T exas Instruments (TI) reserves the right to make changes to its products or to discontinue any
semiconductor product or service without notice, and advises its customers to obtain the latest
version of relevant information to verify , before placing orders, that the information being relied
on is current.
TI warrants performance of its semiconductor products and related software to the specifications
applicable at the time of sale in accordance with TI’s standard warranty . T esting and other quality
control techniques are utilized to the extent TI deems necessary to support this warranty.
Specific testing of all parameters of each device is not necessarily performed, except those
mandated by government requirements.
Certain applications using semiconductor products may involve potential risks of death,
personal injury , or severe property or environmental damage (“Critical Applications”).
TI SEMICONDUCTOR PRODUCTS ARE NOT DESIGNED, INTENDED, AUTHORIZED, OR
WARRANTED TO BE SUITABLE FOR USE IN LIFE-SUPPORT APPLICATIONS, DEVICES
OR SYSTEMS OR OTHER CRITICAL APPLICATIONS.
Inclusion of TI products in such applications is understood to be fully at the risk of the customer.
Use of TI products in such applications requires the written approval of an appropriate TI officer .
Questions concerning potential risk applications should be directed to TI through a local SC
sales office.
In order to minimize risks associated with the customer’s applications, adequate design and
operating safeguards should be provided by the customer to minimize inherent or procedural
hazards.
TI assumes no liability for applications assistance, customer product design, software
performance, or infringement of patents or services described herein. Nor does TI warrant or
represent that any license, either express or implied, is granted under any patent right, copyright,
mask work right, or other intellectual property right of TI covering or relating to any combination,
machine, or process in which such semiconductor products or services might be or are used.
The TVP3010C and the TVP3010M palettes are commercial and military versions, respectively, of an
advanced Video Interface Palette (VIP) from Texas Instruments implemented in the EPIC 0.8-micron
CMOS process. Differences between the two versions are outlined in separate tables. In both versions,
maximum flexibility is provided by the pixel multiplexing scheme. The scheme accommodates 64-, 32-, 16-,
8-, and 4-bit pixel buses without any circuit modification. This enables the system to be easily reconfigured
for varying amounts of available video RAM. The device supports selection of little- or big-endian data format
for the pixel-bus/frame-buffer interface. Data can be split into 1, 2, 4, or 8 bit-planes for pseudo-color mode
or split into 12-, 16- or 24-bit true-color and direct-color modes. For the 24-bit direct-color modes, an 8-bit
overlay plane is available. The 16-bit direct-color and true-color modes can be configured to IBM XGA
(5, 6, 5), T ARGA (5, 5, 5, 1), or (6, 6, 4) as another existing format. An additional 12-bit mode (4, 4, 4, 4)
is supported with 4 bits for each color and overlay. An on-chip, IBM XGA-compatible hardware cursor is
incorporated so that further increases in graphics-system performance are possible. Both devices are
software compatible with the INMOS IMSG176/8 and Brooktree Bt476/8 color palettes.
An internal-frequency doubler is incorporated, allowing convenient and cost-effective clock-source
alternatives to be utilized. An auxiliary windowing function and a pixel-port select function are provided so
that overlay or VGA graphics can be displayed on top of direct color inside or outside a specified auxiliary
window. Color-keyed switching of direct color and overlay is also supported.
Clocking is provided through one of five TTL inputs, CLK0–CLK4, and is software selectable. Additionally,
CLK1/CLK2 and CLK3/CLK4 can be selected as differential ECL clock sources. The video, shift-clock, and
reference-clock outputs provide a software-selected divide ratio of the chosen clock input. The reference
clock can optionally be provided as an output on CLK3, and a data-latch clock can optionally be input on
CLK4.
The TVP3010C and the TVP3010M have three 256 8 color look-up tables with triple 8-bit video
digital-to-analog converters (DACs) capable of directly driving a doubly-terminated 75-Ω line. The lookup
tables are designed with a dual-ported RAM architecture that enables ultra-high speed operation. Sync
generation is incorporated on the green output channel. Horizontal sync and vertical sync are fed through
the device and optionally inverted to indicate screen resolution to the monitor. A palette-page register
provides the additional bits of palette address when 1, 2, or 4 bit-planes are used. This allows the screen
colors to be changed with only one microprocessor-interface unit (MPU) write cycle.
Each device features a separate VGA bus that allows data from the feature connector of most
VGA-supported personal computers to be fed directly into the palette without the need for external data
multiplexing. This allows a replacement graphics board to remain downwards compatible by utilizing the
existing graphics circuitry often located on the motherboard.
Both the TVP3010 VIP and the TVP3010M VIP are highly system integrated. Either device can be
connected to the serial port of a VRAM device without external buffer logic and each device can be
connected to many graphics engines directly. The split shift register transfer function, which is supported
by VRAM, is also supported by the TVP3010C and TVP3010M.
