronously with the pixel timing,
there is an uncertainty in this data valid output delay of one
PXCK period. This uncertainty does not apply to tDOZ.
and Blank Insert block are monitored. When the Control
Register pointer is loaded with 60h, the D7-0 port will output
8-bit luminance pixels synchronous with respect to PXCK.
To halt the pixel flow from D7-0, bring CS HIGH.
Operational Timing
The TMC22x91 operates in three distinct modes:
The RESET pin restores the TMC22x91 to field 1 line 1 and
places the encoder in a power-down state (if HRESET is
LOW). Bit 4 of the Global Control Register (SRESET) is set
LOW. All other control words and CLUT contents are left
unchanged. Returning RESET HIGH synchronizes the internal clock with PXCK and restores the device outputs to
active states.
Reading Pixel Data from the D7-0 Port
The microprocessor port of the TMC22x91 may be used to
monitor digital video outputs. The eight MSBs of the upsampled and interpolated pixel data that go to the
COMPOSITE D/A converter can also be accessed via the
D7-0 port. When the Test Control Register is loaded with 28h
and the Control Register pointer is loaded with 40h, the D7-0
port will output the 8-bit composite pixels synchronous with
PXCK. To halt the pixel flow from D7-0, simply bring CS
HIGH.
1.
Master mode. The encoder independently produces all
internal timing and provides digital sync to the host
controller.
2.
Slave mode. The encoder accepts horizontal and vertical
sync from the controller and synchronizes the video output accordingly.
3.
Genlock mode. The encoder accepts horizontal and vertical sync from the companion TMC22071 Genlocking
Video Digitizer, synchronizes itself to the incoming
video, and provides appropriate H Sync and V Sync to
the host. It synchronizes Pixel Data input in two ways:
a.
Internal PDC. The encoder internally generates the
Pixel Data Control (PDC) signal which calls for
data input from the external pixel source.
External PDC. The encoder receives a PDC signal
from the host and accepts Pixel Data based on that
input.
b.
Luminance pixel data may also be read from D7-0. In this
case, the eight MSBs of luminance at the input of the Sync
1
2
3
PXCK
tSR
tHR
tSR
RESET
24330A
Figure 12. Reset Timing – PCK Synchronization
Reset Timing
The TMC22x91 operates from a master clock (PXCK) at
twice the pixel rate. In Master mode, the PCK to PXCK timing relationship is set on the rising edge of RESET. In Figure
12, PCK is denoted by odd PXCK counts.
When RESET is taken LOW with sufficient setup time (tSR)
before a rising edge of PXCK, the internal state machines are
reset and the device is put into a mode as dictated by the Global Control Register bits 0 and 4. In Master mode, when the
RESET pin is taken HIGH, the internal clock timing is established. In Slave and Genlock mode, this timing is established
by VHSYNC and GHSYNC respectively. The first PXCK
following this RESET rising edge is designated as PXCK 1.
Where it is significant, reference PXCK timing will be
shown with numbered rising edges. A designation of 2N
clocks refers to an even number of PXCK rising edges from
device reset. If RESET is not shown and clock numbering
36
does not refer to 2N, timing is relative to signals shown in
the diagram only.
Pixel Data Input Timing
PXCK is internally divided by 2 to generate an internal pixel
clock, PCK which is not accessible from the pins of the
TMC22x91. To ensure the correct phase relationship
between PCK and pixel data, PCK is locked to VHSYNC or
GHSYNC (Slave or Genlock mode, respectively). In Master
mode, VHSYNC is produced on the rising edge of PCK
allowing external circuitry to synchronize the generation of
pixel data and LDV which also operates at the rate of PCK.
The rising edge of LDV clocks the 24-bit pixel data into
three 8-bit registers while PCK clocks that data through the
pixel data path within the TMC22x91. It is therefore necessary to meet the set-up and hold timing between pixel data
and LDV as well as LDV and PCK as shown in Figure 13.
PRODUCT SPECIFICATION
TMC22091/TMC22191
tPWHPX
tPWLPX
2N+1
PXCK
2N+2
tSP
2N+3
tPWLVH
VHSYNC
(GHSYNC)
PCK
tPWHLDV
tXL
tPWLLDV
LDV
tSP
tHP
PD
KEY
24340A
Figure 13. Slave Mode PD Port Interface Timing (Genlock Mode)
0
1
2
3
4
16
17
18
19
20
21
51
52
53
54
55
56
57
58
59
60
PXCK
RESET
VHSYNC
COMPOSITE
OUTPUT
50% Sync Amplitude
24353A
Figure 14. Master Mode Timing
Master Mode
Slave Mode
In Master mode, initial timing is determined from the
RESET input, and subsequent cycles result from programmed values in the Timing Control Registers. The Horizontal Sync output, VHSYNC, goes LOW 18 PXCK clock
cycles after the device is reset. The 50% point of the falling
edge of sync LOW on line 4 of field 1 (NTSC) or line 1 of
field 1 (PAL) occurs at the COMPOSITE and LUMA outputs 56 clocks after reset, or 38 clocks after VHSYNC. See
Figure 14, Master Mode Timing.
In Slave mode, the 50% point of the falling edge of sync
occurs 46 PXCK clocks after the falling edge of VHSYNC,
which is an input signal to the TMC22x91. This must be provided by the host to begin every line. If it is early, the line
will be started early, maintaining the 52 clock delay to output. If it comes late, the front porch portion of the output
waveform will be extended as necessary. See Figure 15,
Slave Mode Timing.
37
2N+56
2N+55
2N+54
2N+53
2N+52
2N+51
2N+50
2N+49
2N+48
2N+47
2N+46
2N+45
2N+44
2N+43
2N+42
2N+41
2N+40
2N+4
2N+3
PRODUCT SPECIFICATION
2N+2
2N+1
2N+0
TMC22091/TMC22191
PXCK
VHSYNC1
VVSYNC1
for field 1
COMPOSITE
OUTPUT
50% Sync Amplitude
24355A
2N+54
2N+53
2N+52
2N+51
2N+50
2N+49
2N+48
2N+47
2N+14
2N+13
2N+12
2N+11
2N+10
2N+9
2N+8
2N+7
2N+6
2N+3
2N+2
2N+1
2N+0
Figure 15. Slave Mode Timing
PXCK
GHSYNC
VHSYNC
COMPOSITE
OUTPUT
50% Sync Amplitude
24356A
Figure 16. Genlocked Mode Timing
Genlocked Mode
In Genlocked mode, the encoder receives sync signals over
the GHSYNC and GVSYNC inputs, and provides VHSYNC
and VVSYNC to the host. The 50% sync amplitude point
occurs 50 PXCK clocks after GHSYNC goes LOW, while
VHSYNC is produced at clock 13. If GHSYNC is late, the
front porch is lengthened, if is is early, front porch is
shortened. See Figure 16, Genlock Mode Timing.
