DATASHEET

SIGNS
NEW DE
R
O
F
D
NT
N DE
L ACEM E a t
COMME
P
E
E
R
R
T
D
O
E
N
r
D
rt Cente
OM M E N
NO R EC
al Suppo il.com/tsc
ic
n
h
c
e
T
s
our
Sheet
w.inter
contact ERSData
IL or ww
T
N
-I
8
1-88
HSP43168
July 27, 2009
FN2808.12
Dual FIR Filter
Features
The HSP43168 Dual FIR Filter consists of two independent
8-tap FIR filters. Each filter supports decimation from 1 to 16
and provides on-board storage for 32 sets of coefficients.
The Block Diagram shows two FIR cells each fed by a
separate coefficient bank and one of two separate inputs.
The outputs of the FIR cells are either summed or
multiplexed by the MUX/Adder. The compute power in the
FIR Cells can be configured to provide quadrature filtering,
complex filtering, 2-D convolution, 1-D/2-D correlations, and
interpolating/decimating filters.
• Two Independent 8-Tap FIR Filters Configurable as a
Single 16-Tap FIR
The FIR cells take advantage of symmetry in FIR coefficients
by pre-adding data samples prior to multiplication. This
allows an 8-tap FIR to be implemented using only 4
multipliers per filter cell. These cells can be configured as
either a single 16-tap FIR filter or dual 8-tap FIR filters.
Asymmetric filtering is also supported.
• Standard Microprocessor Interface
Decimation of up to 16 is provided to boost the effective
number of filter taps from 2 to 16 times. Further, the Decimation
Registers provide the delay necessary for fractional data
conversion and 2-D filtering with kernels to 16x16.
• 10-Bit Data and Coefficients
• On-Board Storage for 32 Programmable Coefficient Sets
• Up To: 256 FIR Taps, 16x16 2-D Kernels, or 10x19-Bit
Data and Coefficients
• Programmable Decimation to 16
• Programmable Rounding on Output
• Pb-Free Available (RoHS Compliant)
Applications
• Quadrature, Complex Filtering
• Image Processing
• Polyphase Filtering
• Adaptive Filtering
The flexibility of the Dual is further enhanced by 32 sets of
user programmable coefficients. Coefficient selection may
be changed asynchronously from clock to clock. The ability
to toggle between coefficient sets further simplifies
applications such as polyphase or adaptive filtering.
The HSP43168 is a low power fully static design
implemented in an advanced CMOS process. The
configuration of the device is controlled through a standard
microprocessor interface.
Ordering Information
PART NUMBER
PART MARKING
TEMP. RANGE
(°C)
PACKAGE
PKG.
DWG. #
HSP43168VC-45
HSP43168VC-45
0 to +70
100 Ld MQFP
HSP43168VC-45Z (Note)
HSP43168VC-45Z
0 to +70
100 Ld MQFP (Pb-free)
Q100.14x20
Q100.14x20
HSP43168JC-33
HSP43168JC-33
0 to +70
84 Ld PLCC
N84.1.15
HSP43168JC-33Z (Note)
HSP43168JC-33Z
0 to +70
84 Ld PLCC
N84.1.15
NOTE: These Intersil Pb-free plastic packaged products employ special Pb-free material sets, molding compounds/die attach materials, and 100%
matte tin plate plus anneal (e3 termination finish, which is RoHS compliant and compatible with both SnPb and Pb-free soldering operations). Intersil
Pb-free products are MSL classified at Pb-free peak reflow temperatures that meet or exceed the Pb-free requirements of IPC/JEDEC J STD-020.
1
CAUTION: These devices are sensitive to electrostatic discharge; follow proper IC Handling Procedures.
1-888-INTERSIL or 1-888-468-3774 | Intersil (and design) is a registered trademark of Intersil Americas Inc.
Copyright Intersil Americas Inc. 2000, 2001, 2004, 2007, 2009. All Rights Reserved
All other trademarks mentioned are the property of their respective owners.
HSP43168
Block Diagram
10
CIN0 - 9
A0 - 8
WR
CSEL0 - 4
CONTROL/
CONFIGURATION
9
COEFFICIENT
BANK A
INA0 - 9
INB0 - 9/
OUT0 - 8
COEFFICIENT
BANK B
10
FIR CELL A
FIR CELL B
MUX
MUX
10
MUX /
ADDER
9
19
OUT9 - 27
OEL
OEH
Pinouts
CIN 9
CSEL 4
CSEL 3
CSEL 2
CSEL 1
CSEL 0
VCC
A8
A7
A6
A5
A4
A3
A2
A1
A0
GND
WR
MUX 1
MUX 0
CIN 8
HSP43168
(84 LD PLCC)
TOP VIEW
11 10 9 8 7 6 5 4 3 2 1 84 83 82 81 80 79 78 77 76 75
CIN 7
CIN 6
CIN 5
CIN 4
GND
CIN 3
CIN 2
CIN 1
CIN 0
INA 9
INA 8
INA 7
INA 6
INA 5
VCC
INA 4
INA 3
INA 2
INA 1
INA 0
INB 9
74
73
72
71
70
69
68
67
66
65
64
63
62
61
60
59
58
57
56
55
54
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
RVRS
FWD
SHFTEN
TXFR
ACCEN
VCC
CLK
GND
OEH
OUT 27
OUT 26
OUT 25
OUT 24
OUT 23
OUT 22
OUT 21
OUT 20
OUT 19
OUT 18
OUT 17
VCC
INB 8
INB 7
INB 6
INB 5
GND
INB 4
INB 3
INB 2
INB 1
INB 0
OEL
OUT 9
OUT 10
VCC
OUT 11
OUT 12
OUT 13
OUT 14
OUT 15
OUT 16
GND
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
2
FN2808.12
July 27, 2009
HSP43168
(Continued)
CIN9
CSEL4
CSEL3
CSEL2
CSEL1
CSEL0
VCC
VCC
A8
A7
A6
A5
A4
A3
A2
A1
A0
GND
GND
WR
HSP43168
(100 LD MQFP)
TOP VIEW
100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81
CIN8
NC
CIN7
NC
CIN6
CIN5
CIN4
GND
GND
CIN3
CIN2
CIN1
CIN0
INA9
INA8
INA7
INA6
INA5
VCC
VCC
INA4
INA3
INA2
INA1
INA0
NC
NC
INB9
INB8
INB7
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
80
79
78
77
76
75
74
73
72
71
70
69
68
67
66
65
64
63
62
61
60
59
58
57
56
55
54
53
52
51
MUX1
MUX0
RVRS
NC
FWRD
SHIFTEN
TXFR
ACCEN
VCC
VCC
CLK
GND
GND
OEH
OUT27
OUT26
OUT25
OUT24
OUT23
OUT22
OUT21
OUT20
OUT19
OUT18
OUT17
NC
VCC
VCC
GND
GND
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
INB6
INB5
GND
GND
INB4
INB3
INB2
INB1
INB0
OEL
OUT9
OUT10
VCC
VCC
OUT11
OUT12
OUT13
OUT14
OUT15
OUT16
Pinouts
3
FN2808.12
July 27, 2009
HSP43168
Pin Description
SYMBOL
TYPE
DESCRIPTION
VCC
VCC: +5V power supply pin
GND
Ground
CIN0-9
I
Control/Coefficient Data Bus. Processor interface for loading control data and coefficients. CIN0 is the LSB
A0-8
I
Control/Coefficient Address Bus. Processor interface for addressing Control and Coefficient Registers. A0 is the
LSB
WR
I
Control/Coefficient Write Clock. Data is latched into the Control and Coefficient Registers on the rising edge of WR
CSEL0-4
I
Coefficient Select. This input determines which of the 32 coefficient sets are to be used by FIR A and B. This input
is registered and CSEL0 is the LSB.
INA0-9
I
Input to FIR A. INA0 is the LSB
INB0-9
I/O
Bidirectional Input for FIR B. INB0 is the LSB and is input only. When used as output, INB1-9 are the LSBs of the
output bus, and INB9 is the MSB of these bits.
OUT9-27
O
19 MSBs of Output Bus. Data format is either unsigned or two's complement depending on configuration. OUT27
is the MSB.
SHFTEN
I
Shift Enable. This active low input enables clocking of data into the part and shifting of data through the Decimation
Registers.
FWRD
FWD
I
Forward ALU Input Enable. When active low, data from the forward decimation path is input to the ALUs through
the “a” input. When high, the “a” inputs to the ALUs are zeroed.
RVRS
I
Reverse ALU Input Enable. When active low, data from the reverse decimation path is input to the ALUs through
the “b” input. When high, the “b” inputs to the ALUs are zeroed.
TXFR
I
Data Transfer Control. This active low input switches the LIFO being read into the reverse decimation path with
the LIFO being written from the forward decimation path (see Figure 1).
MUX0-1
I
Adder/Mux Control. This input controls data flow through the output Adder/Mux. Table 5 lists the various
configurations.
CLK
I
Clock. All inputs except those associated with the processor interface (CIN0-9, A0-8, WR) and the output enables
(OEL, OEH) are registered by the rising edge of CLK.
OEL
I
Output Enable Low. This three-state control enables the LSBs of the output bus to INB1-9 when OEL is low.
OEH
I
Output Enable High. This three-state control enables OUT9-27 when OEH is low.
ACCEN
I
Accumulate Enable. This active high input allows accumulation in the FIR Cell Accumulator. A low on this input
latches the FIR Accumulator contents into the Output Holding Registers while zeroing the feedback pass in the
Accumulator.