The system-integration concept is carried to manufacturing testing and field diagnosis levels. To support
these testing and diagnostic levels, several highly-integrated test functions have been designed to enable
simplified testing of the palette, the graphics board, and the graphics system.
EPIC is a trademark of Texas Instruments Incorporated.
XGA is a registered trademark of International Business Machines Corporation.
TARGA is a registered trademark of Truevision Incorporated.
INMOS is a trademark of INMOS International Limited.
Brooktree is a trademark of Brooktree Corporation.
1–1
The TVP3010C and TVP3010M are 32-bit devices and both are pin compatible with the TLC3407X VIP,
allowing convenient performance upgrades when using devices in the TI Video Interface Palette family.
NOTE:
The TVP3010C and TVP3010M include circuits that are patented as well as circuit
designs that have patents pending.
CLK3[RCLK]74M1I/ODot clock 3 TTL input or reference clock output. When
CLK3[LCLK]73L3IDot clock 4 TTL input or pixel-port latch clock. CLK3[LCLK]
COMP52K11ICompensation. COMP provides compensation for the
NOTE 1: All unused inputs should be tied to a logic level and not be allowed to float.
55, 57J1, L11, G12Analog power. All AVDD terminals must be connected.
(TTL
compatible)
(TTL/ECL
compatible)
Dot clock 0 input. CLK0 can be selected to drive the dot
clock at frequencies up to 140 MHz. When VGA mode is
active, the default clock source is CLK0. The maximum
frequency in VGA mode is 85 MHz.
Dual-mode dot clock input. These inputs are essentially
ECL-compatible inputs, but two TTL clocks may be used
on the CLK1 and CLK2 if so selected in the input clock
select register. These inputs may be selected as the dot
clock up to the device limit while in the ECL mode or up to
140 MHz in the TTL mode.
configured as CLK3, this terminal is similar to CLK0 and
can be selected to drive the dot clock at frequencies up to
140 MHz. When configured as RCLK, this terminal outputs
the reference clock signal, which is similar to the SCLK
signal but not gated off during blanking. This signal can be
used for pixel-port timing reference or other system
synchronization. The terminal defaults to CLK3 after reset.
can be configured to drive dot clock frequencies up to 140
MHz, or it can be configured as a latch-clock input to latch
pixel-port input data. It defaults to CLK4 after reset, and
LCLK is internally connected to RCLK to latch pixel-port
data.
internal reference amplifier. A 0.1-µF ceramic capacitor is
required between COMP and A VDD. The COMP capacitor
must be as close to the device as possible to avoid noise
pick up.
1–7
1.5Terminal Functions (TVP3010C and TVP3010M) Continued
I/O
DESCRIPTION
TERMINAL
NAMENO. (FN)NO. (GA)
DV
DD
D(0–7)36–43B12, C12,
FS ADJUST51L12IFull-scale adjustment. A resistor connected between
GND44, 54,
HSYNCOUT46H12O
IOR, IOG, IOB48, 49, 50J12, J11,
MUXOUT [SENSE]63M7O
P(0–31)1–29,
NOTE 1: All unused inputs should be tied to a logic level and not be allowed to float.
45, 81M11Digital power. All DVDD terminals must be connected
I/O
(TTL
compatible)
(TTL
compatible)
OAnalog current outputs. These outputs can drive a
together.
MPU interface data bus. Data terminals are used to
transfer data in and out of the register map and
palette/overlay RAM.
FS ADJUST and ground controls the full-scale range
of the DACs.
Ground. All GND terminals must be connected. The
GNDs are connected internally.
Horizontal sync output after pipeline delay. For
system mode the horizontal-sync output can be
programmed, but for the VGA mode the output
carries the same polarity as the input.
37.5-Ω load directly (doubly terminated 75-Ω line),
thus eliminating the requirement for any external
buffering.
Multiplexer output control or DAC comparator output
signal. When MUXOUT
plexer output control, it is software programmable
through the configuration register. When the
multiplexer control register is set to VGA mode, this
output terminal and corresponding configuration
register bit are set low to indicate to external devices
that the VGA pass-through mode is being used.
Alternatively , SENSE
comparator output. In this case, the SENSE
when one or more of the DAC output analog levels is
above the internal comparator reference of 350 mV
"
50 mV.
I
Pixel input port. The port can be used in various
modes as shown in the multiplexer control register.
All the unused terminals need to be tied to GND.
is configured as a multi-
can be configured as the DAC
is low
1–8
1.5Terminal Functions (TVP3010C and TVP3010M) Continued
I/O
DESCRIPTION
TERMINAL
NAMENO. (FN)NO. (GA)
REF53M12V oltage reference for DACs. An internal voltage reference
RD31B10I
RS(0–2)32–34A12, C10,
RS3 [PSEL]35C11I
SCLK79K1O
SFLAG62M8I
SYSBL60M9I
HSYNC,
VSYNC
VCLK78L1O
VGABL61L8I
VGA(0–7)65–72M6, L6, M5,
NOTE 1: All unused inputs should be tied to a logic level and not be allowed to float.