Pixel Data Control
The Pixel Data Control (PDC) signal determines the active
picture area. It may be an input or an output, as determined
by the Interface Control Register bit 1.
38
The position (number of PCK cycles) of the rising edge of
PDC relative to the falling edge of VHSYNC can be found
by summing SY, BU, BR, and CBP. See Figure 17.
External Pixel Data Control
When used as an input, PDC goes HIGH four PXCK cycles
before the first valid pixel of a line is presented to the PD
input port. If this signal is late (with respect to the horizontal
blanking interval programmed in the timing control registers), the Color Back Porch (CBP) will be extended. If it is
early, incoming pixel data will be ignored until the end of the
CBP.
2N+50
2N+49
2N+48
2N+47
2N+46
2N+45
2N+44
2N+43
2N+42
P19
2N+41
2N+39
P4
2N+40
2N+9
2N+8
2N+7
2N+6
2N+5
2N+4
2N+3
TMC22091/TMC22191
2N+2
2N+1
2N+0
PRODUCT SPECIFICATION
PXCK
PDC
P1
PD
P2
P3
P21
P20
P22
P23
P24
COMPOSITE
OUTPUT
P1 PI1
P2 PI2 P3
POST-FILTER
OUTPUT
24358A
Figure 17. External Pixel Data Control
Internal Pixel Data Control
2N+50
2N+49
2N+48
2N+47
2N+46
2N+45
2N+44
2N+43
P19
2N+42
2N+39
P4
2N+41
2N+9
2N+8
2N+7
2N+6
2N+5
2N+4
2N+3
2N+2
2N+1
2N+0
When programmed as an output, PDC goes HIGH four
PXCK periods prior to the end of CBP (as programmed in
the horizontal timing registers) which is also four PXCK
cycles prior to required input of the first pixel of a line.
2N+40
Pixels produced by the encoder appear at the analog outputs
(COMPOSITE, LUMA, CHROMA) 40 clocks after they are
registered into the PD port. Note that the pixels enter at onehalf the PXCK rate. The encoded signal passes through
interpolation filters which generate intermediate output values, improving the output frequency response and greatly
simplifying the external reconstruction filter. The interpolated pixels are designated PI in the diagram.
PXCK
PDC
tDO
PD
P1
P2
P3
P20
P21
P22
P23
P24
COMPOSITE
OUTPUT
P1 PI1
P2 PI2 P3
POST-FILTER
OUTPUT
24357A
Figure 18. Internal Pixel Data Control
39
TMC22091/TMC22191
PRODUCT SPECIFICATION
Layering with the TMC22191
A 4-Layer Example
Layering is a video production process where various images
or patterns are superimposed (keyed) over each other to form
a layered composite of the input images. Four layers with the
following priority are provided by the TMC22191:
For this layering example, a BACKGROUND image is
generated. This image comprises shaded matte levels varying
from black at the top of the screen to white at the bottom.
This could just as well be a color image which will be seen
wherever no other image appears through the layering
process.
1.
The DOWNSTREAM KEY layer keys over all other
layers.
2.
The FOREGROUND layer keys over MIDGROUND
and BACKGROUND, but not over DOWNSTREAM
KEY.
3.
The MIDGROUND layer keys over BACKGROUND,
but not over FOREGROUND or DOWNSTREAM
KEY.
4.
The BACKGROUND layer never keys over any other
layer.
It is important not to confuse layers with sources. The
TMC22191 can be programmed to assign any of its input
sources (RGB, YCRCB, CVBS bus, Overlay bits) to any of
the four layers.
The ability to combine various video sources into a 4-layer
composite image is a very powerful tool in the production of
live video. The TMC22191 performs layering operations
entirely in the digital domain, enabling precise digital
control.
The MIDGROUND image comprises a happy face superimposed over a white rectangle. Only the happy face and the
white rectangle are of interest for this image and therefore,
the portion of the image outside that area will be replaced by
the BACKGROUND image when MIDGROUND is keyed
over BACKGROUND. A key signal is generated on a pixelby-pixel basis. It indicates which image is active. The key
signal for keying MIDGROUND over BACKGROUND is
shown to the right of the MIDGROUND image. This represents a single bit signal mapped over the image. When the
signal is black (logic LOW), the MIDGROUND image is
active, when it is white (logic HIGH), the BACKGROUND
is active.
The results of layering MIDGROUND over BACKGROUND images are shown in the 2-layer composite image
Figure 19.
BACKGROUND
KEY
2-LAYER COMPOSITE
MIDGROUND
27036A
Figure 19. 2-Layer Image Construction
A FOREGROUND image comprises a shaded matte rectangle with "HI KIDS !" alpha characters in its center. This is to
be superimposed over the previous 2-layer composite image.
The key signal needed for superimposing FOREGROUND
over other images is shown to the right of the FOREGROUND image. This represents a single bit signal mapped
over the image. When the signal is black (logic LOW), the
40
FOREGROUND image is active, when it is white (logic
HIGH), the composite image is active.
A new 3-layer composite image, FOREGROUND over
MIDGROUND over BACKGROUND, is shown in Figure
20.
PRODUCT SPECIFICATION
TMC22091/TMC22191
2-LAYER COMPOSITE
KEY
HI KIDS !
3-LAYER COMPOSITE
HI KIDS !
FOREGROUND
HI KIDS !
27037A
Figure 20. Adding a 3rd Layer
A DOWNSTREAM KEY image comprises the white alpha
characters "HAPPY FACE", and black alpha characters
"Time". This is to be superimposed over the previous 3-layer
composite image. The key signal needed for superimposing
DOWNSTREAM KEY image over the other composite
images is shown to the right. This represents a single bit signal mapped over the image. When the signal is black (logic
LOW), the DOWNSTREAM KEY image is active, when it
is white (logic HIGH), the previous composite image is
active.
The final 4-layer composite image, DOWNSTREAM KEY
over FOREGROUND over MIDGROUND over BACKGROUND, is shown in Figure 21.
