FAIRCHILD TMC2250AKEC

www.fairchildsemi.com
TMC2250A
Matrix Multiplier
12 x 10 bit, 50 MHz
Features
Applications
• Four user-selectable filtering and transformation
functions:
– Triple dot product (3 x 3) matrix multiply
– Cascadeable 9-tap systolic FIR filter
– Cascadeable 3 x 3-pixel image convolver
– Cascadeable 4 x 2-pixel image convolver
• 50 MHz (20ns) pipelined throughput
• 12-bit input and output data, 10-bit coefficients
• 6-bit cascade input and output ports in all filter modes
• Onboard coefficient storage, with three-cycle updating of
all nine coefficients
•
•
•
•
•
•
•
Image filtering and manipulation
Video effects generation
Video standards conversion and encoding/decoding
Three-dimensional image manipulation
Medical image processing
Edge detection for object recognition
FIR filtering for communications systems
Description
The TMC2250A is a flexible high-performance nine-multiplier
array VLSI circuit which can execute a cascadeable 9-tap
FIR filter, a cascadeable 4 x 2 or 3 x 3-pixel image convolution, or a 3 x 3 color space conversion. All configurations
offer throughput at up to the maximum guaranteed 50 MHz
clock rate with 12-bit data and 10-bit coefficients. All inputs
and outputs are registered on the rising edges of the clock.
The 3 x 3 matrix multiply or color conversion configuration
can perform video standard conversion (YIQ or YUV to
RGB, etc.) or three-dimensional perspective translation at
real-time video rates.
The 9-tap FIR filter configuration, useful in Video, Telecommunications, and Signal Processing, features a 16-bit cascade
input to allow construction of longer filters.
The cascadeable 3 x 3 and 4 x 2-pixel image convolver functions allow the user to perform numerous image processing
functions, including static filters and edge detectors. The 16-bit
cascade input port facilitates two-chip 50 MHz cubic convolution (4 x 4-pixel kernel).
The TMC2250A is fabricated in a sub-micron CMOS process
and operates at clock speeds of up to 50 MHz over the full
commercial (0°C to 70°C) temperature and supply voltage
ranges. It is available in 120-pin Plastic Pin Grid Array
(PPGA) packages, 120-lead Ceramic Pin Grid Array package (CPGA), 120-lead PQFP to PPGA package (MPGA) and
120-lead Plastic Quad FlatPack (PQFP). All input and output
signals are TTL compatible.
REV. 1.0.2 10/25/00
PRODUCT SPECIFICATION
TMC2250A
Functional Description
The TMC2250A is a nine-multiplier array with the internal bus
structure and summing adders needed to implement a 3 x 3
matrix multiplier (triple dot product) a cascadeable 9-tap FIR
filter, a 3 x 3-pixel convolver, or a 4 x 2-pixel convolver all in
one monolithic circuit. With a 50MHz guaranteed maximum
clock rate, this device offers video and imaging system
designers a single-chip solution to numerous common image
and signal-processing problems.
The three data input ports (A, B, C) accept 12-bit two's complement integer data, which is also the format for the output
ports (X, Y, Z) in the matrix multiply mode (Mode 00). In the
filter configurations (Modes 01, 10, and 11) the cascade ports
assume 12-bit integer, 4-bit fractional two's complement data
on both input and output. The coefficient input ports (KA,
KB, KC) are always 10-bit two's complement fractional.
Table 1 details the bit weighting of the input and output data
in all configurations.
Operating Modes
The TMC2250A can implement four different digital filter
architectures. Upon selection of the desired function by the
user (MODE1-0), the device reconfigures its internal data
paths and input and output buses appropriately. The output
ports (XC, YC and ZC) are configured in all filter modes a
16-bit Cascade In and Cascade Out ports so that multiple
devices can be connected to build larger filters. These modes
are described individually below. The I/O function configurations for all four modes are shown in Table 1.
Definitions
The calculations performed by the TMC2250A in each mode
are also shown below, utilizing the following notation:
A(1), B(5), C(2), CASIN(3)
Indicates the data word presented to that input port during
the specified clock rising edge(x). Applies to all input ports
A11-0, B11-0, C11-0, and CASIN15-0.
KA1(1), KB3(4)
Indicates coefficient data stored in the specified one of the
nine onboard coefficient registers KA1 through KC3, as
shown in the block diagram for that mode, input during or
before the specified clock rising edge (x).
X(1), Y(4), Z(6), CASOUT (6)
Indicated data available at that output port tDO after that
specified clock rising edge (x). Applies to all output ports
X11-0, Y11-0, Z11-0, and CASOUT15-0.