58, 59M10, L9I
B11
L5, M4, L4,
M3, M2
(TTL
compatible)
(TTL
compatible)
(TTL
compatible)
(TTL
compatible)
(TTL
compatible)
(TTL
compatible)
(TTL
compatible)
(TTL
compatible)
(TTL
capability)
(TTL
capability)
of nominally 1.235 V is provided, which requires an
external 0.1-µF ceramic capacitor between REF and
analog GND. However, the internal reference voltage can
be overdriven by an externally supplied reference voltage.
A typical connection is shown in Appendix A.
Read strobe inputs When cleared to 0, RD initiates a
read from the register map. Reads are performed
asynchronously and are initiated on the low-going edge of
RD
(see Figure 3–1).
I
Register-select inputs. The RS terminals specify the
location in the register map that is to be accessed (see
Table 2–1).
Register-select input or port-select input. When configured
as the RS3 input, this terminal has no effect. When
configured as the port-select input, RS3 [PSEL] allows the
creation of VGA or overlay windows in a direct-color
background on a pixel-by-pixel basis.
Shift clock output. SCLK is selected as a division of the dot
clock input. The output signals are gated off during
blanking, although SCLK is still used internally to
synchronize with the activation of Blank
Split shift register transfer flag. The TVP3010 detects a
low-to-high transition on SFLAG during a blanking
sequence and immediately generates an SCLK pulse. This
early SCLK pulse replaces the first SCLK pulse in the
normal sequence.
System blank input. SYSBL is active (low).
Horizontal and vertical sync inputs. These signals
generate the sync level on the green current output. They
are active (low) inputs, but the HSYNCOUT and
VSYNCOUT outputs can be programmed through the
general control register.
Video clock output. VCLK is the user-programmable
output for synchronization to the graphics processor.
VGA blank input. VGABL is active (low).
I
VGA pass-through bus. These buses can be selected as
the pixel bus for VGA mode, but it does not allow for any
multiplexing.
.
1–9
1.5Terminal Functions (TVP3010C and TVP3010M) Continued
I/O
DESCRIPTION
TERMINAL
NAMENO. (FN)NO. (GA)
VSYNCOUT47H11O
(TTL
capability)
WR30A11I
(TTL
capability)
8/6 [OVS]64L7I
(TTL
capability)
NOTE 1: All unused inputs should be tied to a logic level and not be allowed to float.
Vertical sync output after pipeline delay. For system
mode, the output can be programmed, but for the
VGA mode the output carries the same polarity as
the input.
Write strobe input. A low on WR initiates a write to the
register map. As with RD
asynchronous and initiated on the low-going edge of
WR
, (see Figure 3–1).
DAC resolution selection or overscan input. The 8/6
terminal selects the data-bus width (8 or 6 bits) for
the DAC and is essentially provided in order to
maintain compatibility with the IMSG176. When 8/6
[OVS] is high, 8-bit bus transfers are used with D7
the MSB and D0 the LSB. For 6-bit bus operation,
while the color palette still has the 8-bit information,
D5 shifts to the bit 7 position with D0 shifted to the bit
2 position and the 2 LSBs are filled with zeros at the
output multiplexer to DAC. The palette-holding
register zeroes the two MSBs when it is read in the
6-bit mode. The terminal can also be configured to
function as the overscan input facilitating the
creation of custom screen borders. This terminal
defaults to 8/6
after reset.
, write transfers are
1–10
2 Detailed Description
The TVP3010C and TVP3010M VIPs are identical in their operation. Both the TVP3010C and TVP3010M
are 32-bit devices; both devices are terminal compatible with the TLC34076 and each device offers
advanced features. To facilitate the enhanced functionality, some terminals have dual functions. The
dual-function terminals are controlled by the configuration register discussed in subsection 2.16.1. At reset,
all pins default to the TLC34076 terminal functions.
2.1MPU Interface
The microprocessor unit (MPU) interface is controlled using read and write strobes (RD, WR), three
register-select terminals [RS(0 – 2)], and the 8/6
6-bit-wide data path to the color-palette RAM and is provided in order to maintain compatibility with the
IMSG176. Since the 8/6
[OVS] pin is a dual-function pin, 2 bits are provided in the configuration register to
control this function. Configuration-register bit 1 determines whether the 8/6 [OVS] pin operates as 8/6 or
OVS. If configuration register bit 1 is cleared to 0 (default), then 8/6
held low, data on the lowest 6 bits of the data bus are internally shifted up by 2 bits to occupy the upper
8/6
6 bits at the output multiplexer and the bottom 2 bits are then cleared to 0. This operation is carried out in
order to utilize the maximum range of the DACs.