NC
No connect
4
FN2808.12
July 27, 2009
† DATA REVERSAL ENABLE
1
TXFR
DELAY 4
0
DELAY 3
M
U
X
DATA FEEDBACK
CIRCUITRY
† FIR A ODD/EVEN # TAPS
FIR A REVERSE PATH
5
INA0-9
INB0
INB1-9/
OUT0-8
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
D
E M
M U
U X
X
LIFO B
DELAY
1-16 ††
DELAY 3
† ODD/EVEN SYMMETRY
† MODE SELECT
LIFO A
M
U
X
DECIMATION REGISTERS
FIR A FORWARD PATH
SHFTEN
DELAY 4
1
M
U
X
0
DATA FEEDBACK
CIRCUITRY
† FIR B
ODD/EVEN # TAPS
M
U
X
DECIMATION REGISTERS
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
DELAY
1-16 ††
† ODD/EVEN
LIFO A
D
E M
M U
U X
X
LIFO B
DELAY
1-16
NUMBER OF
TAPS
DELAY 3
10
9
DELAY 3
M
U
X
DELAY
1-16 ††
DELAY 3
10
M
U
X
† DATA REVERSAL ENABLE
FIR B REVERSE PATH
FIR B FORWARD PATH
† ODD/EVEN
† ODD/EVEN
10
SYMMETRY
SYMMETRY
A
B
ALU
FWRD
DELAY 3
RVRS
DELAY 3
A
B
ALU
A
B
ALU
A
B
ALU
REG
REG
A
B
ALU
A
B
ALU
A
B
ALU
A
B
ALU
REG
REG
REG
REG
11
† FIR B INPUT
REG
SOURCE
REG
† MODE SELECT
HSP43168
11
X
CSEL0-4
5
10
COEF
BANK
X
COEF
BANK
0
X
COEF
BANK
1
X
2
COEF
BANK
COEF
BANK
X
X
COEF
BANK
0
3
X
COEF
BANK
1
X
2
COEF
BANK
3
DELAY 4
21
REG
REG
REG
REG
REG
0
M
U
X
R
E
G
MUX0-1
CIN0-9
A0-8
WR
0
ADDER
22
REG
ACCEN
REG
REG
FIR B
ACCUMULATOR
FIR A
ACCUMULATOR
CLK
REG
OUTPUT
HOLDING
REGISTER
M
U
X
FIR CELL A
R
E
G
ADDER
REG
OUTPUT
HOLDING
REGISTER
FIR CELL B
DELAY 5
2
DELAY 6
10
9
CONTROL
† MODE SELECT
† ODD EVEN SYMMETRY
† FIR A ODD/EVEN # TAPS
† FIR B ODD/EVEN # TAPS
† FIR B INPUT SOURCE
† DATA REVERSAL ENABLE
† ROUND ENABLE
† DECIMATION FACTOR
† ROUND ENABLE
MUX/
ADDER
28
DELAY 2
9
19
OUT9-27
FN2808.12
July 27, 2009
† Processor control words
†† Decimation factor
OEL
OEH
FIGURE 1. DUAL FIR FILTER
HSP43168
Functional Description
As shown in Figure 1, the HSP43168 consists of two
4-multiplier FIR filter cells which process 10-bit data and
coefficients. The FIR cells can operate as two independent
8-tap FIR filters or two 4-tap asymmetric filters at maximum
I/O rates. A single filter mode is provided which allows the
FIR cells to operate as one 16-tap FIR filter or one 8-tap
asymmetric filter. On board coefficient storage for up to 32
sets of 8 coefficients is provided. The coefficient sets are
user selectable and are programmed through a
microprocessor interface. Programmable decimation to 16 is
also provided. By utilizing Decimation Registers together
with the coefficient sets, polyphase filters are realizable
which allow the user to trade data rate for filter taps. The
MUX/Adder can be configured to either add or multiplex the
outputs of the filter cells depending upon whether the cells
are operating in single or dual filter mode. In addition, a
shifter in the MUX/Adder is provided for implementation of
filters with 10-bit data and 20-bit coefficients or vice versa.
Preparing the Dual FIR for Operation
Two configuration steps are required to prepare the Dual FIR
Filter for normal operation: 1) loading the Configuration
Control Registers, and 2) loading the FIR Filter Coefficients.
Configuration Control Registers are loaded by placing the
control register address on address lines A0-8, placing the
configuration data on the configuration input lines CIN0-9,
and asserting the WR line (followed by a release of the
assertion). This action creates a rising edge on the WR line,
which clocks the address and configuration data into the part.
The details of the “Load Configuration” process are outlined in
“Microprocessor Interface” on page 6.
FIR Coefficients are loaded by placing the address of the
Coefficient Data Bank on the address lines A0-8, placing
the FIR 10-bit coefficient values on the configuration input
lines CIN0-9 and then asserting the WR line (followed by a
release of the assertion). This action creates a rising edge
on the WR line, which clocks the FIR Coefficient Band
address and FIR Coefficient data into the part. The details
of the “Load FIR Coefficient” process are outlined in
“Coefficient Bank” on page 8.
Both the Configuration Load and FIR Coefficient Load can
be done as a sequence of asynchronous write commands
to the Dual FIR Filter. Once these actions are complete, the
part is ready for normal filter operation. The CLK, TXFR,
FWRD, RVRS, ACCEN, and SHFTEN signals must be
asserted in a manner determined by the application.
MUX0-1 must meet the setup and hold times with respect
to clock for proper filter operation. Details of the MUX1-0
control can be found in “Output MUX/Adder” on page 9.
Details of the ACCEN control can be found in “FIR Cell
Accumulator” on page 9. Bit locations for the various filter
control/configuration signals can be found in “Input/Output
Formats” on page 10.
6
The Dual FIR Filter has a “pipeline” delay of 8 CLK periods,
once normal filtering operations begin. Five typical filtering
operation examples are provided in “Application Examples”
on page 10 as a guide to configuration and control of the
Dual FIR Filter.
During normal filter operations, the location and duration of
the TXFR signal assertions are determined by the filter
configuration and operation mode. Once set, these signal
parameters must be maintained during normal operation to
ensure proper data alignment in the part. Once the part is
reset, do not change TXFR unless you load the configuration
again.
Note: The fixed or periodic relationship between the
TXFR signal and CLK must be maintained for valid filter
operation. This relationship can only change when CLK
is halted and new configuration control words are
loaded into the device.
Microprocessor Interface
The Dual FIR has a 20 pin write only microprocessor interface
for loading data into the Control Block and Coefficient Banks.
The interface consists of a 10-bit data bus (CIN0-9), a 9-bit
address bus (A0-8), and a write input (WR) to latch the data
into the on-board registers on a rising edge. The configuration
control and coefficient data loading is asynchronous to CLK.
Control Block
The Dual FIR is configured by writing to the registers within
the Control Block. Figure 2 shows the timing diagram for
writing to the Configuration Control Registers. These Control
Registers are memory mapped to Address 000H (H =
Hexadecimal) and 001H on A0-8. The Filter Coefficient
Registers are mapped to 1XXH (X = value described in
“Coefficient Bank” on page 8).
RESET
WR
A8-0
000H
001H
C9-0
FIGURE 2. LATCHING C9-0 VALUES INTO ADDRESS A8-0
REGISTERS
The format of the Control Registers is shown in Table 1 and
Table 2. Writing to any of the Control/Configuration Registers
causes a reset which lasts for 6 CLK cycles following the
assertion of WR. The reset caused by Writing Registers in
the Control Block will not clear the contents of the Coefficient
Bank. As shown in Figure 2, either Configuration Control
Register can be written to during reset.
FN2808.12
July 27, 2009
HSP43168
TABLE 1. CONFIGURATION/CONTROL WORD 0 BIT
DEFINITIONS
CONTROL ADDRESS 000H
BITS
3-0
4
FUNCTION
DESCRIPTION
Decimation Factor (N) R = N + 1
0000 = No Decimation.
1111 = Decimation by 16.
Mode Select
0 = Single Filter Mode.
1 = Dual Filter Mode.
(also 20-Bit Coefficient Filter)
5
Odd/Even Filter
Coefficient Symmetry
0 = Even Symmetric Coefficients.
1 = Odd Symmetric Coefficients.
6
FIR A Odd/Even
Number of Taps
0 = Odd Number of Taps in Filter.
1 = Even Number of Taps in Filter.
7
FIR B Odd/Even
Number of Taps
(Defined Same as FIR A Above).
8
FIR B Input Source
0 = Input from INA0-9.
1 = Input from INB0-9.
9
Not Used
Set to 0 for Proper Operation.
TABLE 2. CONFIGURATION/CONTROL WORD 1 BIT
DEFINITIONS
CONTROL ADDRESS 001H
BITS
FUNCTION
DESCRIPTION
0
FIR A Input Format
0 = Unsigned.
1 = Two's Complement.
1
FIR A Coefficient Format
(Defined same as FIR A input).
2
FIR B Input Format
(Defined same as FIR A input).
3
FIR B Coefficient
(Defined same as FIR A input).
4
Data Reversal Enable
0 = Enabled.
1 = Disabled.
8-5
Round Position
0000 = 2-10.
1011 = 21.
(See Figure 4)
9
Round Enable
0 = Enabled.
1 = Disabled.
NOTES:
1. Address locations 002H to 011H are reserved, and writing to
these locations will have unpredictable effects on part
configuration.
The 4 LSBs of the control word loaded at address 000H are
used to select the decimation factor. The Decimation Factor
is programmed to one less than the number of delays
between filter taps as shown in Equation 1.
DF =  CLK delays between taps  – 1
(EQ. 1)
For example, if the 4 LSBs are programmed with a value of
0010, the Forward and Reverse Shifting Decimation Registers
are each configured with a delay of 3. Bit 4 is used to select
whether the FIR cells operate as two independent filters or
one extended length filter. Dual filter mode assumes Filter A
and Filter B are separate independent filters. In the single filter
mode, the data is routed through the forward paths of Filters A
and B before entering the reverse paths of Filters A and B
7
(see Figure 1). Coefficient symmetry is selected by bit 5. Bits 6
and 7 are programmed to configure the FIR cells for odd or
even filter lengths (number of taps). Bit 8 selects the FIR B
input source when the FIR cells are configured for
independent operation. Bit 9 must be programmed to 0.