Numeric Format
Table 2 shows the binary weightings of the input and output
ports of the TMC2250A. Although the internal sums of products could grow to 23 bits, in the matrix multiply mode
(Mode 00) the outputs X, Y and Z are rounded to yield 12-bit
integer words. Thus the output format is identical to the input
data format. In the filter configurations (Modes 01, 10, and
11) the cascade output is always half-LSB rounded to 16
bits, specifically 12 integer bits and 4 fractional guard bits,
with no overflow "headroom". The user is of course free to
half-LSB round the output word to any size less than 16 bits
by forcing a 1 into the bit position of the cascade input
immediately below the desired LSB. In all modes, bit
weighting is easily adjusted if desired by applying the same
scaling correction factor to both input and output data words.
If the coefficients are rescaled, the relative weightings of the
CASIN and CASOUT ports will differ accordingly.
Data Overflow
As shown in Table 2, the TMC2250A's matched input and
output data formats accommodate 0dB (unity) gain. Therefore, the user must be aware of input conditions that could
lead to numeric overflow. Maximum input data and coefficient word sizes must be taken into account with the specific
algorithm performed to ensure that no overflow occurs.
Table 1. Data Port Formatting by Mode
Mode
2
Inputs
A11-0 B11-0 C11-0 KA9-0 KB9-0 KC9-0
Inputs/Output
Outputs
XC11-0
YC11-8
Y7-4
YC3-0
ZC11-0
X11-0
Y11-8
Y7-4
Y3-0
Z11-0
00
A11-0
B11-0 C11-0 KA9-0 KB9-0 KC9-0
01
A11-0
B11-0
KA9-0 KB9-0 KC9-0 CASIN15-4 CASIN3-0
NC
CASOUT3-0 CASOUT15-4
10
A11-0
B11-0 C11-0 KA9-0 KB9-0 KC9-0 CASIN15-4 CASIN3-0
NC
CASOUT3-0 CASOUT15-4
11
A11-0
B11-0
NC
CASOUT3-0 CASOUT15-4
NC
NC
KA9-0 KB9-0 KC9-0 CASIN15-4 CASIN3-0
REV. 1.0.2 10/25/00
TMC2250A
PRODUCT SPECIFICATION
Table 2. Bit Weightings for Input and Output Data Words
Bit Weights
211
210
29
28
27
26
25
24
23
22
21
20 . 2-1 2-2 2-3 2-4 2-5 2-6 2-7 2-8 2-9
-I11
I10
I9
I8
I7
I6
I5
I4
I3
I2
I1
I0
Inputs
All Modes
Data A, B, C
Coefficients
KA, KB, KC
.
-K9 . K8 K7 K6 K5 K4 K3 K2 K1 K0
Modes 01,
-CI15 CI14 CI13 CI12 CI11 CI10 CI9 CI8 CI7 CI6 CI5 CI4 . CI3 CI2 CI1 CI0
10, 11 CASIN
Internal Sum
X20
X19 X18 X17 X16 X15 X14 X13 X12 X11 X10 X9 . X8 X7 X6 X5 X4 X3 X2 X1 X0
Outputs
Mode 00
X, Y, Z
-O11 O10 O9
O8
O7
O6 O5 O4 O3 O2 O1 O0 .
Modes 01,
10, 11
CASOUT
- CO1 CO1 CO1 CO1 CO1 CO CO CO CO CO CO4 . CO CO CO CO
CO15 4
3
2
1
0
9
8
7
6
5
3
2
1
0
Note: A minus sign indicates a two’s complement sign bit.
REV. 1.0.2 10/25/00
3
PRODUCT SPECIFICATION
TMC2250A
Pin Assignments
120 Pin Plastic Quad Flat Pack (MQFP), KE Package
120
1
30
31
4
91
90
61
60
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
26
27
28
29
30
Name
XC6
XC5
XC4
XC3
XC2
XC1
XC0
GND
YC11
YC10
YC9
VDD
YC8
Y7
Y6
GND
Y5
Y4
YC0
VDD
YC1
YC2
YC3
GND
ZC0
ZC1
ZC2
ZC3
ZC4
ZC5
Pin
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Name
ZC6
ZC7
ZC8
GND
ZC9
ZC10
ZC11
KC0
KC1
KC2
KC3
GND
KC4
KC5
KC6
VDD
KC7
KC8
KC9
KB0
KB1
KB2
KB3
KB4
KB5
KB6
KB7
KB8
KB9
KA0
Pin
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
Name
KA1
KA2
KA3
KA4
KA5
KA6
KA7
KA8
KA9
CWE1
CWE0
GND
A0
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
B0
B1
B2
CLK
B3
B4
Pin
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