In this illustration, all four source images are static (not moving). The images input to the TMC22191 can just as well be
"live" (from video camera or VCR sources) as long as:
• Data from those sources is in an input format that the
TMC22191 can accept, and
• The sources either synchronize the TMC22191
(Genlock mode) or are synchronized by the TMC22191
(Master or Slave mode).
Key signals may be generated external to the TMC22191
(Hardware Keying) and use the KEY input pin for control.
Key signals may also be generated within the TMC22191
(Data Keying) by the comparison of input color data with
color data stored in the TMC22191.
3-LAYER COMPOSITE
HI KIDS !
KEY
DOWNSTREAM KEY
HAPPY
FACE
TIME
HAPPY
FACE
TIME
4-LAYER COMPOSITE
HI KIDS !
HAPPY
FACE
TIME
27038A
Figure 21. Adding a 4th Layer
41
TMC22091/TMC22191
PRODUCT SPECIFICATION
2-Layer Keying with the TMC22091
The TMC22091 facilitates the keying of PD port input data
over the CVBS bus input data. Keying is controlled on a
pixel-by-pixel basis by either the KEY input pin or the internal Data Key. The first two layers in the previous 4-Layer
Example apply to the TMC22091. The result of keying is an
effect where a MIDGROUND source image (i.e. Happy Face
from PD data) is superimposed over a BACKGROUND
source image (i.e. variable matte color from CVBS data).
Assigning Video Sources to Layers with the
TMC22191
Digital video inputs to the TMC22191 (PD, CVBS, Overlay)
are assigned to the four layers by choosing one of the 16
modes of the Layering Control Register. OVERLAY is
always keyed (switched on a pixel-by-pixel basis from active
to transparent) by the OL4-0 inputs. OVERLAY can not be
programmed to the BACKGROUND layer. The CVBS digital video bus can be assigned to any of the four layers and is
keyed by the KEY input signal or internal Data Key. In
modes 0 thru 7, the CLUTs are not bypassed and the
BYPASS input is ignored.
Table 14. Layer Assignments, Image Sources, and Keying Controls (TMC22191)
LCR 04
Background
Midground
Foreground
Keying
Control
Downstream Key
Image Source:
Keying
Control
Image Source:
Keying
Control
LAYMODE
Image Source
Image Source
0
PD(YCBCR, RGB, CI)
CVBS
KEY or
Data Key
—
—
—
—
1
PD(YCBCR, RGB, CI)
CVBS
KEY or
Data Key
OVERLAY
OL4-0
—
—
2
PD(YCBCR, RGB, CI)
CVBS
KEY
PD(YCBCR, RGB, CI)
Data Key
OVERLAY
OL4-0
3
PD(YCBCR, RGB, CI)
CVBS
KEY
PD(YCBCR, RGB, CI)
Data Key
OVERLAY
OL4-0
4
CVBS
OVERLAY
OL4-0
PD(YCBCR, RGB, CI)
KEY or
Data Key
—
—
5
CVBS
PD(YCBCR, RGB, CI)
KEY or
Data Key
OVERLAY
OL4-0
—
—
6
PD(YCBCR, RGB, CI)
CVBS
KEY
OVERLAY
OL4-0
PD(YCBCR, RGB, CI) Data Key
OVERLAY
OL4-0
PD(YCBCR, RGB, CI) Data Key
7
PD(YCBCR, RGB, CI)
CVBS
KEY
8
PD(YCBCR, CI)
CVBS
KEY or
Data Key
9
PD(RGB)
PD(YCBCR, CI)
BYPASS
A
PD(RGB)
CVBS
B
PD(RGB)
C
D
—
—
—
CVBS
KEY or
Data Key
OVERLAY
OL4-0
KEY or
Data Key
PD(YCBCR, CI)
BYPASS
OVERLAY
OL4-0
CVBS
KEY or
Data Key
OVERLAY
OL4-0
PD(YCBCR, CI)
BYPASS
PD(RGB)
PD(YCBCR, CI)
BYPASS
OVERLAY
OL4-0
CVBS
KEY or
Data Key
CVBS
PD(RGB)
KEY
PD(YCBCR, CI)
BYPASS
OVERLAY
OL4-0
E
CVBS
OVERLAY
OL4-0
PD(RGB)
KEY
PD(YCBCR, CI)
BYPASS
F
PD(RGB)
OVERLAY
OL4-0
CVBS
KEY or
Data Key
PD(YCBCR, CI)
BYPASS
Notes:
1. For LAYMODE = 0 to 7, Pixel Data always passes through the CLUTs. FORMAT, INMODE, and the BYPASS pin selects the
input format for PD23-0 according to Table 6.
2. For LAYMODE = 8 to F and BYPASS = HIGH, Data Key is disabled.
3. Asserting the signal listed under "Keying Control:" enables the corresponding "Signal Source:". Signals with " " are asserted
by a logic LOW.
42
PRODUCT SPECIFICATION
TMC22091/TMC22191
Hardware Keying
The KEY input switches the COMPOSITE D/A converter
input from the luminance and chrominance combiner output
to the CVBS data bus on a pixel-by-pixel basis. This is a
"soft" switch, executed over four PXCK periods to minimize
out-of-band spurious signals. The video signal from the
CVBS bus can only present on the COMPOSITE output. The
CHROMA and LUMA outputs continue to present encoded
PD port data when CVBS is active.
Hardware keying is enabled by the Key Control Register bit
6. Normally, keying is only effective during the Active Video
portion of the waveform as determined by the VA registers
15 and 18. The Horizontal Blanking interval is generated by
the encoder state machine even if the KEY signal is held
HIGH through Horizontal Blanking. However, it is possible
to allow digital Horizontal Blanking to be passed through
from the CVBS bus to the COMPOSITE output by setting
0
1
2
3
4
5
6
7
8
9
Key Control Register bit 5 HIGH. In this mode, KEY is
always active, and may be exercised at will.
The KEY input is registered into the encoder just like Pixel
Data is clocked into the PD port. It may be considered a 25th
Pixel Data bit. It is internally pipelined, so the midpoint of
the key transition occurs at the output of the pixel that was
input at the same time as the KEY signal.
Data Keying
Data Keying internally generates a Key signal that acts
exactly as the external KEY signal. There are three Key
Value Registers 05, 06, and 07 that are matched against the
input data to the three tables in the CLUT. These tables are
designated D, E, and F. They contain different information
depending on the input mode selected as shown in Table 16.