The direct register map is shown in Table 2–1. Extended registers can be accessed through the index
register. The index register map is shown in Table 2–2. In general, the index register must first be loaded
with the target address value. Successive reads or writes from and to the data register then access the target
location. The MPU interface operates asynchronously, with data transfers being synchronized by internal
logic.
RS3 is a do not care for register addressing but is used as the PSEL input (see
Section 2.6).
T able 2–1. Direct Register Map
RS2RS1RS0REGISTER ADDRESSED BY MPUR/WDEFAULT (HEX)
-select terminal. The 8/6 pin selects between an 8- or
operation is controlled by the pin. With
NOTE:
2–1
T able 2–2. Indirect Register Map (Extended Registers)
INDEX REGISTER
(HEX)
00R/W00Cursor Position X LSB
01R/W00Cursor Position X MSB
02R/W00Cursor Position Y LSB
03R/W00Cursor Position Y MSB
04R/W1FSprite Origin X
05R/W1FSprite Origin Y
06R/W00Cursor Control Register
07Reserved
08WXXCursor RAM Address LSB
09WXXCursor RAM Address MSB
0AR/WXXCursor RAM Data
0BReserved
0C–0FReserved-Undefined
10R/WXXWindow Start X LSB
11R/WXXWindow Start X MSB
12R/WXXWindow Stop X LSB
13R/WXXWindow Stop X MSB
14R/WXXWindow Start Y LSB
15R/WXXWindow Start Y MSB
16R/WXXWindow Stop Y LSB
17R/WXXWindow Stop Y MSB
18R/W80Multiplexer Control Register 1
19R/W98Multiplexer Control Register 2
1AR/W00Input-Clock Selection Register
1BR/W3EOutput-Clock Selection Register
1CR/W00Palette Page Register
1DR/W20General Control Register
1ER/W00Configuration Register
1FReserved-Undefined
20R/WXXOverscan Color Red
21R/WXXOverscan Color Green
NOTE 1: Reserved registers should be avoided; otherwise, circuit behavior could deviate
from that specified. Reserved-undefined registers are nonexistent locations on
the register map.
R/W
DEFAULT
(HEX)
REGISTER ADDRESSED
BY INDEX REGISTER
2–2
T able 2–2. Indirect Register Map (Extended Registers) (Continued)
NOTE 1: Reserved registers should be avoided; otherwise, circuit behavior could deviate
from that specified. Reserved-undefined registers are nonexistent locations on
the register map.
R/W
DEFAULT
(HEX)
REGISTER ADDRESSED
BY INDEX REGISTER
2–3
2.2Color Palette
The color palette is addressed by an internal 8-bit address register for reading/writing data from/to the RAM.
This register is automatically incremented following a RAM transfer, allowing the entire palette to be
read/written with only one access of the address register. When the address register increments beyond
the last location in RAM, it is reset to the first location (address 0). All read and write accesses to the RAM
are asynchronous to SCLK, VCLK, and dot clock but performed within one dot clock. Therefore, they do not
cause any noticeable disturbance on the display .
The color RAM is 24 bits wide for each location and 8 bits wide for each color. Since the MPU access is
8 bits wide, the color data stored in the palette is 8 bits even when the 6-bit mode is chosen
= 0). If the 6-bit mode is chosen, the 2 MSBs of color data in the palette have the values previously
(8/6
written. However, if they are read back in the 6-bit mode, the 2 MSBs are 0s to be compatible with IMSG176
and Bt176. The output multiplexer shifts the six LSB to the six MSB positions and fills the 2 LSBs with 0s
after the color palette. The multiplexer then feeds the data to the DAC. The test register and the CRC
calculation both take data after the output multiplexer, enabling total system verification. The color-palette
access is described in the following two sections, and it is fully compatible with IMSG176/8 and Bt476/8.
2.2.1Writing to Color-Palette RAM
T o load the color palette, the MPU must first write to the address register (write mode) with the address where
the modification is to start. This is then followed by three successive writes to the palette-holding register
with 8 bits of red, green, and blue data. After the blue write cycle, the three bytes of color data are
concatenated into a 24-bit word that is then written to the RAM location specified by the address register.
The address register then increments to the next location, which the MPU may modify by simply writing
another sequence of red, green, and blue data. A block of color values in consecutive locations may be
written to by writing the start address and performing continuous red, green, and blue write cycles until the
entire block has been written.