Name
B5
B6
B7
B8
B9
B10
B11
C0
C1
C2
C3
VDD
C4
C5
C6
GND
C7
C8
C9
C10
C11
MODE1
MODE0
GND
XC11
XC10
XC9
VDD
XC8
XC7
REV. 1.0.2 10/25/00
TMC2250A
PRODUCT SPECIFICATION
Pin Assignments (continued)
120 Pin Plastic Pin Grid Array, H5 Package and 120 Pin Ceramic Pin Grid Array, G1 Package and
120 Pin Plastic Quad Flatpack to 120-Pin Pin Grid Array (MPGA)
13
12
11
10
9
8
Top View
Cavity Up
7
6
5
KEY
4
3
2
1
A
B
C
D
REV. 1.0.2 10/25/00
E
F
G
H
J
K
L
M N
Pin
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
A12
A13
B1
B2
B3
B4
B5
B6
B7
B8
B9
B10
B11
B12
B13
C1
C2
C3
C4
Name
XC7
XC9
XC10
MODE0
C11
C8
C7
C5
C3
C1
B10
B7
B4
XC4
XC5
XC8
XC11
MODE1
C9
C6
C4
C2
B11
B9
B6
B2
XC1
XC2
XC6
VDD
Pin
C5
C6
C7
C8
C9
C10
C11
C12
C13
D1
D2
D3
D11
D12
D13
E1
E2
E3
E11
E12
E13
F1
F2
F3
F11
F12
F13
G1
G2
G3
Name
GND
C10
GND
VDD
C0
B8
B5
B3
B1
YC11
XC0
XC0
CLK
B0
A10
YC9
YC10
GND
A11
A9
A8
Y7
YC8
VDD
A7
A6
A5
Y5
Y6
GND
Pin
G11
G12
G13
H1
H2
H3
H11
H12
H13
J1
J2
J3
J11
J12
J13
K1
K2
K3
K11
K12
K13
L1
L2
L3
L4
L5
L6
L7
L8
L9
Name
A3
A2
A3
Y4
YC0
VDD
GND
A0
A1
YC1
YC2
GND
KA8
CWE1
CWE0
YC3
ZC0
ZC3
KA4
KA7
KA9
ZC1
ZC4
ZC6
GND
KC0
GND
VDD
KB0
KB4
Pin
L10
L11
L12
L13
M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
M11
M12
M13
N1
N2
N3
N4
N5
N6
N7
N8
N9
N10
N11
N12
N13
Name
KB8
KA1
KA5
KA6
ZC2
ZC7
ZC9
ZC11
KC2
KC4
KC6
KC9
KB2
KB5
KB9
KA2
KA3
ZC5
ZC8
ZC10
KC1
KC3
KC5
KC7
KC8
KB1
KB3
KB6
KB7
KA0
5
PRODUCT SPECIFICATION
TMC2250A
Pin Descriptions
Pin Number
Pin Name CPGA/PPGA/
MQFP
MPGA
Function
Pin Description
Supply
Voltage
The TMC2250A operates from a single +5V supply.
All pins must be connected.
Ground
The TMC2250A operates from a single +5V supply.
All pins must be connected.
Power
VDD
F3, H3, L7, C8,
C4
12, 20, 46,
102, 118
GND
E3, G3, J3, L4, 8, 16, 24, 34,
L6, H11, C7, 42, 72, 106,
C5
114
Clock
CLK
D11
88
System Clock
The TMC2250A operates from a single system clock
input. All timing specifications are referenced to the
rising edge of clock.
B4, A4
112, 113
Mode Control
The TMC2250A will switch to the configuration
selected by the user (as shown in Table 3) on the
next clock. This registered control is usually static;
however, should the user wish to switch between
modes, the internal pipeline latencies of the device
must be taken into account. Valid data will not be
available at the outputs in the new configuration until
enough clocks in the new mode have passed to flush
the internal registers.
J12, J13
70, 71
Coefficient
Write Enable
Data presented to the coefficient input ports (KA, KB,
and KC) will update three of the internal coefficient
storage registers, as indicated by the simultaneous
Coefficient Write Enable select, on the next clock.
See Table 4 and the Functional Block Diagram.
A11-0
E11, D13, E12, 84, 83, 82, 81, Data Input A
E13, F11, F12, 80, 79, 78, 77,
F13, G13,
76, 75, 74, 73
G11, G12,
H13, H12
B11-0
B10, A11, B11, 97, 96, 95, 94, Data Input B
C10, A12, B12, 93, 92, 91, 90,
C11, A13,
89, 87, 86, 85
C12, B13,
C13, D12
Data presented to the 12-bit registered data input
ports A, B, and C are latched into the multiplier input
registers for the currently selected configuration
(Table 3). In all modes except Mode 00, new data are
internally right-shifted to the next filter tap on each
rising edge of CLK.