Note: When the filter is programmed for even-taps, the
TXFR signal is delayed by only three CLKS (see Figure 1).
For odd-taps, the TXFR signal is delayed by four CLKS.
The 4 LSBs of the control word loaded at address 001H are
used to configure the format of the FIR cell's data and
coefficients. Bit 4 is programmed to enable or disable the
reversal of data sample order prior to entering the Reverse
Path Decimation Registers. Data reversal is required for
symmetric filter coefficient sets of both even or odd numbers
of filter taps. Asymmetric filters and some decimated
symmetric filters require the data reversal to be off. Bits 5 to
9 are used to support programmable rounding on the output.
FIR Filter Cells
Each FIR filter cell is based on an array of four 11x10-bit two's
complement multipliers. One input of the multipliers comes
from the ALU’s which combine data shifting through the
Forward and Reverse Decimation Registers. The second
multiplier input comes from the user programmable coefficient
bank. The multiplier outputs are fed to an accumulator whose
result is passed to the output section where it is multiplexed or
added with the result from the other FIR cell.
Decimation Registers
The Forward and Reverse Decimation Shift Registers can
be configured for decimation factors from 1 to 16 (see
Table 1, bits 0-3). Note: Setting the decimation factor
only affects the Delay Registers between filter taps, not
the filter control multiplexers. Example 4 and Example 5
in “Application Examples” on page 10 discuss how to
configure the part for actual decimation applications.
The Reverse Shifting Registers with the data reversal logic
are used to take advantage of symmetry in linear phase filters
by aligning data at the ALUs for pre-addition prior to
multiplication by the common coefficient. When the FIR cells
are configured in single filter mode, the Decimation Registers
in FIR cell A and FIR cell B are cascaded. This extended filter
tap delay path allows computation of a filter which is twice the
size of that capable using a single cell. The Decimation
Registers also provide data storage for polyphase or 2-D
filtering applications (See “Application Examples” on
page 10).
The Data Feedback Circuitry in each FIR cell is responsible
for transferring data from the Forward to the Reverse
Shifting Decimation Registers. This circuitry feeds blocks of
samples into the reverse shifting decimation path in either
reversed or non-reversed sample order. The MUX/DEMUX
structure at the input to the Feedback Circuitry routes data
to the LIFOs or the delay stage depending on the selected
FN2808.12
July 27, 2009
HSP43168
configuration. The MUX on the Feedback Circuitry Output
selects which storage element feeds the Reverse Shifting
Decimation Registers.
In applications requiring reversal of sample order, the FIR
cells are configured with data reversal enabled (see
Table 2, CW5, bit 4 = 0). In this mode, data is transferred
from the forward to the backward Shifting Registers
through a pingponged LIFO structure. While one LIFO is
being read into the backward shifting path, the other LIFO
is written with data samples. The MUX/DEMUX controls
which LIFO is being written, and the MUX on the Feedback
Circuitry output controls which LIFO is being read. A low on
TXFR and SHIFTEN, switches the LIFOs being read and
written, which causes the block of data to be read from the
structure in reversed in sample order (See Example 4 in
“Application Examples” on page 10).
The frequency with which TXFR is asserted determines size
of the data blocks in which sample order is reversed. For
example, if TXFR is asserted once every three CLKs, blocks
of 3 data samples with order reversed, would be fed into the
Backward Decimation Registers. Note: Altering the
frequency or phase of TXFR assertion once a filtering
operation has begun will invalidate the filtering result.
In applications which do not require sample order reversal,
the FIR cells must be configured with data reversal
disabled (see Table 2, CW5, Bit 4 = 1). In addition, TXFR
must be asserted to ensure proper data flow. In this
configuration, data to the backward shifting decimation path
is routed though a delay stage instead of the pingpong
LIFOs. The number of registers in the delay stage is based
on the programmed decimation factor. Note: Data
reversal must be disabled and TXFR must be asserted
for filtering applications which do not use decimation.
The configuration of the FIR cells as even or odd length filters
determines the point in the forward decimation path from
which data is multiplexed to the Data Feedback Circuitry. For
example, if the FIR cell is configured as an odd length filter,
data prior to the last register in the third forward decimation
stage is routed to the Feedback Circuitry. If the FIR cell is
configured as an even length filter, data output from the third
forward decimation stage is multiplexed to the Feedback
Circuitry. This is required to ensure proper data alignment with
symmetric filter coefficients (See “Application Examples” on
page 10).
ALUs
Data shifting through the forward and reverse decimation
paths feed the “a” and “b” inputs of the ALUs respectively.
The ALUs perform an “b+a” operation if the FIR cell is
configured for even symmetric coefficients or an “b-a”
operation if configured for odd symmetric coefficients.
Control Word 0, Bit 5 is used to set the ALU operation.
For applications in which a pre-add or subtract is not required,
the “a” or “b” input can be zeroed by disabling FWRD or
RVRS respectively. This has the effect of producing an ALU
output which is either “a”, “-a”, or “b” depending on the filter
symmetry chosen. For example, if the FIR cell is configured
for an even symmetric filter with FWRD low and RVRS high,
the data shifting through the Forward Decimation Registers
would appear on the ALU output.
Table 3 details the ALU configurations, where “a” is the ALU
data input from the front decimation delay registers and “b” is
the ALU data from the back decimation delay registers.
TABLE 3. ALU CONFIGURATIONS
ALU
OUT
SYMMETRY FWD RVS
The shifting of data through the Forward and Reverse
Decimation Registers is enabled by asserting the SHFTEN
input. When SHFTEN is high, data shifting is disabled, and
the data sample latched into the part on the previous clock is
the last input to the filter structure. The data sample at the
filter input when SHFTEN is asserted, will be the next data
sample into the forward decimation path.
a+b
0 (Even)
0
0
Even Number of Taps, Even
Symmetry (Example 1)
+b
0 (Even)
0
1
Even Symmetry
+a
0 (Even)
1
0
Even Symmetry
-
0 (Even)
1
1
Even Symmetry
b-a
1 (Odd)
0
0
Even Number of Taps, Odd
Symmetry (Example 2)
When operating the FIR cells as two independent filters, FIR
A receives input data via INA0-9 and FIR B receives data
from either INA0-9 or INB0-9 depending on the application
(see Table 1).
+b
1 (Odd)
0
1
Odd Symmetry
-a
1 (Odd)
1
0
Odd Symmetry
-
1 (Odd)
1
1
Odd Symmetry
When the FIR cells are configured as a single extended
length filter, the forward and reverse decimation paths of the
two FIR cells are cascaded. In this mode, data is transferred
from the forward decimation path to the reverse decimation
path by the Data Feedback Circuitry in FIR B. Thus, the
manner in which data is read into the reverse decimation
path is determined by FIR B's configuration. When the
decimation paths are cascaded, data is routed through the
fourth delay stage in FIR A's forward path to FIR B.
8
DESCRIPTION
Coefficient Bank
The output of the ALU is multiplied by a coefficient from one
of 32 user programmable coefficient sets. Each set consists
of 8 coefficients (4 coefficients for FIR A and 4 for FIR B).
CSEL0-4 is used to select a coefficient set to be used.
Coefficient sets may be switched every clock to support
polyphase filtering operations.
FN2808.12
July 27, 2009
HSP43168
The coefficients are loaded into On-Board Registers using
the microprocessor interface, CIN0-9, A0-8, and WR. Each
multiplier within the FIR Cells is driven by a coefficient bank
with one of 32 coefficients. These coefficients are addressed
as shown in Table 4. The inputs A0-1 specify the Coefficient
Bank for one of the four multipliers in each FIR Cell; A2
specifies FIR Cell A or B; Bits A7-3 specify one of 32 sets in
which the coefficient is to be stored. For example, an
address of 10dH would access the coefficient for the second
multiplier in FIR B in the second coefficient set.
TABLE 4. FIR COEFFICIENT WRITE ADDRESSES
FIR
COEFF.
CSEL (4-0)
COEFF.
SET
CELL
A/B
MULTIPLIER
A8
A7-3
A2
A1-0
FIR
BANK
1
xxxx x
0
00
A
0
1
xxxx x
0
01
A
1
1
xxxx x
0
10
A
2
DESTINATION
Output MUX/Adder
The contents of each FIR Cell's Output Holding Register is
summed or multiplexed in the Mux/Adder. The operation of
the Mux/Adder is controlled by the MUX1-0 inputs as
shown in Table 5. Applications requiring 10-bit data and
20-bit coefficients or 20-bit data and 10-bit coefficients are
made possible by configuring the MUX/Adder to scale FIR
B's output by 2-10 prior to summing with FIR A. When the
Dual FIR is configured as two independent filters, the
MUX1-0 inputs would be used to multiplex the filter outputs
of each cell. For applications in which FIR A and B are
configured as a single filter, the MUX/Adder is configured to
sum the output of each FIR cell.
NOTE: While a 20-bit coefficient filter is a single filter, the mode
select is set to 1 and MUX1-0 is set to 00.
TABLE 5. MUX1-0 BIT DEFINITIONS
MUX1-0 DECODING
MUX1-0
OUT0-27
1
xxxx x
0
11
A
3
1
xxxx x
1
00
B
0
00
FIRA + FIRB (FIR B Scaled by 2-10)
1
xxxx x
1
01
B
1
01
FIRA + FIRB
1
xxxx x
1
10
B
2
10
FIRA
1
xxxx x
1
11
B
3
11
FIRB
FIR Cell Accumulator
The registered outputs from the multipliers in each FIR cell
feed an accumulator. The ACCEN input controls each
accumulator's running sum and the latching of data from the
accumulator into the Output Holding Registers. When
ACCEN is low, feedback from the accumulator adder is
zeroed which disables accumulation. Also, output from the
accumulator is latched into the Output Holding Registers.
