Microchip Technology MA330012 데이터 시트
©
2007 Microchip Technology Inc.
Preliminary
DS70165E-page 35
dsPIC33F
2.6.1
MULTIPLIER
The 17-bit x 17-bit multiplier is capable of signed or
unsigned operation and can multiplex its output using a
scaler to support either 1.31 fractional (Q31) or 32-bit
integer results. Unsigned operands are zero-extended
into the 17th bit of the multiplier input value. Signed
operands are sign-extended into the 17th bit of the
multiplier input value. The output of the 17-bit x 17-bit
multiplier/scaler is a 33-bit value which is
sign-extended to 40 bits. Integer data is inherently rep-
resented as a signed two’s complement value, where
the MSb is defined as a sign bit. Generally speaking,
the range of an N-bit two’s complement integer is -2
unsigned operation and can multiplex its output using a
scaler to support either 1.31 fractional (Q31) or 32-bit
integer results. Unsigned operands are zero-extended
into the 17th bit of the multiplier input value. Signed
operands are sign-extended into the 17th bit of the
multiplier input value. The output of the 17-bit x 17-bit
multiplier/scaler is a 33-bit value which is
sign-extended to 40 bits. Integer data is inherently rep-
resented as a signed two’s complement value, where
the MSb is defined as a sign bit. Generally speaking,
the range of an N-bit two’s complement integer is -2
N-1
to 2
N-1
– 1. For a 16-bit integer, the data range is
-32768 (0x8000) to 32767 (0x7FFF) including ‘
0
’. For a
32-bit integer, the data range is -2,147,483,648
(0x8000 0000) to 2,147,483,647 (0x7FFF FFFF).
(0x8000 0000) to 2,147,483,647 (0x7FFF FFFF).
When the multiplier is configured for fractional multipli-
cation, the data is represented as a two’s complement
fraction, where the MSb is defined as a sign bit and the
radix point is implied to lie just after the sign bit (QX
format). The range of an N-bit two’s complement
fraction with this implied radix point is -1.0 to (1 – 2
cation, the data is represented as a two’s complement
fraction, where the MSb is defined as a sign bit and the
radix point is implied to lie just after the sign bit (QX
format). The range of an N-bit two’s complement
fraction with this implied radix point is -1.0 to (1 – 2
1-N
).
For a 16-bit fraction, the Q15 data range is -1.0
(0x8000) to 0.999969482 (0x7FFF) including ‘
(0x8000) to 0.999969482 (0x7FFF) including ‘
0
’ and
has a precision of 3.01518x10
-5
. In Fractional mode,
the 16 x 16 multiply operation generates a 1.31 product
which has a precision of 4.65661 x 10
which has a precision of 4.65661 x 10
-10
.
The same multiplier is used to support the MCU multi-
ply instructions which include integer 16-bit signed,
unsigned and mixed sign multiplies.
ply instructions which include integer 16-bit signed,
unsigned and mixed sign multiplies.
The
MUL
instruction may be directed to use byte or
word sized operands. Byte operands will direct a 16-bit
result, and word operands will direct a 32-bit result to
the specified register(s) in the W array.
result, and word operands will direct a 32-bit result to
the specified register(s) in the W array.
2.6.2
DATA ACCUMULATORS AND
ADDER/SUBTRACTER
ADDER/SUBTRACTER
The data accumulator consists of a 40-bit
adder/subtracter with automatic sign extension logic. It
can select one of two accumulators (A or B) as its
pre-accumulation source and post-accumulation desti-
nation. For the
adder/subtracter with automatic sign extension logic. It
can select one of two accumulators (A or B) as its
pre-accumulation source and post-accumulation desti-
nation. For the
ADD
and
LAC
instructions, the data to be
accumulated or loaded can be optionally scaled via the
barrel shifter prior to accumulation.
barrel shifter prior to accumulation.
2.6.2.1
Adder/Subtracter, Overflow and
Saturation
Saturation
The adder/subtracter is a 40-bit adder with an optional
zero input into one side, and either true, or complement
data into the other input. In the case of addition, the
carry/borrow input is active-high and the other input is
true data (not complemented), whereas in the case of
subtraction, the carry/borrow input is active-low and the
other input is complemented. The adder/subtracter
generates Overflow Status bits, SA/SB and OA/OB,
which are latched and reflected in the STATUS
register:
zero input into one side, and either true, or complement
data into the other input. In the case of addition, the
carry/borrow input is active-high and the other input is
true data (not complemented), whereas in the case of
subtraction, the carry/borrow input is active-low and the
other input is complemented. The adder/subtracter
generates Overflow Status bits, SA/SB and OA/OB,
which are latched and reflected in the STATUS
register:
• Overflow from bit 39: this is a catastrophic
overflow in which the sign of the accumulator is
destroyed.
destroyed.
• Overflow into guard bits 32 through 39: this is a
recoverable overflow. This bit is set whenever all
the guard bits are not identical to each other.
the guard bits are not identical to each other.
The adder has an additional saturation block which
controls accumulator data saturation, if selected. It
uses the result of the adder, the Overflow Status bits
described above and the SAT<A:B> (CORCON<7:6>)
and ACCSAT (CORCON<4>) mode control bits to
determine when and to what value to saturate.
controls accumulator data saturation, if selected. It
uses the result of the adder, the Overflow Status bits
described above and the SAT<A:B> (CORCON<7:6>)
and ACCSAT (CORCON<4>) mode control bits to
determine when and to what value to saturate.
Six STATUS register bits have been provided to
support saturation and overflow; they are:
support saturation and overflow; they are:
1.
OA:
AccA overflowed into guard bits
AccA overflowed into guard bits
2.
OB:
AccB overflowed into guard bits
AccB overflowed into guard bits
3.
SA:
AccA saturated (bit 31 overflow and saturation)
or
AccA overflowed into guard bits and saturated
(bit 39 overflow and saturation)
AccA saturated (bit 31 overflow and saturation)
or
AccA overflowed into guard bits and saturated
(bit 39 overflow and saturation)
4.
SB:
AccB saturated (bit 31 overflow and saturation)
or
AccB overflowed into guard bits and saturated
(bit 39 overflow and saturation)
AccB saturated (bit 31 overflow and saturation)
or
AccB overflowed into guard bits and saturated
(bit 39 overflow and saturation)
5.
OAB:
Logical OR of OA and OB
Logical OR of OA and OB
6.
SAB:
Logical OR of SA and SB
Logical OR of SA and SB
The OA and OB bits are modified each time data
passes through the adder/subtracter. When set, they
indicate that the most recent operation has overflowed
into the accumulator guard bits (bits 32 through 39).
The OA and OB bits can also optionally generate an
arithmetic warning trap when set and the correspond-
ing Overflow Trap Flag Enable bits (OVATE, OVBTE) in
the INTCON1 register (refer to Section 6.0 “Interrupt
Controller”) are set. This allows the user to take
immediate action, for example, to correct system gain.
passes through the adder/subtracter. When set, they
indicate that the most recent operation has overflowed
into the accumulator guard bits (bits 32 through 39).
The OA and OB bits can also optionally generate an
arithmetic warning trap when set and the correspond-
ing Overflow Trap Flag Enable bits (OVATE, OVBTE) in
the INTCON1 register (refer to Section 6.0 “Interrupt
Controller”) are set. This allows the user to take
immediate action, for example, to correct system gain.