C11-0
A5, C6, B6, A6,
A7, B7, A8, B8,
A9, B9, A10,
C9
Controls
MODE1,0
CWE1,0
Input/Output
6
111, 110,
109, 108,
107, 105,
104, 103,
101, 100, 99,
98
Data Input C
REV. 1.0.2 10/25/00
TMC2250A
PRODUCT SPECIFICATION
Pin Descriptions (continued)
Pin Number
Pin Name CPGA/PPGA/
MQFP
MPGA
Function
KA9-0
K13, J11, K12, 69, 68, 67, 66, Coefficient
L13, L12, K11, 65, 64, 63, 62, Input A1, A2,
M13, M12,
61, 60
A3
L11, N13
KB9-0
M11, L10,
59, 58, 57, 56, Coefficient
N12, N11, 55, 54, 53, 52, Input B1, B2,
M10, L9, N10,
51, 50
B3
M9, N9, L8
KC9-0
M8, N8, N7, 49, 48, 47, 45, Coefficient
M7, N6, M6, 44, 43, 41, 40, Input B1, B2,
N5, M5, N4, L5
39, 38
B3
XC11-0
B4, A3, A2, B3,
115, 116,
CASIN15-4/
A1, C3, B2, B1,
117, 119,
Output X
D3, C2, C1, D2 120, 1, 2, 3, 4,
5, 6, 7
YC11-8
Y7-4
YC3-0
ZC11-0
Pin Description
Data presented to the 10-bit registered coefficient
input ports KA, KB and KC are latched three at a time
into the internal coefficient storage register set
indicated by the Coefficient Write Enable CWE1,0 on
the next clock, as shown in Table 4.
In all modes except Mode 00, the x port and four bits
of the Y output port are reconfigured as the 16-bit
registered Cascade Input port CASIN15-0. Data
presented to this input will be added to the weighted
sums of the data words which were presented to the
D1, E2, E1, F2 9, 10, 11, 13 CASIN3-0/
input ports (A, B and C).
Output Y11-0
In the matrix multiply mode, data are available at the
F1, G2, G1, H1 14, 15, 17, 18 Output7-4 only 12-bit registered output ports X, Y AND Z t after
DO
K1, J2, J1, H2 23, 22, 21, 19 CASOUT3-0/
every clock. These ports are reconfigured in the
Output Y3-0
filtering modes as 16-bit Cascade Input and Output
M4, N3, M3, 37, 36, 35, 33, CASOUT15-4/ ports.CASOUT15-0
In all modes except Mode 00, the Z port and four bits
N2, M2, L3, 32, 31, 30, 29, Output Z11-0
of the Y output port are reconfigured as the 16-bit
N1, L2, K3, 28, 27, 26, 25
registered Cascade Output port CASOUT15-0.
M1, L1, K2
Notes:
1. The output ports X, Y, Z and CASOUT, and input port CASIN are internally reconfigured by the device as required for each
mode of the device. The multiple-function pins have names which are combinations of these titles, as appropriate.
2. The output drivers on pins XC11-0 and YC11-8 are not necessarily disabled until after the first rising edge of CLK following
power-up. If these pins are to be tied to other output drivers, to each other, or to ground or VDD, the user should ensure that
a clock pulse arrives within a few seconds of power-up, to avoid bus contention.
Table 3. Configuration Mode Word
MODE1,0
Configuration Mode
Table 4. Coefficient Write Enable Word
CWE1,0
Coefficient Set Selected
00
3 x 3 Matrix Multiply
00
Hold all registers
01
9-Tap One Dimensional FIR
01
Update KA1, KB1, KC1
10
3 x 3 -Pixel Convolver
10
Update KA2, KB2, KC2
11
4 x 2 -Pixel Convolver
11
Update KA3, KB3, KC3
REV. 1.0.2 10/25/00
7
PRODUCT SPECIFICATION
TMC2250A
Table 5. Coefficient Input Ports
Input Port
multiplied by the appropriate stored coefficients. The three
corresponding sums of products are available at the outputs
five clock cycles after the input data are latched, and three
new data words half-LSB rounded to 12 bits are then available every clock cycle.
Registers Available
KA
KA1, KA2, KA3
KB
KB1, KB2, KB3
KC
KC1, KC2, KC3
X(5)=A(1)KA1(1)+B(1)KB1(1)+C(1)KC1(1)
3 x 3 Matrix Multiplier (Mode 00)
Y(5)=A(1)KA2(1)+B(1)KB2(1)+C(1)KC2(1)
This mode utilizes all six input and output ports in the basic
configuration to realize a "triple dot product", in which each
output is the sum of all three input words in that column
1
2
Z(5)=A(1)KA3(1)+B(1)KB3(1)+C(1)KC3(1)
3
4
5
6
7
8
CLK
CWE
KA, KB, KC
DATA IN A, B, C
MODE CONTROL
01
10
11
K_1
K_2
K_3
0
0
1.0
00
0
0
0
00
KA1 + KB1 + KC1
X OUT
KA2 + KB2 + KC2
Y OUT
KA3 + KB3 + KC3
Z OUT
Figure 1. 3 x 3 Matrix Multiplier Impulse Response (Mode 00)
8
REV. 1.0.2 10/25/00
TMC2250A
A
PRODUCT SPECIFICATION
1
12
2
2
12
KA1
KA
B
12
KA2
10
10
10
10
21
3
3
3
4
4
4
10
12
1
KB1
2
2
12
KB2
10
10
12
KB3
10
10
21
10
10
21
21
3
3
3
4
4
4
10
1
12
2
2
12
KC1
2
12
KC2
10
10
KC
KA3
21
12
C
12
10
10
21
2
KB
2
KC3
10
10
21
12
10
10
21
21
3
3
3
4
4
4
10
RND
RND
RND
12 (MSB)
12 (MSB)
5
5
12
X
12 (MSB)
5
12
Y
12
Z
Figure 2. 3 x 3 Matrix Multiplier Configuration (Mode 00)
REV. 1.0.2 10/25/00
9
PRODUCT SPECIFICATION
TMC2250A
9-Tap FIR Filter Mode (01)
The architecture for this configuration is shown in Figure 4.