When ACCEN is asserted, accumulation is enabled and the
contents of the Output Holding Registers remain unchanged.
9
FN2808.12
July 27, 2009
HSP43168
Input/Output Formats
The Dual FIR supports mixed mode arithmetic with both
unsigned and two's complement data and coefficients. The
input and output formats for both data types are shown in
Figure 3. If the Dual FIR is configured as an even symmetric
filter with unsigned data and coefficients, the output will be
unsigned. Otherwise, the output will be two's complement.
The MUX/Adder can be configured to implement
programmable rounding at bit locations 2-10 through 21. The
round is implemented by adding a 1 to the specified location
(see Table 2). Figure 4 illustrates the rounding operation. For
example, to configure the part such that the output is
rounded to the 10 MSBs, OUT18 - 27, the round position
would be chosen to be 2-1. The negative sign on the MSB
indicates 2’s complement format.
INPUT DATA FORMAT INA0-9, INB0-9
FRACTIONAL TWO’S COMPLEMENT
9
8
7
6
5
4
3
2
1
0
-20
.2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
27
26
25
24
23
22
21
20
19
IOUT
18
9-27
17
16
15
14
13
12
11
10
9
8
7
6
5
IOUT
4
0-8
3
2
1
0
21
20
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
2-10
8
9
10
11
12
13
14
15
16
17
18
19
1011
1010
1001
1000
0111
0110
0101
0100
0011
0010
0001
0000
“ROUND POSITION” VALUE
NUMBER OF OUTPUT BITS
LOCATION OF ADDITION OF 1
OUTPUT BITS
OUTPUT DATA FORMAT OUT9-27
FRACTIONAL TWO'S COMPLEMENT
FIGURE 4. ROUND POSITION BIT DEFINITION
27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
-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
OUTPUT DATA FORMAT OUT0-8
FRACTIONAL TWO'S COMPLEMENT
8
7
6
5
4
3
2
1
0
2-10
2-11
2-12
2-13
2-14
2-15
2-16
2-17
2-18
INPUT DATA FORMAT INA0-9, INB0-9
FRACTIONAL UNSIGNED
In this section a number of examples are presented which
detail even, odd, symmetric, asymmetric, decimating and
dual FIR filter configurations. These examples are intended
to illustrate the different operational features of the
HSP43168 and should be used as a guide in developing an
application specific filter configuration. Use Table 6 to select
and find the example that best matches your application.
TABLE 6. FILTER EXAMPLE SELECTION GUIDE
9
8
7
6
5
4
3
2
1
0
20
.2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
OUTPUT DATA FORMAT OUT9-27
FRACTIONAL UNSIGNED
27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
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
OUTPUT DATA FORMAT OUT0-8
FRACTIONAL UNSIGNED
8
7
6
5
4
3
2
1
0
2-10
2-11
2-12
2-13
2-14
2-15
2-16
2-17
2-18
FIGURE 3. INPUT/OUTPUT FORMAT DEFINITIONS
10
Application Examples
FILTER TYPE
EXAMPLE NUMBER
Even Tap Even Symmetric
1
Odd Tap Even Symmetric
2
Asymmetric
3
Even Tap Decimating
4
Odd Tap Decimating
5
Dual Decimating
6
Examples 1 through 5 are explained using a single four tap
FIR cell, but the same concept applies to FIR filters which
use both FIR cells (A and B) in a single filter configuration.
Example 6 details a dual filter mode where FIR cell A and B
implement different digital filters. All examples are
functionally verified configurations. Each example details a
complete design solution, including a block diagram, a
data/coefficient alignment illustration, a data flow diagram
and a control signal timing diagram.
FN2808.12
July 27, 2009
HSP43168
Two programmable Configuration Control Registers define a
unique FIR filter configuration. Register 000H has all filter
configuration unique parameters, while Register 001H, Bit 4, is
filter configuration unique. Table 7 details the configuration
control register values, the number of filter coefficient banks
required and the MUX1-0 control values for each filter example.
Figure 7A shows the data sample alignment at the preadders for the data/coefficient alignment shown in Figure 6.
C3
8 TAPS
C3
C2
C2
C1
h(n)
C1
C0
C0
TABLE 7. CONFIGURATION CONTROL REGISTER VALUES
REG
000
HEX
REG # OF FILTER
001 COEFFICIENT
BANKS
HEX
Even Tap Even
Symmetric
1d0
010
1
10
Odd Tap Even Symmetric
110
010
1
10
Asymmetric
110
010
2
10
Even Tap Decimate by
N+1
1dN
000
N+1
10
11N
000
N+1
10
15N or
19N
000
N+1
10 and
11
FILTER TYPE
Odd Tap Decimate by N+1
Dual: Even and Odd Tap
Decimate by N+1
MUX
1-0
Bit 4
Example 1: Even-Tap Even Symmetric Filter
Example
The HSP43168 may be configured as two independent
8-tap symmetric filters as shown by the Block Diagram in
Figure 5. Each of the FIR cells takes advantage of
symmetric filter coefficients by pre-adding data samples
common to a given coefficient. As a result, each FIR cell
can implement an 8-tap symmetric filter using only four
multipliers. Similarly, when the HSP43168 is configured in
single filter mode a 16-tap symmetric filter is possible by
using the multipliers in both cells.
HSP43168
8-TAP EVEN SYMMETRIC
INA0-9
A
A
B
B
INB0-9
FIR A
8-TAP EVEN SYMMETRIC
M
U
X
OUT9-27
x(n)
X9
X8
X7
X6
X5
X4
X3
X2
X1
X0
FIGURE 6. DATA/COEFFICIENT ALIGNMENT FOR 8-TAP
EVEN SYMMETRIC FILTER
The dual filter application is configured by writing 1d0H to
address 000H via the microprocessor interface, CIN0-9, A0-8,
and WR. Since this application does not use decimation, the
4th bit of the Control Register at Address 001H must be set to
disable data reversal (see Table 2). Failure to disable data
reversal will produce erroneous results.
Using this architecture, only the unique coefficients need to
be stored in the Coefficient Bank. For example, the above
filter would be stored in the first coefficient set for FIR A by
writing C0, C1, C2, and C3 to Address 100H, 101H, 102H,
and 103H respectively. To write the same filter to the first
coefficient set for FIR B, the address sequence would
change to 104H, 105H, 106H, and 107H.
To operate the HSP43168 in this mode, TXFR is tied low to
ensure proper data flow; both FWRD and RVRS are tied low
to enable data samples from the forward and reverse data
paths to the ALUs for pre-adding; ACCEN is tied low to
prevent accumulation over multiple CLKs; SHFTEN is tied low
to allow shifting of data through the Decimation Registers;
MUX0-1 is programmed to multiplex the output the of either
FIR A or FIR B; CSEL0-4 is programmable to access the
stored coefficient set, in this example CSEL = 00000.
FIR B
0
1
2
6
5
4
3
FIGURE 5. USING HSP43168 AS TWO INDEPENDENT FILTERS
The operation of the FIR cell is better understood by comparing
the data and coefficient alignment for a given filter output,
Figure 6, with the data flow through the FIR cell, as shown in
Figure 7. The Block Diagrams in Figure 7 are a simplification of
the FIR cell shown in Figure 1. For simplicity, the ALUs and FIR
Cell Accumulators were replaced by adders, and the Pipeline
Delay Registers were omitted. In this example, we will only
show the data flow through one of the two FIR cells.
In Figure 7, the order of the data samples within the filter
cell is shown by the numbers in the forward and backward
shifting decimation paths. The output of the filter cell is
given by the equation at the bottom of each block diagram.
11
7
+
C0
+
C1
+
C2
+
C3
+
(X7 + X0)C0 + (X6 + X1)C1 + (X5 + X2)C2 + (X4 + X3)C3
FIGURE 7A. DATA FLOW AS DATA SAMPLE 7 IS CLOCKED
INTO THE FEED-FORWARD STAGE
FN2808.12
July 27, 2009
HSP43168
8
1
2
3
7
6
5
+
+
+
The operation of the FIR cell for odd length filters is better
understood by comparing the data/coefficient alignment in
Figure 9 with the Data Flow Diagrams in Figure 10. The
Block Diagrams in Figure 10 are a simplification of the FIR
cell shown in Figure 1.
4
+
C0
C1
C2
7-TAPS
C3
C2
C1
h(n)
C3
C2
C1
C0
C0
+
(X8 + X1)C0 + (X7 + X2)C1 + (X6+X3)C2 + (X5 + X4)C3
x(n)
FIGURE 7B. DATA FLOW AS DATA SAMPLE 8 IS CLOCKED
INTO THE FEED-FORWARD STAGE
9
2
3
4
8
7
6
+
C0
+
+
C1
C2
X9
X8
X7
X6
X5
X4
X3
X2
X1
X0
FIGURE 9. DATA/COEFFICIENT ALIGNMENT FOR 7-TAP
SYMMETRIC FILTER
5
For odd length filters, proper data/coefficient alignment is
ensured by routing data entering the last register in the third
forward decimation stage to the Backward Shifting
Registers. In this configuration, the center coefficient must
be scaled by 1/2 to compensate for the summation of the
same data sample from both the Forward and Backward
Shifting Registers.