38
39
40
41
42
43
44
45
46
47
48
49
PXCK
CVBS
VN+2
KEY
*
PD
PN
VN+3
VN+4
VN+5
VN+6
VN+25
VN+26
VN+27
VN+28
VN+29
VN+30
PN+1
PN+2
PN+3
PN+4
PN+23
PN+24
PN+25
PN+26
PN+27
PN+28
KEY MIDPOINT
VN+3
PIN+3VIN+2
PN+3VIN+2
PIN+2VIN+1
PN+2
PIN+1
PIN
PN
is advanced five PXCK cycles when
* KEY
Control Register OE bit 4 is HIGH (TMC22191).
PN+1
COMPOSITE
OUTPUT
24359A
Figure 22. Hardware Keying
The key registers may be individually enabled using bits
3,2,1 of the Key Control Register. Bit 4 of the same register
enables/disables Data Keying in its entirety. Data Keying and
Hardware Keying are logically ORed: when both are
enabled, either one will result in a key switch to the CVBS
channel.
The key comparison is based on the input data to the tables
in the CLUT. When operating in color-index mode, all three
tables receive the same input value, so any one of the three
registers is sufficient to identify a key value. The outputs of
all enabled key registers are ANDed to produce the KEY signal. If more than one key register are enabled and their key
values are not identical, no key will be generated.
Table 16. Table D, E, F Contents
Mode
Table D
Table E
Table F
GBR
Green
Blue
Red
RGB
Red
Green
Blue
YCBCR
Y
CB
CR
CI
CI
CI
CI
43
TMC22091/TMC22191
PRODUCT SPECIFICATION
Genlock Interface
The TMC22x91 can process digital composite video
connected to its CVBS port. It has been designed to couple
tightly with the companion TMC22071 Genlocking Video
Digitizer, but it will work with other sources as well.
Subcarrier frequency and phase data are transmitted to the
encoder over the CVBS bus as 4-bit nibbles on CVBS3-0
during the horizontal sync period. Field identification is also
required for the TMC22x91 internal sync generator. The
14th nibble of the sequence contains no relevant data.
The digital composite video has to be in standard 8-bit
binary format at a PXCK/2 rate. Synchronization with the
internal PXCK/2 is established by the phasing of the
GHSYNC input, as shown in Figures 24 and 25.
0
1
2
3
4
5
6
7
8
9
38
39
40
41
42
43
44
45
46
47
48
49
PXCK
CVBS
PD
VN+2
VN+3
VN+4
VN+5
VN+6
VN+25
VN+26
VN+27
VN+28
VN+29
VN+30
PN
PN+1
PN+2
PN+3
PN+4
PN+23
PN+24
PN+25
PN+26
PN+27
PN+28
KEY MIDPOINT
FIRST KEY WORD
VN+7
PIN+3VIN+6
PN+3VIN+6
PIN+2VIN+5
PN+2
PIN+1
PN+1
PIN
PN
COMPOSITE
OUTPUT
24360A
Figure 23. Data Keying
2N+1
PXCK
tSGI
2N+2
2N+3
tHGI
GHSYNC
tSGI
tHGI
CVBS7-0
24341A
Figure 24. Genlock Interface Timing
44
PRODUCT SPECIFICATION
TMC22091/TMC22191
PXCK
0
1
2
3
4
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
GHSYNC
CVBS7-0
PIXEL
PIXEL
PIXEL
f3:0 φ23:20 φ19:16 φ15:12 φ11:8 φ7:4 φ3:0
FID f23:20 f19:16 f15:12 f11:8 f7:4
FREQUENCY
PIXEL
PIXEL
PHASE
24383A
FIELD IDENTIFICATION
Figure 25. Frequency/Phase Data Transfer
Filtering
The TMC22x91 incorporates internal digital filters to establish appropriate bandwidths and simplify external analog filter designs.
Since these are fixed-coefficient digital filters, their filter
characteristics depend upon clock rate. Figures 26 and 27
show the frequency response for two pixel rates, 12.27 MHz
and 14.75 MHz.
0
-10
The color-difference low-pass filters in the TMC22x91
establish chrominance bandwidths which meet the specifications outlined in CCIR Report 624-3, Table II, Item 2.6, for
system I over a range of pixel rates from 12.27 Mpps to
14.75 Mpps. Equal bandwidth is established for both colordifference channels.
Attenuation (dB)
Color-Difference Low-Pass Filters
0
CCIR Rep 624 Specification
-50
-60
-70
-80
-90
f = 12.27 Mpps
3
0
6
-30
-40
-50
-60
-70
-80
0
9
12
Frequency (MHz)
15
24394A
Figure 27. Chroma Modulator and Luminance
Interpolation Filter Full Spectrum Response
f = 14.27 Mpps
f = 12.27 Mpps
f = 14.75 Mpps
0
1
2
3
4
5
Frequency (MHz)
6
7
8
24363A
Figure 26. Color-Difference Low-Pass Filter Response
Interpolation Filters
The Chroma Modulator output and the luminance data path
are digitally filtered with sharp-cutoff low-pass interpolation
filters. These filters ensure that aliased subcarrier, chrominance, and luminance frequencies are sufficiently suppressed
in the frequency band above base-band video and below the
pixel frequency (fS/4 to 3fS/4, where fS is the PXCK
frequency).
Attenuation (dB)
-20
Attenuation (dB)
-30
-40
CCIR Rep 624 Specification
-10
-90
f = 14.75 Mpps
-20
-1
-2
f = 12.27 Mpps
-3
-4
0
1
2
3
4
5
6
Frequency (MHz)
7
24365A
Figure 28. Chroma Modulator and Luminance
Interpolation Filter Passband Detail
All digital-to-analog reconstruction systems exhibit a high
frequency roll-off as a result of the zero-order hold characteristic of D/A converters. This response is commonly
referred to as a sin(x)/x response. It is a function of the sampling rate of the output D/A.
45
TMC22091/TMC22191
PRODUCT SPECIFICATION
The digital interpolation filters in TMC22x91 convert the
data stream to a sample rate of twice the pixel rate. As shown
in Figures 27 and 28, the filters decrease the sin(x)/x rolloff
and the output spectrum between fS/4 and 3fS/4 contains
very little energy. Since there is so little signal energy in this
frequency band, the demands placed on the output reconstruction filter are greatly reduced. The output filter needs to
be flat to fS/4 and have good rejection at 3fS/4. The relaxed
requirements greatly simplify the design of a filter with good
phase response and low group delay distortion. A small
amount of peaking may be used to compensate residual
sin(x)/x rolloff.