The user loads the desired coefficient set, presents input data
to ports A and B simultaneously (most applications will wire
the A and B inputs together), and receives the resulting 9sample response, half-LSB rounded to 16 bits, 5 to 13 clock
cycles later. A new output data word is available every clock
cycle.
CASOUT(13) =
A(9)KA3(9)+A(8)KA2(8)+A(7)KA1(7)
+B(6)KB3(9)+B(5)KB2(8)+B(4)KB1(7)
+B(3)KC3(9)+B(2)KC2(8)+B(1)KC1(7)
+CASIN(10)
Latency: Impulse in to center of 9-tap response =9 registers.
Cascade In to Cascade Out=4 registers.
The figure shows that the input data are automatically rightshifted by one position through the row of multiplier input
registers on every clock in anticipation of a new input data
word.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
CLK
CWE
KA, KB, KC
01
10
11
K_1
K_2
K_3
DATA IN A, B
1.0
01
MODE CONTROL
Q13
CASIN
CASOUT
KA3
KA2
KA1
KB3
KB2
KB1
KC3
KC2
KC1
Q13
Figure 3. 9-Tap FIR Filter Impulse Response (Mode 01)
10
REV. 1.0.2 10/25/00
TMC2250A
A
PRODUCT SPECIFICATION
12
1
2
2
12
KA1
KA
B
12
KA2
10
10
2
KA3
10
10
21
12
10
10
21
21
3
3
3
4
4
4
10
12
1
2
3
4
5
5
12
KB1
5
6
12
KB2
10
10
5
KB3
10
10
21
12
10
10
21
21
7
KB
6
6
6
7
7
7
10
C
8
8
12
KC1
KC
12
KC2
10
10
8
KC3
10
10
21
12
10
10
21
21
9
9
9
10
10
10
10
16 (MSB)
CASIN
(0 - 15) 16
21
21
21
21
21
5
‘10000’
HALF LSB
ROUNDING
2, 5,
8, 11
3, 5, 6
8, 9, 11,
12
16 (MSB)
4 - 13
16
Z = CASOUT
(0 - 15)
Figure 4. 9-Tap FIR Filter Configuration (Mode 01)
REV. 1.0.2 10/25/00
11
PRODUCT SPECIFICATION
TMC2250A
3 x 3 Pixel Convolver (Mode 10)
This filter configuration accepts a 3 pixel-square neighborhood, side-loaded three pixels at a time through input ports
A, B and C, and multiplies the 9 most recent pixel values by
the coefficient set currently stored in the registers. These
products are summed with the data presented to the cascade
input, and a new 3-cycle impulse response, rounded to 16
bits, is available at the output port 5 to 7 clocks later, with a
new output available on every clock cycle.
CASOUT(7)=
A(3)KA3(3)+A(2)KA2(2)+A(1)KA1(1)
+B(3)KB3(3)+B(2)KB2(2)+B(1)KB1(1)
+C(3)KC3(3)+C(2)KC2(2)+C(1)KC1(1)
+CASIN(4)
The input pixel data are automatically shifted one location to
the right through the three rows of multiplier input registers
on every clock in anticipation of three new input data words,
effectively sliding the convolutional window over one column in an image plane.
2
1
3
4
Latency: Impulse in to center of 3-tap response = 6 registers.
Cascade In to Cascade Out=4 registers.
5
6
7
8
9
10
11
CLK
CWE
01
KA, KB, KC
K_1
DATA IN A, B, C
MODE 10
CASIN
CASOUT
10
11
K_2
K_3
1.0
01
Q17
ΣK3
ΣK2
ΣK1
Q7
ΣKj = KAj + KBj + KCj
Figure 5. 3 x 3-Pixel Convolver Impulse Response (Mode 10)
12
REV. 1.0.2 10/25/00
TMC2250A
A
PRODUCT SPECIFICATION
12
1
2
2
12
KA1
KA
B
12
KA2
10
10
21
10
10
21
3
3
3
4
4
4
10
1
12
KB1
2
2
12
KB2
10
10
KB3
10
21
12
10
10
10
21
21
3
3
3
4
4
4
10
1
12
2
2
12
KC1
2
12
KC2
10
10
KC
KA3
21
12
C
12
10
10
2
KB
2
KC3
10
10
21
12
21
10
10
21
3
3
3
4
4
4
10
16 (MSB)
CASIN
(0 - 15) 16
21
21
21
21
21
1
5
‘10000’
HALF LSB
ROUNDING
2, 5
3, 5, 6
16 (MSB)
4-7
16
Z = CASOUT
(0 - 15)
Figure 6. 3 x 3-Pixel Convolver Configuration (Mode 10)
REV. 1.0.2 10/25/00
13
PRODUCT SPECIFICATION
TMC2250A
4 x 2-Pixel Cascadeable Convolver
(Mode 11)
Similar to Mode 10, the 4 x 2 -Pixel convolver allows the use
to perform full-speed cubic convolution with only two
TMC2250A devices and the TMC2111A Pipeline Delay
Register to synchronize the cascade ports (see the Applications Discussion section).