+
C3
+
0
1
2
5
4
3
3
(X9 + X2)C0 + (X8 + X3)C1 + (X7 + X4)C2 + (X6 + X5)C3
FIGURE 7C. DATA FLOW AS DATA SAMPLE 9 IS CLOCKED
INTO THE FEED-FORWARD STAGE
6
+
FIGURE 7. DATA FLOW DIAGRAMS FOR 8-TAP SYMMETRIC
FILTER
C0
+
C1
C2
Example 2: Odd-Tap Even Symmetric
Filter Example
(X6 + X0)C0 + (X5 + X1)C1 + (X4 + X2)C2 + (X3 + X3)C3/2
1
B
B
INB0-9
C0
2
6
7
+
7-TAP EVEN SYMMETRIC
A
C3/2
FIGURE 10A. DATA FLOW AS DATA SAMPLE 6 IS CLOCKED
INTO THE FEED-FORWARD STAGE
HSP43168
A
+
+
The HSP43168 may be configured as two independent
7-tap symmetric filters with a Functional Block Diagram
shown in Figure 8. Again, this example shows data flow
through one of the two FIR cells. As in the 8-tap filter
example, the HSP43168 implements the filtering operation
by summing data samples sharing a common coefficient
prior to multiplication by that coefficient. However, for odd
length filters the pre-addition requires that the center
coefficient be scaled by 1/2.
INA0-9
+
5
+
C1
4
+
C2
4
3
+
C3/2
FIR A
7-TAP EVEN SYMMETRIC
M
U
X
OUT9-27
+
FIR B
(X7 + X1)C0 + (X 6 + X2)C1 + (X5 + X3)C2 + (X4 + X4)C3/2
FIGURE 8. USING HSP43168 AS TWO INDEPENDENT FILTERS
12
FIGURE 10B. DATA FLOW AS DATA SAMPLE 7 IS CLOCKED
INTO THE FEED FORWARD STAGE
FN2808.12
July 27, 2009
HSP43168
2
3
5
4
HSP43168
8-TAP ASYMMETRIC
7
8
+
C0
6
5
+
C1
+
C2
FIR A
INA0-9
+
A
A
B
B
INB0-9
M
U
X
8-TAP ASYMMETRIC
OUT9-27
FIR B
C3/2
FIGURE 11. USING HSP43168 AS TWO INDEPENDENT
FILTERS
+
(X8 + X2)C0 + (X7 + X3)C1 + (X6 + X4)C2 + (X5 + X5)C3/2
FIGURE 10C. DATA FLOW AS DATA SAMPLE 8 IS CLOCKED
INTO THE FEED FORWARD STAGE
FIGURE 10. DATA FLOW DIAGRAMS FOR 7-TAP SYMMETRIC
FILTER
In the Data Flow Diagrams of Figure 10, the order of the data
samples input in to the filter cell is shown by the numbers in
the forward and backward shifting decimation paths. The
output of the filter cell is given by the equation at the bottom
of the block. The diagram in Figure 10A shows data sample
alignment at the pre-adders for the Data/Coefficient
Alignment shown in Figure 9.
This dual filter application is configured by writing 110H to
Address 000H via the microprocessor interface, CIN0-9, A0-8,
and WR. Also, data reversal must be disabled by setting bit 4
of the Control Register at Address 0001H. As in the 8-tap
example, only the unique coefficients need to be stored in the
Coefficient Bank. These coefficients are stored in the first
coefficient set for FIR A by writing C0, C1, C2, and C3 to
Address 100H, 101H, 102H, and 103H respectively. To write
the same filter to the first coefficient set for FIR B, the address
sequence would change to 104H, 105H, 106H, and 107H.
The control signals TXFR, FWRD, RVRS, ACCEN, SHFTEN,
and CSEL0-4 are controlled as described in Example 1.
The operation of this configuration is better understood by
comparing the Data/Coefficient Alignment in Figure 12 with
the Data Flow Diagrams in Figure 13. The ALUs have been
omitted from the FIR cell diagrams because data is fed to the
multipliers directly from the forward and reverse decimation
paths. The data samples within the FIR cell are shown by the
numbers in the decimation paths.
C7
h(n)
C6
C5
C4
8-TAPS
C3
C2
C1
C0
x(n)
X9
X8
X7
X6
X5
X4
X3
X2
X1
X0
FIGURE 12. DATA/COEFFICIENT ALIGNMENT FOR 8-TAP
ASYMMETRIC FILTER
Example 3: Asymmetric Filter Example
The FIR cells within the HSP43168 can each calculate 4
asymmetric taps on each clock. Thus, a single FIR cell can
implement an 8-tap asymmetric filter if the HSP43168 is
clocked at twice the input data rate. Similarly, if the Dual is
configured as a single filter, a 16-tap asymmetric filter is
realizable. Only one of the two FIR cells are used in this
example for the Block Diagram shown in Figure 11.
For this example, the FIR cells are configured as two 8-tap
asymmetric filters which are clocked at twice the input data
rate. New data is shifted into the forward and backward
decimation paths every other CLK by the assertion of
SHFTEN. The filter output is computed by passing data from
each decimation path to the multipliers on alternating clocks.
Two sets of coefficients are required, one for data on the
forward decimation path, and one for data on the reverse
path. The filter output is generated by accumulating the
multiplier outputs for two CLKs.
13
FN2808.12
July 27, 2009
HSP43168
0
1
2
6
5
4
C1
C2
C3
7
C0
0
1
2
6
5
4
C6
C5
C4
C7
ACCUMULATOR
ACCUMULATOR
(X0)C0 + (X1)C1 + (X2)C2 + (X3)C3
+ (X7)C7 + (X6)C6 + (X5)C5 + (X4)C4
(X0)C0 + (X1)C1 + (X2)C2 + (X3)C3
FIGURE 13A. DATA SHIFTING DISABLED, BACKWARD
SHIFTING DECIMATION REGISTERS FEEDING
MULTIPLIERS
1
2
3
FIGURE 13B. SHIFTING OF DATA SAMPLE 7 INTO FIR CELL
ENABLED, FORWARD SHIFTING REGISTERS
FEEDING MULTIPLIERS
7
6
5
C1
C2
C3
1
2
3
7
6
5
C6
C5
C4
4
4
8
C7
C0
3
3
ACCUMULATOR
ACCUMULATOR
(X1)C0 + (X2)C1 +( X3)C2 + (X4)C3
FIGURE 13C. DATA SHIFTING DISABLED, BACKWARD
SHIFTING DECIMATION REGISTERS FEEDING
MULTIPLIERS
(X1)C0 + (X2)C1 + (X3)C2 + (X4)C3
+ (X8)C7 + (X7)C6 + (X6)C5 + (X5)C4
FIGURE 13D. SHIFTING OF DATA SAMPLE 8 INTO FIR CELL
ENABLED, FORWARD SHIFTING REGISTERS
FEEDING MULTIPLIERS
FIGURE 13. DATA FLOW DIAGRAMS FOR 8-TAP ASYMMETRIC FILTER
For this application, each filter cell is configured as an odd
length filter by writing 110H to the Control Register at
Address 000H. Even though an even tap filter is being
implemented, the filter cells must be configured as odd
length to ensure proper data flow. In addition, the filters
must be set to even symmetry. Also, the 4th bit at Control
Address 001H must be set to disable data reversal, and
TXFR must be tied low. Since an 8-tap asymmetric filter is
being implemented, two sets of coefficients must be stored.
14
These eight coefficients could be loaded into the first two
coefficient sets for FIR A by writing C0, C1, C2, C3, C7, C6,
C5, and C4 to address 100H, 101H, 102H, 103H, 108H,
109H, 10aH, and 10bH respectively.
The sum of products required for this 8-tap filter require
dynamic control over FWRD, RVRS, ACCEN, and
CSEL0-4. The relative timing of these signals is shown in
Figure 14.
FN2808.12
July 27, 2009
HSP43168
0
1
2
3
13
14
15
The alignment of data relative to the 24 filter coefficients for
a particular output is depicted graphically in Figure 16. As in
previous examples, the HSP43168 implements the filtering
operation by summing data samples prior to multiplication by
the common coefficient. In this example an output is required
every third CLK which allows 3 CLKs for computation. On
each CLK, one of three sets of coefficients are used to
calculate 8 of the filter taps. The Block Diagrams in Figure 17
show the data flow and accumulator output for the
data/coefficient alignment in Figure 16.
16
CLK
(NOTE)
X0
INA0-9
CSEL0-4
1
X1
0
X6
1
0
X7
0
1
X8
0
1
0
ACCEN
FWRD
Proper data and coefficient alignment is achieved by
asserting TXFR once every three CLKs to switch the LIFOs
which are being read and written. This has the effect of
feeding blocks of three samples into the backward shifting
decimation path which are reversed in sample order. In
addition, ACCEN is deasserted once every three clocks to
allow accumulation over three CLKs. The three sets of
coefficients required in the calculation of a 24-tap symmetric
filter are cycled through using CSEL0-4. The timing
relationship between the CSEL0-4, ACCEN, and TXFR are
shown in Figure 18.
RVRS
SHFTEN
(TIED LOW)
TXFR
NOTE: CLK is 2X data rate.
FIGURE 14. CONTROL TIMING FOR 8-TAP ASYMMETRIC
FILTER
Example 4: Even-Tap Decimating Filter Example
The HSP43168 supports filtering applications requiring
decimation to 16. In these applications the output data rate is
reduced by a factor of N. As a result, N clock cycles can be
used for the computation of the filter output. For example,
each FIR cell can calculate 8 symmetric or 4 asymmetric
taps in one clock. If the application requires decimation by
two, the filter output can be calculated over two clocks thus,
boosting the number of taps per FIR cell to 16 symmetric or
8 asymmetric. For this example, each FIR cell is configured
as an independent 24-tap decimate x3 filter. Again, the data
flow diagrams show only one of the FIR cells shown in
Figure 15.
HSP43168
h(n)
EVEN-TAP DECIMATING
INA0-9
A
A
B
B
INB0-9
C0 C1
FIR A
EVEN-TAP DECIMATING
To operate in this mode the Dual is configured by writing 1d2
to Address 000H via the microprocessor interface, CIN0-9,
A0-8, and WR. Data reversal must be enabled see (Table 2).