2.2.2Reading From Color-Palette RAM
Reading from the palette is performed by writing to the address register (read mode) with the location to be
read. This then initiates a transfer from the palette RAM into the holding register, followed by an increment
of the address register. Three successive MPU reads from the holding register produce red, green, and blue
color data (6 or 8 bits depending on the 8/6
the contents of the color-palette RAM at the address specified by the address register are copied into the
holding register and the address register is again incremented. As with writing to the palette, a block of color
values in consecutive locations may be read by writing the start address and performing continuous red,
green, and blue read-cycles until the entire block has been read. Since the color-palette RAM is dual ported,
the RAM may be read during active display without disturbing the video.
mode) for the specified location. Following the blue read-cycle,
2.2.3Palette Page Register
The palette page register appears as an 8-bit register on the extended register map (see Section 2.1). Its
purpose is to provide high-speed color changing by removing the need for palette reloading. When using
1, 2, or 4 bit-planes, the additional planes are provided from the page register. When using four bit-planes,
the pixel inputs specify the lower 4 bits of the palette address with the upper 4 bits specified from the page
register. This gives the user the capability of selecting from 16 palette pages with only one-chip access, thus
allowing all the screen colors to be changed at the line frequency. A bit-to-bit correspondence is used;
therefore, in the above configuration, page-register bits 7 through 4 map onto palette-address bits 7 through
4, respectively. This is illustrated in Table 2–3.
2–4
NOTE:
The additional bits from the page register are inserted after the read mask.
The palette page register specifies the additional bit-planes for the overlay field in
direct-color modes with less than 8 bits per pixel overlay .
T able 2–3. Allocation of Palette Page Register Bits
The read-mask register is an 8-bit register used to enable or disable a bit-plane from addressing the
color-palette RAM in the pseudo-color modes. Each palette address bit is logically ANDed with the
corresponding bit from the read mask register before going to the palette page register and addressing the
palette RAM.
In order to provide maximum flexibility to control palette data, the read mask operation is performed before
the addition of the page register bits. Therefore, care must be taken in those modes that have less than 8
bits per pixel of pseudo-color or overlay data. Be aware of the palette page register settings in these modes.
2.3Clock Selection and Output-Clock (SCLK, RCLK, and VCLK) Generation
The TVP3010C and the TVP3010M VIP provide a maximum of five clock inputs. CLK0 is dedicated as a
TTL input. The other four clock inputs can be selected as either two differential ECL input or two extra TTL
inputs. The TTL inputs can be used for video rates up to 140 MHz. The dual-mode clock input (ECL/TTL)
is primarily an ECL input but can be used as TTL-compatible inputs if the input-clock selection register is
so programmed. The clock source used at power up is CLK0; an alternative source can be selected by
software during normal operation. This chosen clock input can be used unmodified as the dot clock
(representing pixel rate to the monitor). Alternatively , when the input-clock selection register is programmed
to use the internal frequency-doubler , the chosen clock source is used as a reference for multiplication. Each
device also allows for user programming of RCLK, SCLK and VCLK outputs (reference, shift and video
clocks) by using the output-clock selection register. The input-clock and output-clock selection registers are
located in the indirect register map (see Table 2–2).
The ECL inputs can be used as differential or single-ended inputs. When CLK1 or CLK3 is used as a
single-ended ECL input, CLK2 or CLK4 needs to be externally terminated to set the input common-mode
signal level. This can be done with a simple resistor divider, as is the case with fully dif ferential ECL. Care
needs to be taken when choosing the resistor values to ensure that the dc level on CLK2 or CLK4 is in the
middle of the CLK1 or CLK3 ECL-input signal range.
2.3.1RCLK, SCLK, VCLK
Both VIP devices provide a user-programmable reference clock (RCLK), a shift clock (SCLK), and video
(VCLK) clock outputs that can be set as divisions of the dot clock. RCLK is a continuously-running reference
clock and is not disabled during the Blank signal. RCLK can be selected as divisions of 1, 2, 4, 8, 16, 32 or
64 of the
2–5
dot clock (see Table 2–4). It is provided as a clock reference and is typically connected back to the LCLK
FUNCTION (
5)
input to latch pixel-port data. Since pixel-port data is latched on the rising edge of LCLK, the RCLK frequency
must be set as a function of the desired multiplexing ratio (that depends on the pixel-bus width and number
of bit-planes, see Section 2.4).
T able 2–4. Output-Clock Selection Register Format
OUTPUT-CLOCK SELECTION-REGISTER BITS (see Note 2)
6543210
000xxxVCLK/1 output ratio
001xxxVCLK/2 output ratio
010xxxVCLK/4 output ration
011xxxVCLK/8 output ratio
100xxxVCLK/16 output ratio
101xxxVCLK/32 output ratio
110xxxVCLK/64 output ratio
111xxxVCLK output held at logic 1
xxx000RCLK/1 output ratio (see Notes 2 and 5)
xxx001RCLK/2 output ratio (see Notes 2 and 5)
xxx010RCLK/4 output ratio (see Notes 2 and 5)
xxx011RCLK/8 output ratio (see Notes 2 and 5)
xxx100RCLK/16 output ratio (see Notes 2 and 5)
xxx101RCLK/32 output ratio (see Notes 2 and 5)
xxx110RCLK/64 output ratio (see Notes 2 and 5)
0xxx110RCLK/64, SCLK output held at logic 0
0xxx111RCLK, SCLK outputs held at logic 0
x111111Clock counter reset (6)
†
These lines indicate the reset conditions as required for VGA pass-through.