As shown below, the column of input pixel data is automatically shifted one location to the right through the two rows of
multiplier input registers on every clock in anticipation of
two new input data words, effectively sliding the convolutional window over one column in an image plane.
Pixel data are side-loaded into ports A and B, multiplied by
the onboard coefficients, summed with the cascade input,
and half-LSB rounded to 16 bits. The four-cycle impulse
response emerges at the cascade output port 5 to 8 clock
cycles later. A new output word is available on every clock
cycle. Note that Multiplier KC2 is not used in this mode and
that its stored coefficient is ignored.
CASOUT(8)=
1
2
3
4
5
A(4)KA3(4)+A(3)KA2(3)+A(2)KA1(2)
+A(1)KB3(4)+B(4)KB3(4)+B(3)KB2(3)
+B(2)KB1(2)+B(1)KC1(2)+CASIN(5)
6
7
8
9
10
11
CLK
CWE
01
10
11
KA, KB, KC
K_1
K_2
K_3
DATA IN A, B
00
1.0
MODE
11
CASIN
Q8
KA3 + KB3
KA1 + KB1
Q8
CASOUT
KA2 + KB2
KC1 + KC3
Figure 7. 4 x 2-Pixel Convolver Impulse Response (Mode 11)
14
REV. 1.0.2 10/25/00
TMC2250A
A
12
PRODUCT SPECIFICATION
1
2
2
12
12
KA1
KA2
10
10
KA
2
KA3
10
10
21
12
10
10
21
21
3
3
3
4
4
4
10
2
B
12
3
1
2
2
2
12
4
2
12
12
2
KB1
KB2
10
10
KB
KB3
10
10
21
10
10
21
21
3
3
3
4
4
4
10
C
0
3
12
KC1
KC
12
KC2
10
10
5
KC3
10
10
21
12
21
10
10
21
4
6
5
7
10
16 (MSB)
CASIN
(0 - 15) 16
21
21
1
2, 5
6
21
21
3, 5,
6, 7
21
5
16 (MSB)
‘10000’
HALF LSB
ROUNDING
4-8
16
Z = CASOUT
(0 - 15)
Figure 8. 4 x 2-Pixel Convolver Configuration (Mode 11)
REV. 1.0.2 10/25/00
15
PRODUCT SPECIFICATION
TMC2250A
tCY
1
2
tPWH
3
4
5
CLK
CWE
tPWL
KA, KB, KC
tD
tS
X, Y, Z
CASOUT
tH
PREVIOUS
NEW
tHO
Figure 9. Input/Output Timing Diagram
VDD
VDD
p
p
Digital
Input
Digital
Output
n
n
GND
GND
Figure 10. Equivalent Digital Input Circuit
Figure 11. Equivalent Digital Output Circuit
Absolute Maximum Ratings (beyond which the device may be damaged)1
Parameter
Min
Supply Voltage
-0.5
7.0
V
Input Voltage
-0.5
VDD + 0.5
V
-0.5
VDD + 0.5
V
-3.0
6.0
mA
1
sec
Applied Voltage2
Externally Forced
Current3,4
Short Circuit Duration (single output in HIGH state to ground)
Operating, Ambient Temperature
-20
Junction Temperature
Storage Temperature
Lead Soldering Temperature (10 seconds)
-65
Typ
Max
Unit
110
°C
140
°C
150
°C
300
°C
Notes:
1. Functional operation under any of these conditions is NOT implied. Performance and reliability are guaranteed only if
Operating Conditions are not exceeded.