The 12 unique coefficients for this example are stored as
three sets of coefficients for either FIR cell. For FIR A, the
coefficients are loaded into the Coefficient Bank by writing
C2, C5, C8, C11, C1, C4, C7, C10, C0, C3, C6, and C9 to
Address [100H, 101H, 102H, 103H], CSEL = 0; [108H,
109H, 10aH, 10bH], CSEL = 1; [110H, 111H, 112H, and
113H], CSEL = 2, respectively.
M
U
X
OUT9-27
C4
C2 C3
C7
C5 C6
C10C11C11C10 C9
C8 C9
C8
24-TAPS
C7 C6
C5
C4 C3
C2
C1 C0
x(n)
FIR B
23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
FIGURE 15. EVEN-TAP DECIMATING FILTER, 24-TAP DEC = 3
15
8
7
6
5
4
3
2
1
0
FIGURE 16. DATA/COEFFICIENT ALIGNMENT FOR 24-TAP
DECIMATE BY 3 FIR FILTER
FN2808.12
July 27, 2009
HSP43168
2 1 0
5 4 3
8 7 6
20 19 18
+
+
C5
21 20 19
18 17 16
15 14 13
+
+
C1
C2
7 6 11
10 9
14 13 12
17 16 15
+
4 3 8
12
22
21
1 0 5
11 10 9
C8
C11
+
+
C4
C7
+
C10
CSEL = 1
CSEL = 0
ACCUMULATOR
ACCUMULATOR
(X2 + X21)C2 + (X5 + X18)C5 + (X8 + X15)C8 + (X11 + X12)C11
(X1 + X22)C1 + (X4 + X19)C4 + (X7 + X16)C7 + (X10 + X13)C10
+(X2 + X21)C2 + (X5 + X18)C5 + (X8 + X15)C8 + (X11 + X12)C11
FIGURE 17A. COMPUTATIONAL FLOW AS DATA SAMPLE 21 IS
CLOCKED INTO THE FEED-FORWARD STAGE
FIGURE 17B. COMPUTATIONAL FLOW AS DATA SAMPLE 22 IS
CLOCKED INTO THE FEED-FORWARD STAGE
0 5 4
3 8 7
6 11 10
9
12 13
23
22 21 20
19 18 17
+
C0
+
+
C3
5 4 3
16 15 14
14 13 12
+
C6
C9
24
23 22 21
+
CSEL = 2
ACCEN ASSERTED
AND ACTIVE
ACCUMULATOR
11 10 9
8 7 6
C2
+
C5
TXFR ASSERTED
AND ACTIVE
20 19 18
17 16 15
+
C8
+
C11
CSEL = 0
ACCUMULATOR
(X0 + X23)C0 + (X3 + X20)C3 + (X6 + X17)C6 + (X9 + X14)C9
+ (X1 + X22)C1 + (X4 + X19)C4 + (X7 + X16)C7 + (X10 + X13)C10
+ (X2 + X21)C2 + (X5 + X18)C5 + (X8 + X15)C8 + (X11 + X12)C11
(X5 + X24)C0 + (X8 + X21)C5 + (X11 + X18)C8 + (X14 + X15)C11
FIGURE 17C. COMPUTATIONAL FLOW AS DATA SAMPLE 23 IS
CLOCKED INTO THE FEED-FORWARD STAGE
FIGURE 17D. COMPUTATIONAL FLOW AS DATA SAMPLE 24 IS
CLOCKED INTO THE FEED-FORWARD STAGE
FIGURE 17. DATA FLOW DIAGRAMS FOR 24-TAP DECIMATED BY 3 FIR FILTER
0
1
2
4
3
21
5
22
23
CLK
INA0-9
0
1
2
3
4
CSEL0-4
0
1
2
0
1
5
2
21
22
23
0
1
2
ACCEN
Example 5: Odd-Tap Decimating Symmetric Filter
This example highlights the use of the HSP43168 as two
independent, 23-tap, symmetric, decimate by 3 filters. In this
example, the operational differences in the control signals
and data reversal structure may be compared to the
previously discussed even-tap decimating filter. Figure 19
shows two FIR cells. The data flow in this example uses only
one of the FIR cells.
HSP43168
FWRD †
ODD-TAP DECIMATING
RVRS †
INA0-9
SHIFTEN †
A
A
B
B
INB0-9
TXFR
FIR A
ODD-TAP DECIMATING
M
U
X
OUT9-27
FIR B
†Tied low.
FIGURE 18. CONTROL SIGNAL TIMING FOR 24-TAP
DECIMATE X3 FILTER
16
FIGURE 19. USING HSP43168 AS TWO INDEPENDENT FILTERS
FN2808.12
July 27, 2009
HSP43168
As in the 24-tap example, an output is required every third
CLK which allows 3 CLKs for computation. On each CLK,
one of three sets of coefficients are used to calculate the
filter taps. Since this is an odd length filter, the center
coefficient must be scaled by 1/2 to compensate for the
summation of the same data sample from the forward and
backward shifting decimation paths. The Block Diagrams in
Figure 20 show the data flow, and the accumulator output for
the data coefficient alignment is shown in Figure 21.
Proper data and coefficient alignment is achieved by asserting
TXFR once every three CLKs to switch the LIFOs which are
being read and written. In the odd-tap mode, TXFR is internally
delayed by one clock cycle with respect to ACCEN so that the
convolutional sum will be computed correctly. For odd length
filters, data prior to the last register in the forward decimation
path is routed to the feedback circuitry. As a result, TXFR
should be asserted one cycle prior to the input data samples
which align with the center tap. The timing relationship between
the CSEL0-5, ACCEN, and TXFR are shown in Figure 22.
2 1 6
3 2 1
6 5 4
9 8 7
20 19 18
17 16 15
14 13 12
13
21 20 19
22
21
+
+
+
+
C5
+
15 14 13
+
+
C4
C7
CSEL = 1
C10
CSEL = 0
C11/2
C8
18 17 16
+
C1
C2
11 10
8 7 12
5 4 9
12 11 10
ACCUMULATOR
ACCUMULATOR
(X3 + X21)C2 + (X6 + X18)C5 + (X9 + X15)C8 + (X12 + X12)C11/2
FIGURE 20A. COMPUTATIONAL FLOW AS DATA SAMPLE 21 IS
CLOCKED INTO THE FEED-FORWARD STAGE
TXFR TAKES AFFECT ON THIS CLOCK CYCLE
1 6 5
4 9 8
7 12 11
(X2 + X22)C1 + (X5 + X19)C4 + (X8 + X16)C7 + (X11 + X13)C10
+ (X3 + X21)C2 + (X6 + X18)C5 + (X9 + X15)C8 + (X12 + X12)C11/2
FIGURE 20B. COMPUTATIONAL FLOW AS DATA SAMPLE 22 IS
CLOCKED INTO THE FEED-FORWARD STAGE
10
13 14
23
22 21 20
+
C0
19 18 17
+
C3
16 15 14
+
C6
+
C9
ACCUMULATOR
24
6 5 4
9 8 7
12 11 10
23 22 21
20 19 18
17 16 15
+
CSEL = 2
ACCEN ASSERTED
AND ACTIVE
TXFR ASSERTED
(X1 + X23)C0 + (X4 + X20)C3 + (X7 + X17)C6 + (X14 + X10)C9
+ (X2 + X22)C1 + (X5 + X19)C4 + (X8 + X16)C7 + (X11 + X13)C10
+ (X3 + X21)C2 + (X6 + X18)C5 + (X9 + X15)C8 + (X12 + X12)C11/2
FIGURE 20C. COMPUTATIONAL FLOW AS DATA SAMPLE 23 IS
CLOCKED INTO THE FEED-FORWARD STAGE
C2
+
C5
+
C8
15 14 13
+
C11/2
CSEL = 0
ACCUMULATOR
(X6 + X24)C2 + (X9 + X21)C5 + (X12 + X18)C8 + (X15 + X15)C11/2
FIGURE 20D. COMPUTATIONAL FLOW AS DATA SAMPLE 24 IS
CLOCKED INTO THE FEED-FORWARD STAGE
TXFR TAKES AFFECT ON THIS CLOCK CYCLE
FIGURE 20. DATA FLOW DIAGRAMS FOR 23-TAP DECIMATE BY 3 SYMMETRIC FILTER
17
FN2808.12
July 27, 2009
HSP43168
h(n)
C0
C1 C2
C3
C4 C5
C6
C7 C8
C11
C10 C10
C9
C9
Example 6. Dual Decimation Example
23-TAPS
C8 C7
C6
C5 C4
C3
C2 C1
C0
x(n)
The purpose of this example is to give an overview of one of
the more complex applications of the HSP43168. The input
is two data streams (A) and (B) samples. Figure 23 shows
the upper level block diagram of the system being
implemented. The decimation rate was set to N. N-1 is
loaded into the decimation factor in Control Word 000H.
FS
2FS /(N+1)
23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FIGURE 21. DATA/COEFFICIENT ALIGNMENT FOR 23-TAP
DECIMATE BY 3 SYMMETRIC FILTER
B3, B2, B1, B0
INB0-9
A3, A2, A1, A0
HSP43168
INA0-9
DECIMATE BY N
0
1
2
4
3
20
5
21
22
CLK
INA0-9
0
1
2
3
CSEL0-4
0
1
2
0
4
1
5
2
21
22
23
0
1
2
ACCEN
FWRD †
RVRS †
SHIFTEN †
TXFR
†Tied low.