JTAG Test Interface
The JTAG test port accesses registers at every digital I/O pin
except the JTAG test port pins. Table 16 shows the sequence
of the test registers. The register number (Reg) indicates the
order in which the register data is loaded and read (Reg 1 is
loaded and read first, therefore it is at the end of the serial
path). The scan path is 59 registers long. The six TEST pins
of the TMC22091 function as JTAG registers.
The JTAG port is a 4-line interface, following IEEE Std.
1149.1-1990 specifications. The Test Data Input (TDI) and
Test Mode Select (TMS) inputs are referred to the rising
edge of the Test ClocK (TCK) input. The Test Data Output
(TDO) is referred to the falling edge of TCK.
Table 16. JTAG Interface Connections
Reg
Pin
Signal
Reg
Pin
Signal
Reg
Pin
Signal
1
28
BYPASS (TEST)
21
62
PD13
41
2
CVBS1
2
29
OL4 (TEST)
22
63
PD12
42
3
CVBS0
3
44
CVBS7
23
66
PD11
43
4
KEY
4
45
CVBS6
24
67
PD10
44
5
RESET
5
46
CVBS5
25
68
PD9
45
6
CS
6
47
CVBS4
26
69
PD8
46
7
R/W
7
48
OL3 (TEST)
27
70
PD7
47
8
A1
8
49
OL2 (TEST)
28
71
PD6
48
9
A0
9
50
OL1 (TEST)
29
72
PD5
49
11
PDC
10
51
OL0 (TEST)
30
73
PD4
50
12
VHSYNC
11
52
PD23
31
74
PD3
51
13
VVSYNC
12
53
PD22
32
75
PD2
52
14
D7
13
54
PD21
33
76
PD1
53
15
D6
14
55
PD20
34
77
PD0
54
16
D5
15
56
PD19
35
78
LDV
55
17
D4
16
57
PD18
36
79
PXCK
56
18
D3
17
58
PD17
37
82
GVSYNC
57
19
D2
18
59
PD16
38
83
GHSYNC
58
20
D1
19
60
PD15
39
84
CVBS3
59
21
D0
20
61
PD14
40
1
CVBS2
46
PRODUCT SPECIFICATION
TMC22091/TMC22191
tPWLTCK
tPWHTCK
TCK
tSTP
tHTP
TDI
TMS
tHOTP
tDOTP
TDO
24333A
Figure 29. JTAG Test Port Timing
Equivalent Circuits
VDD
VDD
n Substrate
n
RREF
p
VDD
VREF
COMPOSITE
LUMA
CHROMA
27012A
Figure 30. Equivalent Analog Input Circuit
n Substrate
VDD
27013A
Figure 31. Equivalent Analog Output Circuit
VDD
p
p
Input
Output
n
n
27014A
Figure 32. Equivalent Digital Input Circuit
27011A
Figure 33. Equivalent Digital Output Circuit
47
TMC22091/TMC22191
PRODUCT SPECIFICATION
tDOM
CS
tDOZ
tHOM
0.5V
Hi-Z
D7-0
2.0V
0.8V
27029A
0.5V
Figure 34. Transition Levels for Three-State Measurements
Absolute Maximum Ratings (beyond which the device may be damaged)1
Parameter
Min.
Max.
Unit
Power Supply Voltage
-0.5
7.0
V
-0.5
VDD+0.5
V
-20.0
20.0
mA
-0.5
VDD+0.5
V
-20.0
20.0
mA
Short Circuit Duration (Single output in HIGH state to GND)
1
second
Analog Output Short Circuit Duration (Single output to GND)
Infinite
Digital Inputs
Applied Voltage2
Forced
Current3,4
Digital Outputs
Applied Voltage2
Forced
Current3,4
Temperature
110
°C
Operating, junction, plastic package
140
°C
Lead, soldering (10 seconds)
300
°C
220
°C
150
°C
Operating, ambient
-20
Vapor phase soldering (1 minute)
Storage
-65
Notes:
1. Absolute maximum ratings are limiting values applied individually while all other parameters are within specified operating
conditions. Functional operation under any of these conditions is NOT implied.
2. Applied voltage must be current limited to specified range, and measured with respect to GND.
3. Forcing voltage must be limited to specified range.
4. Current is specified as conventional current, flowing into the device.
48
PRODUCT SPECIFICATION
TMC22091/TMC22191
Operating Conditions
Parameter
VDD
Power Supply Voltage
VIH
Input Voltage, Logic HIGH
VIL
Min.
Nom.
Max.
Units
4.75
5.0
5.25
V
TTL Compatible Inputs, all but TCK
2.0
VDD
V
TTL Compatible Input TCK
2.5
VDD
V
CMOS Compatible Inputs
(2/3)VDD
VDD
V
Input Voltage, Logic LOW
TTL Compatible Inputs
GND
0.8
V
CMOS Compatible Inputs
GND
(1/3)VDD
V
IOH
Output Current, Logic HIGH
-2.0
mA
IOL
Output Current, Logic LOW
4.0
mA
VREF
External Reference Voltage
IREF
D/A Converter Reference Current, VREF = Nom.
1.235
V
2.1
3.15
4.4
mA
281
392
588
Ω
(IREF = VREF / RREF, flowing out of the RREF pin)
RREF
Reference Resistor, VREF = Nom.