2. Applied voltage must be current limited to specified range.
3. Forcing voltage must be limited to specified range.
4. Current is specified as conventional current flowing into the device.
16
REV. 1.0.2 10/25/00
TMC2250A
PRODUCT SPECIFICATION
Operating Conditions
Parameter
VDD
Power Supply Voltage
fCLK
Clock Frequency
Min
Nom
Max
4.75
5.0
Units
5.25
V
TMC2250A
30
MHz
TMC2250A-2
40
MHz
TMC2250A-3
50
MHz
tPWH
CLK pulse width, HIGH
6
ns
tPWL
CLK pulse width, LOW
8
ns
tS
Input Data Setup Time
6
ns
tH
Input Data Hold Time
2
ns
VIH
Input Voltage, Logic HIGH
2.0
V
VIL
Input Voltage, Logic LOW
0.8
IOH
Output Current, Logic HIGH
-2.0
mA
IOL
Output Current, Logic LOW
4.0
mA
TA
Ambient Temperature, Still Air
70
°C
0
V
Electrical Characteristics
Parameter
IDD
IDDU
Conditions
Total Power Supply
Current
Power Supply Current,
Unloaded
IDDQ
Power Supply Current,
Quiescent
CPIN
I/O Pin Capacitance
IIH
Input Current,
HIGH1
1
Min
Typ
Max
Units
TMC2250A
125
mA
TMC2250A-2
140
mA
TMC2250A-3
155
mA
TMC2250A
120
mA
TMC2250A-2
135
mA
TMC2250A-3
150
mA
VDD = Max, CLK = LOW
12
mA
VDD = Max, CLOAD = 25pF, fCLK = Max
VDD = Max, OE = HIGH, fCLK=Max
5
VDD = Max, VIN = VDD
pF
±5
µA
IIL
Input Current, LOW
VDD = Max, VIN = 0 V
±5
µA
IOZH
Hi-Z Output Leakage
Current, Output HIGH2
VDD = Max, VIN = VDD
±10
µA
IOZL
Hi-Z Output Leakage
Current, Output LOW2
VDD = Max, VIN = 0 V
±10
µA
IOS
Short-Circuit Current
-80
mA
VOH
Output Voltage, HIGH
IOH = Max, VDD = Min
VOL
Output Voltage, LOW
IOL = Max, VDD = Min
-20
2.4
V
0.4
V
Notes:
1. Except pins XC11-0, YC11-8.
2. Pins XC11-0, YC11-8.
REV. 1.0.2 10/25/00
17
PRODUCT SPECIFICATION
TMC2250A
Switching Characteristics
Parameter
Conditions
tDO
Output Delay Time
CLOAD = 25 pF
tHO
Output Hold Time
CLOAD = 25 pF
Min
Typ
Max
15
3
Units
ns
ns
Application Notes
Performing Large-Kernel Pixel Interpolation
The Cascade Input and Output Ports of the TMC2250A allow the user to stack multiple devices to perform larger interpolation
kernels with no decrease in pixel throughput. Figure 12 illustrates a basic application utilizing Mode 11 to realize a 4 x 4-pixel
kernel, also called Cubic Convolution. This example utilizes the TMC2011A Variable-Length Shift Register to compensate for
the internal latency of each TMC2250A. Alternatively, some applications may utilize RAM, FIFO's, or other methods to store
multiple-line pixel data. In these cases the user may compensate for latency by simply offsetting the access sequencing of the
storage devices.
A
A
12
4 X 2 TMC2250A
B
B
12
CASOUT
16
C
A
12
CASIN
4 X 2 TMC2250A
D
B
12
3 X TMC2111A
CASOUT
16
OUTPUT
Figure 12. Figure 12. Performing Cubic Convolution with Two TMC2250A's
Related Products
•
•
•
•
TMC2301 Image Resampling Sequencer
TMC2302A Image Manipulation Sequencer
TMC2249A Video Mixer
TMC2242B Half-Band Filter
18
REV. 1.0.2 10/25/00
TMC2250A
PRODUCT SPECIFICATION
Mechanical Dimensions
120-Lead CPGA Package
Symbol
Inches
Min.
A
A1
A2
øB
øB2
D
D1
e
L
L1
M
N
P
Max.
Millimeters
Min.
Notes:
Notes
1. Pin #1 identifier shall be within shaded area shown.
Max.
.080
.160
.040
.060
.125
.215
.016
.020
.050 NOM.
1.340
1.380
2.03
4.06
1.01
1.53
3.17
5.46
0.40
0.51
1.27 NOM.
33.27
35.05
1.200 BSC
.100 BSC
.110
.145
.170
.190
13
120
.003
—
30.48 BSC
2.54 BSC
2.79
3.68
4.31
4.83
13
120
.076
—
2. Pin diameter excludes solder dip finish.
3. Dimension "M" defines matrix size.
4. Dimension "N" defines the maximum possible number of pins.
5. Orientation pin is at supplier's option.
2
2
SQ
6. Controlling dimension: inch.
3
4
A
A2
A1
øB
L
D
øB2
P
e
Top View
Cavity Up
D1
Pin 1 Identifier
REV. 1.0.2 10/25/00
19
PRODUCT SPECIFICATION
TMC2250A
Mechanical Dimensions
120-Lead PPGA Package
Symbol
Inches
Min.
A
A1
A2
øB
øB2
D
D1
e
L
L1
M
N
P
Max.
Millimeters
Min.
Notes:
Notes
1. Pin #1 identifier shall be within shaded area shown.
Max.
.080
.160
.040
.060
.125
.215
.016
.020
.050 NOM.