FIGURE 22. CONTROL SIGNAL TIMING FOR 23-TAP
SYMMETRIC FILTER
To operate in this mode, the Dual is configured by writing
112H to Address 000H via the microprocessor interface,
CIN0-9, A0-8, and WR. Data reversal must be enabled (see
Table 2). The 12 unique coefficients for this example are
stored as three sets of coefficients for either FIR cell. For FIR
A, the coefficients are loaded into the Coefficient Bank by
writing [C2, C5, C8, (C11)/ 2], CSEL = 0; [C1, C4, C7, C10],
CSEL = 1; [C0, C3, C6, and C9], CSEL = 2; to address
100H, 101H, 102H, 103H, 108H, 109H, 10aH, 10bH, 110H,
111H, 112H, and 113H, respectively.
18
OUT9-27
BOUT1
AOUT1
BOUT0
AOUT0
FIGURE 23. MULTIPLEXED DECIMATION BLOCK DIAGRAM
To demonstrate the muxed decimation, lets suppose that the
application requires filter A to be configured as an
even-decimate-by-3 filter and filter B to be configured as a
odd-decimate-by-3 filter. The output data is made of the two
decimated data streams multiplexed together and has a data
rate equal to 2x the input sampling rate divided by the
decimation factor. Figure 24 shows the data/coefficient
alignment for FIR A and FIR B.
To operate in this mode, Control Word 000H must be written
with a 0x152. Data reversal must be enabled by setting bit 4
of Control Word 001H = 0. The filter set selected by
CSEL0-4 = 0 should be loaded by writing C2, C5, C8, C11,
D2, D5, D8, and (D11)/ 2 into 100H, 101H, 102H, 103H,
104H, 105H, 106H, and 107H. The filter set selected by
CSEL0-4 = 1 should be loaded by writing C1, C4, C7, C10,
D1, D4, D7, and D10 into 108H, 109H, 10aH, 10bH, 10cH,
10dH, 10eH, and 10fH. The filter set selected by
CSEL0-4 = 2 should be loaded by writing C0, C3, C6, C9,
D0, D3, D6, and D9 into 110H, 111H, 112H, 113H, 114H,
115H, 116H, and 117H.
FN2808.12
July 27, 2009
HSP43168
h2(n)
D1
D0
D2
D3
D4
D5
D6
D7
D8
D9
D11
D10 D10
D9
D8
D7
D6
D5
Figure 25 shows the Timing Diagram required to obtained the
multiplexed/decimated output. The output of the two filters are
provided at by selecting the odd-decimation filter first, then the
even-decimation second using MUX0-1. Figure 26 shows the
Data Flow Diagram for the multiplexed decimation example.
23-TAPS
FIRB
D4
D3
D2
D1
D0
0
B(n)
1
2
4
3
5
20
21
22
CLK
23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8
h1(n)
C0 C1
C2 C3
C4
C5 C6
C7 C8
C11 C11
C10C9
C9C10
7 6 5 4 3 2 1 0
C8 C7
24-TAPS
FIRA
C6 C5
C4
C3 C2
C1 C0
INA0-9
0
1
2
3
CSEL0-4
0
1
2
0
5
1
2
20
21
22
2
0
1
ACCEN
11
MUX0-1
A(n)
4
10
11
11
10
11
TXFR
23 22 21 20 19 18 17 16 15 14 13 12 11 10 9
8
7
6
5
4
3
2
1
0
FIGURE 25. TIMING DIAGRAM FOR MULTIPLEXED
DECIMATION EXAMPLE
FIGURE 24. DATA/COEFFICIENT ALIGNMENT FOR
MULTIPLEXED DECIMATION EXAMPLE
FIR B
FIR B
3 2 1
6 5 4
9 8 7
20 19 18
17 16 15
14 13 12
2 1 6
13
12 11 10
B DATA
STREAM
21 20 19
22
21
+
+
+
+
D5
D8
18 17 16
+
15 14 13
+
+
+
D1
D2
11 10
8 7 12
5 4 9
D11/2
D4
D7
D10
CSEL = 1
CSEL = 0
ACCUMULATOR
ACCUMULATOR
(X2 + X22)D1 + (X5 + X19)D4 + (X8 + X16)D7 + (X11 + X13)D10
+ (X3 + X21)D2 + (X6 + X18)D5 + (X9 + X15)D8 + (X12 + X12)D11/2
(X3 + X21)D2 + (X6 + X18)D5 + (X9 + X15)D8 + (X12 + X12)D11/2
FIR A
A DATA
STREAM
2 1 0
21
20 19 18
8 7 6
1 0 5
11 10 9
10 9
7 6 11
4 3 8
12
+
C2
5 4 3
FIR A
+
C5
22
14 13 12
17 16 15
+
C8
+
C11
+
CSEL = 0
ACCUMULATOR
(X2 + X21)C2 + (X5 + X18)C5 + (X8 + X15)C8 + (X11 + X12)C11
FIGURE 26A. COMPUTATIONAL FLOW AS DATA SAMPLE 21 IS
CLOCKED INTO THE FEED-FORWARD STAGE
19
21 20 19
C1
18 17 16
+
C4
15 14 13
+
C7
+
C10
CSEL = 1
ACCUMULATOR
(X1 + X22)C1 + (X4 + X19)C4 + (X7 + X16)C7 + (X10 + X13)C10
+(X2 + X21)C2 + (X5 + X18)C5 + (X8 + X15)C8 + (X11 + X12)C11
FIGURE 26B. COMPUTATIONAL FLOW AS DATA SAMPLE 22 IS
CLOCKED INTO THE FEED-FORWARD STAGE
FN2808.12
July 27, 2009
HSP43168
FIR B
1 6 5
4 9 8
7 12 11
FIR B
10
6 5 4
9 8 7
12 11 10
23 22 21
20 19 18
17 16 15
15 14 13
13 14
23
22 21 20
+
D0
19 18 17
+
D3
24
16 15 14
+
D6
D9
ACCUMULATOR
+
+
CSEL = 2
D2
+
D5
OUTPUT OF B IS SENT
TO OUT9-27
(X1 + X23)D0 + (X4 + X20)D3 + (X7 + X17)D6 + (X14 + X10)D9
+ (X2 + X22)D1 + (X5 + X19)D4 + (X8 + X16)D7 + (X11 + X13)D10
+ (X3 + X21)D2 + (X6 + X18)D5 + (X9 + X15)D8 + (X12 + X12)D11/2
+
D8
+
D11/2
CSEL = 0
ACCUMULATOR
(X6 + X24)D2 + (X9 + X21)D5 + (X12 + X18)D8 + (X15 + X15)D11/2
FIR A
0 5 4
3 8 7
6 11 10
FIR A
9
5 4 3
11 10 9
8 7 6
12 13
23
22 21 20
+
C0
19 18 17
+
C3
16 15 14
+
C6
14 13 12
24
+
C9
23 22 21
+
CSEL = 2
C2
20 19 18
+
C5
17 16 15
+
C8
+
C11
CSEL = 0
OUTPUT OF A IS SENT
TO OUT9-27
ACCUMULATOR
ACCUMULATOR
(X0 + X23)C0 + (X3 + X20)C3 + (X6 + X17)C6 + (X9 + X14)C9
+ (X1 + X22)C1 + (X4 + X19)C4 + (X7 + X16)C7 + (X10 + X13)C10
+ (X2 + X21)C2 + (X5 + X18)C5 + (X8 + X15)C8 + (X11 + X12)C11
(X5+X24)C0+(X8+X21)C5+(X11+X18)C8+(X14+X15)C11
FIGURE 26C. COMPUTATIONAL FLOW AS DATA SAMPLE 23
IS CLOCKED INTO THE FEED-FORWARD STAGE
FIGURE 26D. COMPUTATIONAL FLOW AS DATA SAMPLE 24 IS
CLOCKED INTO THE FEED-FORWARD STAGE
FIGURE 26. DATA FLOW DIAGRAM FOR MULTIPLEXED DECIMATION EXAMPLE
20
FN2808.12
July 27, 2009
HSP43168
Absolute Maximum Ratings
Thermal Information
Supply Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +6.0V
Input, Output or I/O Voltage . . . . . . . . . . . . GND -0.5V to VCC +0.5V
ESD Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Class 1
Thermal Resistance (Typical, Note 2) . . . . . . . . . . . . . JA (°C/W)
MQFP Package . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33
PLCC Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24
Maximum Junction Temperature
MQFP and PLCC Packages . . . . . . . . . . . . . . . . . . . . . . . +150°C
Maximum Storage Temperature Range . . . . . . . . . .-65°C to +150°C
Pb-Free Reflow Profile. . . . . . . . . . . . . . . . . . . . . . . . .see link below
http://www.intersil.com/pbfree/Pb-FreeReflow.asp
Operating Conditions
Voltage Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5V 5%
Temperature Range, Commercial . . . . . . . . . . . . . . . . 0°C to +70°C
Temperature Range, Industrial. . . . . . . . . . . . . . . . . .-40°C to +85°C
Die Characteristics
Back Side Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .+5V
Number of Transistors or Gates. . . . . . . . . . . . . . . . . . . . . . . . 32529
CAUTION: Do not operate at or near the maximum ratings listed for extended periods of time. Exposure to such conditions may adversely impact product reliability and
result in failures not covered by warranty
NOTE:
2. JA is measured in free air with the component mounted on a high effective thermal conductivity test board with “direct attach” features. See
Tech Brief TB379.
DC Electrical Specifications
PARAMETER
SYMBOL
TEST CONDITIONS
MIN
MAX
UNITS
Power Supply Current
ICCOP
VCC = Max (Notes 4, 5, 6)
CLK Frequency 33MHz
-
363
mA
Standby Power Supply Current
ICCSB
VCC = Max, Outputs Not Loaded
-
500
µA
Input Leakage Current
II
VCC = Max, Input = 0V or VCC
-10
10
µA
Output Leakage Current
IO
VCC = Max, Input = 0V or VCC
-10
10
µA
Logical One Input Voltage
VIH
VCC = Max
2.0
-
V
Logical Zero Input Voltage
VIL
VCC = Min
-
0.8
V
Logical One Output Voltage
VOH
IOH = -400µA, VCC = Min
2.6
-
V
Logical Zero Output Voltage
VOL
IOL = 2mA, VCC = Min
-
0.4
V
Clock Input High
VIHC
VCC = Max
3.0
-
V
Clock Input Low
VILC
VCC = Min
-
0.8
V
Input Capacitance
CIN
CLK Frequency 1MHz
All measurements referenced
to GND. TA = +25°C, (Note 3)
-
12
pF
-
12
pF
Output Capacitance
COUT
NOTES:
3. Controlled via design or process parameters and not directly tested. Characterized upon initial design and after major process and/or changes.