ROUT
Total Output Load Resistance
TA
Ambient Temperature, Still Air
Ω
37.5
0
70
°C
Pixel Interface
fPXL
Pixel Rate
12.27
15
Mpps
fPXCK
Master Clock Rate, 2x pixel rate
24.54
30
MHz
tPWHPX
PXCK Pulse Width, HIGH
10
ns
tPWLPX
PXCK Pulse Width, LOW
10
ns
For PD, VVSYNC, VHSYNC, PDC, KEY
tSP
Setup Time
12
ns
tHP
Hold Time, PD and KEY
0
ns
tHP
Hold Time, PDC, VHSYNC, VVSYNC
5
ns
tXL
Delay Time, LDV
10
ns
tPWHLDV
LDV Pulse Width, HIGH
15
ns
tPWLLDV
LDV Pulse Width, LOW
10
ns
tPWLVH
VHSYNC Pulse Width, LOW
6
15
PXCK
periods
tPWHVV
VVSYNC Pulse Width, LOW
0.5
3
H
Genlock Interface
tSGI
Setup Time, GHSYNC, GVSYNC, CVBS
10
ns
tHGI
Hold Time, GHSYNC, GVSYNC, CVBS
0
ns
Microprocessor Interface
tPWLCS
CS Pulse Width, LOW
55
ns
tPWHCS
CS Pulse Width, HIGH
30
ns
tSA
Address Setup Time
10
ns
tHA
Address Hold Time
0
ns
tSD
Data Setup Time (write)
15
ns
tHD
Data Hold Time (write)
0
ns
49
TMC22091/TMC22191
PRODUCT SPECIFICATION
Operating Conditions (continued)
Parameter
Min.
Nom.
Max.
Units
tSR
Reset Setup Time
24
ns
tHR
Reset Hold Time
2
ns
JTAG Interface
fTCK
Test Clock (TCK) Rate
tPWLTCK
TCK Pulse Width, LOW
10
ns
tPWHTCK
TCK Pulse Width, HIGH
25
ns
tSTP
Test Port Setup Time, TDI, TMS
10
ns
tHTP
Test Port Hold Time, TDI, TMS
3
ns
20
MHz
Note:
1. Timing reference points are at the 50% level.
Electrical Characteristics
Parameter
IDD
Conditions
Power Supply Current1
VDD = Max, fPXCK = 30MHz
Current1
VDD = Max, fPXCK = 30MHz
Min.
Typ.
Max.
Units
250
300
mA
60
mA
1.482
V
IDDQ
Power Supply
(D/A disabled)
VRO
Voltage Reference Output
IBR
Input Bias Current, VREF
VREF = Nom
IIH
Input Current, Logic HIGH
VDD = Max, VIN = VDD
10
µA
IIL
Input Current, Logic LOW
VDD = Max, VIN = 0V
-10
µA
VOH
Output Voltage, Logic HIGH
IOH = Max
VOL
Output Voltage, Logic LOW
IOL = Max
0.988
1.235
µA
100
2.4
V
0.4
V
IOZH
Hi-Z Leakage current, HIGH
VDD = Max, VIN = VDD
10
µA
IOZL
Hi-Z Leakage current, LOW
VDD = Max, VIN = GND
-10
µA
CI
Digital Input Capacitance
TA = 25°C, f = 1MHz
4
10
pF
CO
Digital Output Capacitance
TA = 25°C, f = 1MHz
10
VOC
Video Output Compliance Voltage
ROUT
Video Output Resistance
COUT
Video Output Capacitance
-0.3
2.0
15
IOUT = 0 mA, f = 1 MHz
15
Note:
1. Typical IDD with VDD = +5.0 Volts and TA = 25°C, Maximum IDD with VDD = +5.25 Volts and TA = 0°C.
50
pF
V
kΩ
25
pF
PRODUCT SPECIFICATION
TMC22091/TMC22191
Switching Characteristics
Parameter
Conditions
3
PIPES
Pipeline Delay
PD to Analog Out
tDOZ
Output Delay, CS to low-Z
Min.
Typ.
Max.
Units
44
44
44
PXCK
periods
23
ns
100
ns
6
Valid4
tDOM
Output Delay, CS to Data
tHOM
Output Hold Time, CS to hi-Z
tDOTP
Output Delay, TCK to TDO Valid
tHOTP
Output Hold Time, TCK to TDO
Valid
tDOS
Output Delay
PXCK to VHSYNC,
VVSYNC, PDC
tR
D/A Output Current Risetime
10% to 90% of full-scale
2
ns
tF
D/A Output Current Falltime
90% to 10% of full-scale
2
ns
tDOV
Analog Output Delay
20
ns
10
ns
30
5
ns
ns
25
ns
Notes:
1. Timing reference points are at the 50% level.
2. Analog CLOAD < 10 pF, D7-0 load < 40 pF.
3. Pipeline delay, with respect to PXCK, is a function of the phase relationship between the internally generated PCK (PXCK/2)
and PXCK, as established by the hardware reset.
4. tDOM = 1 PXCK + 54 ns = 100 ns worst-case at PXCK = 24.54 MHz.
System Performance Characteristics
Parameter
Conditions
Min.
Typ.
Max.
Units
10
10
10
Bits
RES
D/A Converter Resolution
ELI
Integral Linearity Error
0.25
%
ELD
Differential Linearity Error
0.20
%
EG
Gain Error
dp
Differential Phase
PXCK = 24.54 MHz,
40 IRE Ramp3
0.5
degree
dg
Differential Gain
PXCK = 24.54 MHz,
40 IRE Ramp3
0.9
%
SKEW
CHROMA to LUMA Output Skew
PSRR
Power Supply Rejection Ratio
±10
0
CCOMP = 0.1 µF, f = 1kHz
0.5
2
% FS
ns
%/
%VDD
Notes:
1. TTL input levels are 0.0 and 3.0 Volts, 10%-90% rise and fall times <3 ns.
2. Analog CLOAD < 10 pF, D7-0 load < 40 pF.
3. NTSC
51
TMC22091/TMC22191
PRODUCT SPECIFICATION
Video
from
Encoder
75Ω
+5V
47 µF
0.1 µF
BYPASS and OL4-0
on TMC22191 only.
OL4-0
BYPASS
PD23-0
VHSYNC
VVSYNC
PDC
KEY
AGND
TMC22x91
DIGITAL VIDEO
ENCODER
LDV
PXCK
330pF
330pF
VDDA
CHROMA
LPF
LUMA
LPF
COMPOSITE
LPF
0.1 µF
COMPOSITE VIDEO
OUTPUT
VDD
COMP
3.3KΩ
VREF
LM185-1.2
CVBS7-0
GHSYNC
GVSYNC
RREF
8
TDO
TCK
TMS
TDI
8
75Ω
+5V
S-VIDEO OUTPUT
RESET
D7-0
A1-0
CS
R/W
GENLOCK
INTERFACE
1.0µH
IN5818
27pF
100
pF
75
Ω
0.1 µF
DGND
VDO
+5V
ANALOG
INTERFACE
24
CLOCKS
FRAME
BUFFER
INTERFACE
5
47 µF
Video
Output
1.8µH
392Ω
0.1µF
2
MICROPROCESSOR
INTERFACE
JTAG TEST
INTERFACE
27007A
Figure 35. Recommended Interface Circuit
Applications Discussion
The TMC22x91 is a complex mixed-signal VLSI circuit. It
converts digital video signals at clock rates of up to 30 MHz
to analog video outputs. A recommended circuit connection
is shown in Figure 35.