1.340
1.380
2.03
4.06
1.01
1.53
3.17
5.46
0.40
0.51
1.27 NOM.
33.27
35.05
1.200 BSC
.100 BSC
.110
.145
.170
.190
13
120
.003
—
30.48 BSC
2.54 BSC
2.79
3.68
4.31
4.83
13
120
.076
—
2. Pin diameter excludes solder dip finish.
3. Dimension "M" defines matrix size.
4. Dimension "N" defines the maximum possible number of pins.
5. Orientation pin is at supplier's option.
2
2
SQ
6. Controlling dimension: inch.
3
4
A
A2
A1
øB
L
D
øB2
P
e
Top View
Cavity Up
D1
Pin 1 Identifier
20
REV. 1.0.2 10/25/00
TMC2250A
PRODUCT SPECIFICATION
Mechanical Dimensions
120-Lead Metric Quad Flat Package to Pin Grid Array Package (MPGA)
Symbol
Inches
Min.
A
A1
A2
A3
øB
øB2
D
D1
e
L
M
N
Max.
Millimeters
Min.
Notes:
Notes
1. Pin #1 identifier shall be within shaded area shown.
Max.
2. Pin diameter excludes solder dip finish.
.309
.311
.145
.155
.080
.090
.050 TYP.
.016
.020
.050 NOM.
1.355
1.365
7.85
7.90
3.68
3.94
2.03
2.29
1.27 TYP.
0.40
0.51
1.27 NOM.
34.42
34.67
2
2
SQ
1.200 BSC
.100 BSC
.175
.185
13
120
30.48 BSC
2.54 BSC
4.45
4.70
13
120
3
4
3. Dimension "M" defines matrix size.
4. Dimension "N" defines the maximum possible number of pins.
5. Orientation pin is at supplier's option.
6. Controlling dimension: inch.
A
A1
A2
L
A3
øB2
øB
e
D
e
Fairchild
TMC2250A
D1
Pin 1 Identifier
REV. 1.0.2 10/25/00
21
PRODUCT SPECIFICATION
TMC2250A
Mechanical Dimensions
120-Lead MQFP Package
Inches
Symbol
Min.
A
A1
A2
B
C
D/E
D1/E1
e
L
N
ND
α
ccc
Max.
—
.154
.010
—
.125
.144
.018
.012
.009
.005
1.219
1.238
1.098
1.106
.0315 BSC
.026
.037
120
30
0°
—
7°
.004
Millimeters
Min.
1. All dimensions and tolerances conform to ANSI Y14.5M-1982.
Max.
—
3.92
.25
—
3.17
3.67
.45
.30
.23
.13
30.95
31.45
27.90
28.10
.80 BSC
.65
.95
120
30
0°
—
Notes:
Notes
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.
4
7°
.10
.20 (.008) Min.
0° Min.
.13 (.005) R Min.
.13/.30
R
.005/.012
D
D1
e
C
PIN 1 IDENTIFIER
E
0.063" Ref (1.60mm)
L
α
Lead Detail
E1
See Lead Detail
Base Plane
A A2
B
A1
22
Seating Plane
-CLEAD COPLANARITY
ccc C
REV. 1.0.2 10/25/00
PRODUCT SPECIFICATION
TMC2250A
Ordering Information
Product Number
Temperature
Range
Speed
Grade
Screening
Package
Package
Marking
TMC2250AG1C
0°C to 70°C
30 MHz
Commercial
120 Pin Ceramic Pin Grid Array
2250AG1C
TMC2250AG1C2
0°C to 70°C
40 MHz
Commercial
120 Pin Ceramic Pin Grid Array
2250AG1C2
TMC2250AG1C3
0°C to 70°C
50 MHz
Commercial
120 Pin Ceramic Pin Grid Array
2250AG1C3
TMC2250AH5C
0°C to 70°C
30 MHz
Commercial
120 Pin Plastic Pin Grid Array
2250AH5C
TMC2250AH5C2
0°C to 70°C
40 MHz
Commercial
120 Pin Plastic Pin Grid Array
2250AH5C2
TMC2250AH5C3
0°C to 70°C
50 MHz
Commercial
120 Pin Plastic Pin Grid Array
2250AH5C3
TMC2250AH6C
0°C to 70°C
30 MHz
Commercial
120 Lead Metric Quad Flatpack
to Pin Grid Array
N/A
TMC2250AH6C2
0°C to 70°C
40 MHz
Commercial
120 Lead Metric Quad Flatpack
to Pin Grid Array
N/A
TMC2250AH6C3
0°C to 70°C
50 MHz
Commercial
120 Lead Metric Quad Flatpack
to Pin Grid Array
N/A
TMC2250AKEC
0°C to 70°C
30 MHz
Commercial
120 Lead Plastic Quad Flatpack
2250AKEC
TMC2250AKEC2
0°C to 70°C
40 MHz
Commercial
120 Lead Plastic Quad Flatpack
2250AKEC2
TMC2250AKEC3
0°C to 70°C
50 MHz
Commercial
120 Lead Plastic Quad Flatpack
2250AKEC3
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
10/25/00 0.0m 002
Stock#DS30002250A
 2000 Fairchild Semiconductor Corporation