4. Power Supply current is proportional to operating frequency. Typical rating for ICCOP is 11mA/MHz.
5. Output load per test load circuit and CL = 40pF.
6. Maximum junction temperature must be considered when operating part at high clock frequencies.
21
FN2808.12
July 27, 2009
HSP43168
AC Electrical Specifications
VCC = +4.75V to +5.25V, TA = 0°C to +70°C Commercial, TA = -40°C to +85°C Industrial (Note 7)
-33 (33MHz)
PARAMETER
SYMBOL
NOTES
-40 (40.8MHz)
-45 (45MHz)
MIN
MAX
MIN
MAX
MIN
MAX
(Note 11) (Note 11) (Note 11) (Note 11) (Note 11) (Note 11) UNITS
CLK Period
tCP
30
-
24.5
-
22
-
ns
CLK High
tCH
12
-
10
-
8
-
ns
CLK Low
tCL
12
-
10
-
8
-
ns
WR Period
tWP
30
-
24.5
-
22
-
ns
WR High
tWH
12
-
10
-
10
-
ns
WR Low
tWL
12
-
10
-
10
-
ns
Setup Time A0-8 to WR Going Low
tAWS
10
-
8
-
8
-
ns
Hold Time A0-8 from WR Going High
tAWH
0
-
0
-
0
-
ns
Setup Time CIN0-9 to WR Going High
tCWS
12
-
11
-
10
-
ns
Hold Time CIN0-9 from WR Going High
tCWH
1
-
1
-
1
-
ns
Setup Time WR Low to CLK Low
tWLCL
(Note 8)
5
-
4
-
3
-
ns
Setup Time CIN0-9 to CLK Low
tCVCL
(Note 8)
7
-
7
-
7
-
ns
Setup Time CSEL0-5, SHFTEN, FWRD, RVRS, TXFR,
INA0-9, INB0-9, ACCEN, MUX0-1 to CLK
Going High
tECS
15
-
13
-
12
-
ns
Hold Time CSEL0-5, SHFTEN, FWRD, RVRS, TXFR,
INA0-9, INB0-9, ACCEN, MUX0-1 to CLK
Going High
tECH
0
-
0
-
0
-
ns
CLK to Output Delay OUT0-27
tDO
-
14
-
13
-
12
ns
Output Enable Time
tOE
-
12
-
12
-
12
ns
Output Disable Time
tOD
(Note 9)
-
12
-
12
-
12
ns
Output Rise, Fall Time
tRF
(Note 9)
-
6
-
6
-
6
ns
NOTES:
7. AC tests performed with CL = 40pF, IOL = 2mA, and IOH = -400µA. Input reference level CLK = 2.0V. Input reference level for all other inputs is
1.5V. Test VIH = 3.0V, VIHC = 4.0V, VIL = 0V, VILC = 0V.
8. Setup time requirement for loading of data on CIN0-9 to guarantee recognition on the following clock.
9. Controlled via design or process parameters and not directly tested. Characterized upon initial design and after major process and/or changes.
10. Parameters with MIN and/or MAX limits are 100% tested at +25°C, unless otherwise specified. Temperature limits established by characterization
and are not production tested.
11. Parameters with MIN and/or MAX limits are 100% tested at +25°C, unless otherwise specified. Temperature limits established by characterization
and are not production tested.
AC Test Load Circuit
DUT
S1
Note: Test head capacitance.
CL (NOTE)
SWITCH S1 OPEN FOR ICCSB AND ICCOP
IOH

1.5V
IOL
EQUIVALENT CIRCUIT
22
FN2808.12
July 27, 2009
HSP43168
Waveforms
tCP
tCH
tCL
CLK
tECS
CSEL0 - 4, MUX0 - 1
tECH
SHFTEN, FWRD
RVRS, TXFR
INA0 - 9, INB0 - 9,
ACCEN
tDO
OUT0 - 27
tWLCL
tWP
tWH
tWL
WR
tAWH
tAWS
A0 - 8
tCWS
tCWH
CIN0 - 9
tCVCL
1.5V
OEL, OEH
1.5V
tOE
OUT0 - 27
HIGH
IMPEDANCE
tOD
1.7V
1.3V
HIGH
IMPEDANCE
FIGURE 27. OUTPUT ENABLE, DISABLE TIMING
2.0V
0.8V
tRF
2.0V
0.8V
tRF
FIGURE 28. OUTPUT RISE AND FALL TIMES
23
FN2808.12
July 27, 2009
HSP43168
Metric Plastic Quad Flatpack Packages (MQFP)
D
Q100.14x20 (JEDEC MS-022GC-1 ISSUE B)
D1
100 LEAD METRIC PLASTIC QUAD FLATPACK PACKAGE
-D-
INCHES
SYMBOL
-A-
-B-
E E1
e
PIN 1
SEATING
A PLANE
-H-
0.076
0.003
-C-
12o-16o
0.40
0.016 MIN
0.20
M
0.008
C A-B S
0o MIN
D S
b
A2 A1
0o-7o
L
12o-16o
b1
MILLIMETERS
MIN
MAX
MIN
MAX
NOTES
A
-
0.134
-
3.40
-
A1
0.010
-
0.25
-
-
A2
0.101
0.113
2.57
2.87
-
b
0.009
0.015
0.22
0.38
6
b1
0.009
0.013
0.22
0.33
-
D
0.908
0.918
23.08
23.32
3
D1
0.782
0.792
19.88
20.12
4, 5
E
0.673
0.681
17.10
17.30
3
E1
0.547
0.555
13.90
14.10
4, 5
L
0.029
0.040
0.73
1.03
-
N
100
100
7
e
0.026 BSC
0.65 BSC
-
ND
30
30
-
NE
20
20
Rev. 1 4/99
NOTES:
1. Controlling dimension: MILLIMETER. Converted inch
dimensions are not necessarily exact.
2. All dimensions and tolerances per ANSI Y14.5M-1982.
3. Dimensions D and E to be determined at seating plane -C- .
0.13/0.17
0.005/0.007
4. Dimensions D1 and E1 to be determined at datum plane
-H- .
BASE METAL
5. Dimensions D1 and E1 do not include mold protrusion.
Allowable protrusion is 0.25mm (0.010 inch) per side.
WITH PLATING
0.13/0.23
0.005/0.009
6. Dimension b does not include dambar protrusion. Allowable
dambar protrusion shall be 0.08mm (0.003 inch) total.
7. “N” is the number of terminal positions.
24
FN2808.12
July 27, 2009
HSP43168
Plastic Leaded Chip Carrier Packages (PLCC)
0.042 (1.07)
0.048 (1.22)
PIN (1) IDENTIFIER
0.042 (1.07)
0.056 (1.42)
0.004 (0.10)
C
0.025 (0.64)
R
0.045 (1.14)
0.050 (1.27) TP
C
L
N84.1.15 (JEDEC MS-018AF ISSUE A)
84 LEAD PLASTIC LEADED CHIP CARRIER PACKAGE
INCHES
SYMBOL
D2/E2
E1 E
C
L
D2/E2
VIEW “A”
0.020 (0.51)
MIN
A1
A
D1
D
MIN
MILLIMETERS
MAX
MIN
MAX
NOTES
A
0.165
0.180
4.20
4.57
-
A1
0.090
0.120
2.29
3.04
-
D
1.185
1.195
30.10
30.35
-
D1
1.150
1.158
29.21
29.41
3
D2
0.541
0.569
13.75
14.45
4, 5
E
1.185
1.195
30.10
30.35
-
E1
1.150
1.158
29.21
29.41
3
E2
0.541
0.569
13.75
14.45
4, 5
N
84
84
SEATING
-C- PLANE
0.020 (0.51) MAX
3 PLCS
0.026 (0.66)
0.032 (0.81)
0.013 (0.33)
0.021 (0.53)
0.025 (0.64)
MIN
0.045 (1.14)
MIN
6
Rev. 2 11/97
VIEW “A” TYP.
NOTES:
1. Controlling dimension: INCH. Converted millimeter dimensions are
not necessarily exact.
2. Dimensions and tolerancing per ANSI Y14.5M-1982.
3. Dimensions D1 and E1 do not include mold protrusions. Allowable
mold protrusion is 0.010 inch (0.25mm) per side. Dimensions D1
and E1 include mold mismatch and are measured at the extreme
material condition at the body parting line.
4. To be measured at seating plane -C- contact point.
5. Centerline to be determined where center leads exit plastic body.
6. “N” is the number of terminal positions.
All Intersil U.S. products are manufactured, assembled and tested utilizing ISO9001 quality systems.
Intersil Corporation’s quality certifications can be viewed at www.intersil.com/design/quality
Intersil products are sold by description only. Intersil Corporation reserves the right to make changes in circuit design, software and/or specifications at any time without
notice. Accordingly, the reader is cautioned to verify that data sheets are current before placing orders. Information furnished by Intersil is believed to be accurate and
reliable. However, no responsibility is assumed by Intersil or its subsidiaries for its use; nor for any infringements of patents or other rights of third parties which may result
from its use. No license is granted by implication or otherwise under any patent or patent rights of Intersil or its subsidiaries.
For information regarding Intersil Corporation and its products, see www.intersil.com
25
FN2808.12
July 27, 2009