NOTES: 2. Register bit 6 enables (1) and disables (default = 0) the SCLK output buffer. Register bit 7 is a don’t
care bit.
3. When the clocks are selected from one mode to the other, a minimum of 30 ns is needed before
the new clocks are stabilized and running.
4. When the output-clock-selection register is written with 3F (hex), the clock counter is reset,
RCLK = SCLK = 0, and VCLK = 1.
5. SCLK is the same as RCLK except that it is disabled during blank. When the RCLK divide ratio is
chosen, this sets the SCLK ratio as well.
see Notes 2, 3, 4, and
†
†
SCLK is the same as RCLK but disabled during the Blank active period. SCLK is designed to be used as
the shift clock to interface directly with the VRAM. If SCLK is not used, the output can be switched off and
held low to protect against VRAM lockup due to invalid SCLK frequencies. The detailed SCLK control timing
is discussed in subsection 2.3.2.
VCLK is designed to be used as the timing reference by the graphics processor or other custom-designed
control logic to generate the graphics system control signals (SYSBL
, HSYNC, and VSYNC). VCLK can be
selected as divisions of 1, 2, 4, 8, 16, 32, or 64 of the dot clock and can also be held at high (see T able 2–4).
The default setup is VCLK held at high since it is not used in VGA pass-through mode. Since these control
signals are sampled by VCLK, VCLK must be enabled for these to function properly .
Even though RCLK/SCLK and VCLK can be selected independently, there is still a relationship between
the two as discussed below. Many system considerations have been carefully covered in their design,
leaving maximum freedom to the user.
2–6
Internally , RCLK, SCLK, and VCLK are generated from a common clock counter that is counted at the rising
edge of the dot clock. Therefore, when VCLK is enabled, it is naturally in phase with RCLK and SCLK as
shown in Figure 2–1.
Normally, the video-control signal inputs HSYNC
, VSYNC, and SYSBL are latched on the falling edge of
VCLK when in a non-VGA mode. When the configuration register is programmed for opposite VCLK polarity ,
these video-control signals are latched on the rising edge of VCLK.
The internal clock counter is initialized any time the output-clock selection register is written with 3F (hex).
This provides a simple mechanism to synchronize multiple palettes or system devices by providing a known
phase relationship for the various system clocks. It is left up to the user to provide some means of disabling
the dot-clock input to the part while this reset is occurring if multiple parts are to be synchronized.
The reset default divide ratio for RCLK is 64:1 with SCLK held low and VCLK held at high. When choosing
certain video timing parameters, exercise caution if the selected RCLK frequency is less than the selected
VCLK frequency (see Appendix B for a more detailed discussion).
Dot Clock
(dot clock/4 as an example)
RCLK = SCLK
(dot clock/2 as an example)
VCLK
Figure 2–1. Dot Clock/VCLK/RCLK/SCLK Relationship
The input-clock-selection register selects the desired input-clock source. T able 2–5 details how to program
the various options.
T able 2–5. Input-Clock Selection Register
INPUT-CLOCK-SELECT REGISTER
†
CLK0 is chosen at reset as required for VGA pass-through.
NOTES: 6. Register bits 3 and 7 are don’t-care bits.
(HEX) (see Note 6)
00Select CLK0 as TTL-clock source
01Select CLK1 as TTL-clock source
02Select CLK2 as TTL-clock source
03Select CLK3 as TTL-clock source
04Select CLK4 as TTL-clock source
06Select CLK3/CLK4 as ECL-clock source up to 140 MHz
07Select CLK1/CLK2 as ECL-clock source up to device limit
10Select CLK0 as doubled TTL-clock source
11Select CLK1 as doubled TTL-clock source
12Select CLK2 as doubled TTL-clock source
13Select CLK3 as doubled TTL-clock source
14Select CLK4 as doubled TTL-clock source
16Select CLK3/CLK4 as doubled ECL-clock source
17Select CLK1/CLK2 as doubled ECL-clock source
7. Register bits 5 and 6 are reserved.
8. When the clocks are selected from one input clock source to another , a minimum of 30 ns is needed before
the new clocks are stabilized and running.
FUNCTION (see Note 7)
†
2–7
The output-clock-selection register is used to program the desired divided-down frequencies for the
reference/shift and video clocks.
2.3.2Frame-Buffer Clocking: Self-Clocked or Externally Clocked
The TVP3010C and the TVP3010M have two pixel-data latching modes, allowing for flexibility in the
frame-buffer interface timing. For the pixel port P(0–31), data is always latched on the rising edge of LCLK.