References
The circuit shown in Figure 35 uses a stable external 1.235V
voltage reference. To use the internal voltage reference, simply delete the 3.3kΩ resistor and the LM185-12. A simple
voltage divider from the power supply should NOT be used,
as any variations in power supply voltage would appear
directly on the video outputs.
Filtering
An simple low-pass output reconstruction filter is shown in
Figure 36. This filter is located in the video signal path after
the COMPOSITE, LUMA, and CHROMA outputs. The
value of RREF may be varied to make up for the filter loss.
Video
from
Encoder
75Ω
1.0µH
Video
Output
1.8µH
75
Ω
75Ω
27pF
330pF
330pF
100
pF
27025A
Figure 36. Recommended Output Reconstruction Filter
52
Interface to the TMC22071 Genlocking Video
Digitizer
The TMC22x91 Digital Video Encoder has been designed to
directly interface to the TMC22071 Genlocking Video Digitizer. An interface circuit is shown in Figure 37. The microprocessor interface for TMC22x91 and TMC22071 are
similar. The R/W, RESET, D0 and A0 signals from the host
microprocessor are shared by the TMC22x91 and
TMC22071. The CS signals are separately driven from the
microprocessor bus.
Grounding Strategy
The TMC22x91 has distinctly separate analog and digital
circuits. To minimize digital crosstalk into the analog signals, the power supplies and grounds are provided over separate pins. In general, the best results are obtained by
connecting all grounds to a ground plane. Power supply pins
should be individually decoupled at the pin.
PRODUCT SPECIFICATION
TMC22091/TMC22191
CVBS7-0
GHSYNC
GVSYNC
PXCK
TMC22x91
LDV
DIGITAL VIDEO ENCODER
RESET
D7-0
A1-0
CS
R/W
8
RESET
D0
A0
CS
R/W
CVBS7-0
GHSYNC
GVSYNC
PXCK
TMC22071
LDV
GENLOCKING VIDEO DIGITIZER
2
8
MICROPROCESSOR
INTERFACE
27009A
Figure 37. TMC22x91-to-TMC22071 Interface Circuit
Printed Circuit Board Layout
Microprocessor I/O Operations
Designing with high-performance mixed-signal circuits
demands printed circuits with ground planes. Overall system
performance is strongly influenced by the board layout.
Capacitive coupling from digital to analog circuits may
result in poor picture quality. Consider the following suggestions when doing the layout:
Various CLUT Read/Write operations are shown in Table 17.
Each step in the table requires a CS pulse (falling edge followed by a rising edge) to execute.
• Keep analog traces (COMP, VREF, RREF) as short and as
far from all digital signals as possible.
• The power plane for the TMC22x91 should be separate
from that which supplies other digital circuitry. A single
power plane should be used for all of the VDD pins. If the
power supply for the TMC22x91 is the same for the
system’s digital circuitry, power to the TMC22x91 should
be filtered with ferrite beads and 0.1µF capacitors to
reduce noise.
• The ground plane should be solid, not cross-hatched.
Connections to the ground plane should be very short.
• Decoupling capacitors should be applied liberally to VDD
pins. For best results, use 0.1µF capacitor in
parallel with 47µF capacitors. Lead lengths should be
minimized. Ceramic chip capacitors are the best choice.
• The PXCK should be handled carefully. Jitter and noise
on this clock or its ground reference will translate to noise
on the video outputs. Terminate the clock line carefully to
eliminate overshoot and ringing.
For Write operations, R/W and A1-0 must conform to setup
and hold timing with respect to the falling edge of CS. D7-0
must meet setup and hold timing with respect to the rising
edge of CS. These timing relationships are illustrated in Figure 10. When writing data into an internal register (i.e.
CLUT Address Register) an extra CS falling edge is required
to transfer the input data to that register. This requirement is
usually accomplished by executing the next step in the
sequence. If there is no planned next step in the sequence,
executing a Control Register Read step will meet the requirement and terminate the sequence.
For Read operations, R/W and A1-0 must conform to setup
and hold timing with respect to the falling edge of CS. Read
data on D7-0 is initiated by the falling edge of CS\ and terminated by the rising edge of CS as shown in Figure 11. When
reading Control Registers, valid data appears tDOM after the
falling edge of CS. When reading CLUT locations, an extra
CLUT Read step is needed to set up the CLUT Read
sequence. This is accomplished in the table by executing an
extra CLUT Read step just before the CLUT Read sequence
which returns successive d, e, and f data. CLUT Read
sequences must be terminated an extra CS falling edge. This
requirement is usually accomplished by executing the next
I/O step. If there is no planned next step in the sequence,
executing a Control Register Read step will meet the requirement and terminate the sequence.
53
TMC22091/TMC22191
PRODUCT SPECIFICATION
Table 17. CLUT Read/Write Sequences
Step
R/W\
A1-0
D7-0
Function
Write Entire CLUT Starting at Address 00
1
0
01
00
Write 00 into CLUT Address Register.
1
0
01
00
Write 00 into CLUT Address Register.
2
0
11
d1
d1 written into D, CLUT address 00.
3
0
11
e1
e1 written into E, CLUT address 00.
4
0
11
f1
f1 written into F, CLUT address 00.
•••
•••
•••
•••
repeat steps 3, 4, 5 until CLUT is full.
767
0
11
d256
d256 written into D, CLUT address FF.
768
0
11
e256
e256 written into E, CLUT address FF.
769
0
11
f256
f256 written into F, CLUT address FF.
770
1
00
xx
Sequence termination.
Write CLUT Location address
1
0
01
addr
Write addr into the CLUT Address Register.
2
0
11
d1
d1 written into D, CLUT address addr.
3
0
11
e1
e1 written into E, CLUT address addr.
4
0
11
f1
f1 written into F, CLUT address addr.
5
1
00
xx
Sequence termination.
Read CLUT Location address
1
0
01
addr
2
1
11
xx
Write addr into the CLUT Address Register.
Set up for CLUT Read sequence.