If auxiliary-control register (ACR) bit 3 is set to 1 (default), the internal circuitry is configured for self-clocked
mode. In this mode, the RCLK or SCLK output of the palette must be used as the timing reference to present
data to the pixel port P(0–31). In self-clocked mode, RCLK can be directly tied back to LCLK or LCLK can
be a delayed version of RCLK within the timing requirements of the VIP. The self-clocked mode of
frame-buffer latching is similar to the operation of the TLC3407X video-interface palette devices.
The VIP internal Blank signal is generated from either VGABL
port is enabled (multiplexer control register 2 (MCR2) bit 7 = 1) or disabled (MCR2 bit 7 = 0). The rising edge
of CLK0 latches VGABL
latch the SYSBL
input when the VGA port is disabled. When the internal Blank signal becomes active, SCLK
when the VGA port is enabled. The falling edge of VCLK is used to sample and
is disabled as soon as possible. For example, if SCLK is high when the sampled SYSBL
is allowed to complete the clock cycle and return to the low state. SCLK then is held low until the sampled
signal goes back high. At this time, SCLK is enabled to clock the first pixel data valid from VRAM.
SYSBL
The VIP video-blanking circuitry is designed with sufficient pipeline delay to allow the internal sampled
and VGABL signals to align with the pipelined RGB data to the video DACs. The logic described
SYSBL
previously works in situations where the SCLK period is shorter than, equal to, or longer than the VCLK
period.
When in the self-clocked mode, the SCLK control timing is designed to interface directly with the external
VRAM. The shift register in the system VRAM is supposed to be updated during the blank active period.
When the SYSBL
input is sampled high by the falling edge of VCLK, the VRAM shift clock (SCLK) is restarted
to clock the VRAM and enable the first group of pixel data to appear on the pixel bus as well as at the
TVP3010 pixel input port. The second SCLK causes the VRAM shift register to shift out the second group
of data. At the same time, LCLK latches the first group of pixel data into the VIP (see Figure 2–2 for a detailed
timing-diagram).
or SYSBL, depending on whether the VGA
goes low, SCLK
VCLK
In Phase
at Input Terminal
Internal Delayed
before dot-clock
at Input Terminal
SYSBL
LD
LCLK = RCLK
Blank
(internal signal
pipeline delay)
Pixel Data
SCLK
Latch Last Group
of Pixel Data
Last Group of Pixel Data
Latch First Group of Pixel Data
1st
2nd
3rd
Group
Group
Group
4th
Group
5th
Group
Latch Last Group
of Pixel Data
6th
Group
Figure 2–2. SCLK/VCLK Control Timing
(SSRT Disabled, RCLK/SCLK Frequency = VCLK Frequency)
The RCLK /SCLK phase relationship is designed so that timing specifications are satisfied for the case
where SCLK is driving a typical 2-MB VRAM load and RCLK is connected to LCLK. If an external buffer is
2–8
required on SCLK so that it can drive a larger load, a similar buffer can be placed on RCLK to match the
signal delay before connecting to LCLK. However, the delay from LCLK to RCLK cannot exceed one RCLK
period –7 ns, (see the timing-parameter specifications for more details).
When the VRAM split shift register operation is performed (see Figure 2–3 and Figure 2–4), the SCLK timing
is adjusted to work with the SFLAG input. Basically , the split shift register operation inserts an SCLK during
the blank period. This causes the first group of pixel data to appear at the pixel port during blank and allows
the first group of data to be displayed as soon as the palette comes out of blank. Figures 2–3 and 2–5 show
the case when the SSRT (split shift register transfer) function is enabled. When a rising edge occurs on the
SFLAG input, one SCLK with a minimum 15-ns pulse duration is generated after the specified delay . Since
this is designed to meet VRAM timing requirements, the SSRT -generated SCLK replaces the first SCLK in
the regular shift register transfer case as described above (see Section 2.15 for a detailed explanation of
the SSRT function).
VCLK
In Phase
at Input Terminal
SYSBL
SFLAG Input
LD
Internal Delayed
LCLK = RCLK
Blank
(internal signal
before dot-clock
pipeline delay)
Pixel Data
at Input Terminal
SCLK
Latch Last Group
of Pixel Data
Last
Group
1st Group of
SCLK Between Split Shift-Register and Regular Shift-Register Transfer
Pixel Data
Latch First Group of Pixel Data
2nd
3rd
Group
Group
4th
Group
5th
Group
Figure 2–3. SCLK/VCLK Control Timing
(SSRT Enabled, RCLK/SCLK Frequency = VCLK Frequency)
Latch Last Group
of Pixel Data
6th
Group
2–9
VCLK
In Phase
at Input Terminal
SYSBL
Internal Delayed
LCLK = RCLK
Blank
(internal signal
before dot-clock
pipeline delay)
Pixel Data
at Input Terminal
SCLK
VCLK
In Phase
SYSBL
at Input Terminal
Latch Last Group
of Pixel Data
LD
Latch first Group of Pixel Data
1st
2nd
Last Group of Pixel Data
Group
Group
3rd
Group
Group
Figure 2–4. SCLK/VCLK Control Timing
(SSRT Disabled, RCLK/SCLK Frequency = 4 x VCLK Frequency)
4th
5th
Group
6th
Group
7th
Group
SFLAG Input
Internal Delayed
LCLK = RCLK
(internal signal
before dot-clock
pipeline delay)
Pixel Data
at Input Terminal
2–10
Latch Last Group
LD
Blank
of Pixel Data
Latch First Group of Pixel Data
Last
Group
First Group of Pixel Data
SCLK Between Split Shift-Register Transfer and Regular Shift-Register Transfer
2nd
Group
Group
SCLK
Figure 2–5. SCLK/VCLK Control Timing
(SSRT Enabled, RCLK/SCLK Frequency = 4 x VCLK Frequency)
Latch Second Group
of Pixel Data
4th
5th
Group
3rd
Group
6th
Group
7th
Group
Externally clocked timing can be chosen for the pixel bus P(10–31) by clearing auxiliary control register bit 3
to 0. In externally clocked mode, the RCLK or SCLK output of the palette is not used as the timing reference
to present data to the pixel bus. Instead, pixel data is presented to the palette with a synchronous clock and
all palette timing is referenced to this clock. In this mode, the external clock should be connected to LCLK
and the selected clock input. (When the VGA port is enabled, the CLK0 input is selected independent of the
input-clock selection register.)
The externally clocked frame-buffer interface mode is intended for applications where windowed or
pixel-by-pixel switching between the VGA port and the pixel port is desired in non-VRAM-based graphics
systems. In such applications, the VGA port is enabled (multiplexer control register bit 7 set to 1) and the
appropriate direct-color mode is set in the multiplexer control register. The auxiliary-window, port-select,
and/or color-key switching functions are then configured and enabled to perform the desired switching. By
setting the frame-buffer interface to the externally clocked mode, the pixel port and VGA port timing and
pipeline delay are made the same. Also, since the VGA port is enabled, all video-control signal timing is
referenced to CLK0, utilizing the VGABL
, HSYNC, and VSYNC inputs.
The externally clocked frame-buffer interface timing can also be used in non-VGA switching applications,
utilizing only the pixel port or only the VGA port. In either case, it is recommended that VGA video-control
signals be used (i.e., VGABL
, HSYNC, VSYNC). In this way, all pixel data and video-control signals are
referenced to CLK0 and video blank and sync are aligned with pixel data.
NOTE:
When the pixel port is used in externally clocked mode (ACR3 = 0), RCLK must be
set to VCLK/1 (DOT/1) in the output-clock selection register and a 1:1 multiplexing
mode must be selected in the multiplexer control registers (see Table 2–6). The
external clock should be connected to the LCLK input as well as the selected clock
input. When the VGA port is also enabled (MCRB7 = 1), CLK0 is selected as the
input clock independent of the input-clock selection-register setting.
VGA switching can only be performed using a 1:1 multiplexing mode.
Overlay switching can only be performed using a 1:1 multiplexing mode when the
pixel port is set for externally clocked mode. When the pixel port is self-clocked, any
of the multiplex ratios may be used (see subsection 2.4.6).
When VGA switching is to be performed using externally clocked mode
(ACR3 = 0), the full VGA port frequency of 85 MHz may be utilized provided that
the VGA port and the pixel port are both synchronized to the CLK0 input clock.
If VGA switching is to be performed using self-clocked mode (ACR3 = 1), the
maximum pixel rate cannot exceed 50 MHz. This is because of internal delay from
the CLK0 input to the RCLK output. For external clocked timing, the LCLK input
needs to be enabled on terminal 73 (TVP3010C) or terminal L3 (TVP3010M) by
programming the configuration register bit 5 to 1.
VGA-data pipeline delay is adjusted within each VIP depending on whether selfor externally clocked frame-buffer interface timing is used (see subsection 2.3.2 ).
If the VIP is programmed for self-clocked timing, three additional dot-clock pipeline
delays are inserted into the internal VGA-data path and into the internal blanking
signal. The additional pipeline delay accounts for the difference between VGABL
or SYSBL and the pixel-data inputs P(0–31) when used in the self- and externally
clocked modes. This is so the VGA and pixel-port data remain synchronous in time
when doing auxiliary window, port select, or color-keyed switching (see Section
2.6). When externally clocked timing is used, the VGA port and the pixel port are
already synchronous since both data and blanking are presented to the palette
during the same CLK0 clock cycle.
2–11
Loading...
+ 63 hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.