3
1
11
d1
d1 read from D, CLUT address addr.
4
1
11
e1
e1 read from E, CLUT address addr.
5
1
11
f1
f1 read from F, CLUT address addr.
6
1
00
xx
Sequence termination.
Read CLUT Address Register Then Write
54
1
1
01
addr
2
0
11
d1
d1 written into D, CLUT address addr.
Read CLUT Address Register.
3
0
11
e1
e1 written into E, CLUT address addr.
4
0
11
f1
f1 written into F, CLUT address addr.
5
1
01
addr+1
Read CLUT Address Register. (terminates Write sequence)
6
0
11
d2
d2 written into D, CLUT address addr+1.
7
0
11
e2
e2 written into E, CLUT address addr+1.
8
0
11
f2
f2 written into F, CLUT address addr+1.
9
1
00
xx
Sequence termination.
PRODUCT SPECIFICATION
TMC22091/TMC22191
Table 17. CLUT Read/Write Sequences (continued)
Step
R/W\
A1-0
D7-0
Function
Read/Modify/Write CLUT Location address
1
0
01
addr
Write addr into the CLUT Address Register.
2
1
11
xx
Set up for CLUT Read.
3
1
11
d1
d1 read from D, CLUT address addr.
4
1
11
e1
e1 read from E, CLUT address addr.
5
1
11
f1
f1 read from F, CLUT address addr.
•••
•••
•••
•••
System Modifies d1, e1, f1 to d1', e1', f1'.
6
0
01
addr
7
0
11
d1’
Write addr into the CLUT Address Register. (terminates Read
sequence)
d1' written into D, CLUT address addr.
8
0
11
e1’
e1' written into E, CLUT address addr.
9
0
11
f1’
f1' written into F, CLUT address addr.
10
1
00
xx
Sequence termination.
Related Products
•
•
•
•
•
•
TMC22071 Genlocking Video Digitizer
TMC2242/2243/2246 Video Filters
TMC2249 Video Mixer
TMC2255 Convolver
TMC2272 Colorspace Converter
TMC2302 Image Manipulation Sequencer
55
TMC22091/TMC22191
Notes:
56
PRODUCT SPECIFICATION
PRODUCT SPECIFICATION
TMC22091/TMC22191
Notes:
57
TMC22091/TMC22191
PRODUCT SPECIFICATION
Mechanical Dimensions – 84-Lead PLCC Package
Inches
Symbol
Min.
A
A1
A2
B
B1
D/E
D1/E1
D3/E3
e
J
ND/NE
N
ccc
Millimeters
Max.
Min.
4.19
5.08
2.29
3.30
.51
—
.33
.53
.66
.81
30.10
30.35
29.21
29.41
25.40 BSC
1.27 BSC
.042
1.07
.056
—
1. All dimensions and tolerances conform to ANSI Y14.5M-1982.
Max.
.165
.200
.090
.130
.020
—
.013
.021
.026
.032
1.185
1.195
1.150
1.158
1.000 BSC
.050 BSC
21
84
Notes:
Notes
2. Corner and edge chamfer = 45°.
3. Dimension D1 and E1 do not include mold protrusion. Allowable
protrusion is .101" (.25mm).
3
1.42
2
21
84
.004
—
0.10
E
E1
J
D
D1
D3/E3
B1
e
A
A1
A2
58
J
B
-CLEAD COPLANARITY
ccc C
PRODUCT SPECIFICATION
TMC22091/TMC22191
Mechanical Dimensions
100 Lead MQFP Package – 3.2mm Footprint
Inches
Symbol
Notes:
Millimeters
Notes
Min.
Max.
Min.
Max.
A
A1
A2
B
—
.010
.100
.008
—
.25
2.55
.22
C
D
.005
.904
.134
—
.120
.015
.009
.923
3.40
—
3.05
.38
.23
23.45
D1
E
E1
e
L
N
ND
NE
.783
.791
.667
.687
.547
.555
.0256 BSC
.028
.040
100
30
20
α
ccc
0°
—
7°
.004
.13
22.95
2. Controlling dimension is millimeters.
3. Dimension "B" does not include dambar protrusion. Allowable
dambar protrusion shall be .08mm (.003in.) maximum in excess of
the "B" dimension. Dambar cannot be located on the lower radius
or the foot.
3, 5
5
4. "L" is the length of terminal for soldering to a substrate.
5. "B" & "C" includes lead finish thickness.
19.90
20.10
16.95
17.45
13.90
14.10
.65 BSC
.73
1.03
100
30
20
0°
—
1. All dimensions and tolerances conform to ANSI Y14.5M-1982.
4
7°
.12
D
.20 (.008) Min.
0° Min.
.13 (.30)
R
.005 (.012)
D1
Datum Plane
B
C
E1
α
.13 (.005) R Min.
Pin 1 Indentifier
E
L
e
0.076" (1.95mm) Ref
Lead Detail
See Lead Detail
Base Plane
A A2
B
A1
Seating Plane
-CLead Coplanarity
ccc C
59
TMC22091/TMC22191
PRODUCT SPECIFICATION
Ordering Information
Product Number
Temperature Range
Screening
Package
Package Marking
TMC22091KHC
TA = 0°C to 70°C
Commercial
100-Lead MQFP
22091KHC
TMC22091R0C
TA = 0°C to 70°C
Commercial
84-Lead PLCC
22091R0C
TMC22191KHC
TA = 0°C to 70°C
Commercial
100-Lead MQFP
22191KHC
TMC22191R0C
TA = 0°C to 70°C
Commercial
84-Lead PLCC
22191R0C
LIFE SUPPORT POLICY
FAIRCHILD’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES
OR SYSTEMS WITHOUT THE EXPRESS WRITTEN APPROVAL OF THE PRESIDENT OF FAIRCHILD SEMICONDUCTOR
CORPORATION. As used herein:
1. Life support devices or systems are devices or systems
which, (a) are intended for surgical implant into the body,
or (b) support or sustain life, and (c) whose failure to
perform when properly used in accordance with
instructions for use provided in the labeling, can be
reasonably expected to result in a significant injury of the
user.
2. A critical component in any component of a life support
device or system whose failure to perform can be
reasonably expected to cause the failure of the life support
device or system, or to affect its safety or effectiveness.
www.fairchildsemi.com
6/12/98 0.0m 002
Stock# DS70022091
 1998 Fairchild Semiconductor Corporation