2009 Microchip Technology Inc.
DS70286C-page 21
dsPIC33FJXXXGPX06/X08/X10
3.0
CPU
The dsPIC33FJXXXGPX06/X08/X10 CPU module has a
16-bit (data) modified Harvard architecture with an
enhanced instruction set, including significant support for
DSP. The CPU has a 24-bit instruction word with a variable
length opcode field. The Program Counter (PC) is 23 bits
wide and addresses up to 4M x 24 bits of user program
memory space. The actual amount of program memory
implemented varies by device. A single-cycle instruction
prefetch mechanism is used to help maintain throughput
and provides predictable execution. All instructions execute
in a single cycle, with the exception of instructions that
change the program flow, the double word move (MOV.D)
instruction and the table instructions. Overhead-free pro-
gram loop constructs are supported using the DO and
REPEAT
instructions, both of which are interruptible at any
point.
The dsPIC33FJXXXGPX06/X08/X10 devices have sixteen,
16-bit working registers in the programmer’s model. Each of
the working registers can serve as a data, address or
address offset register. The 16th working register (W15)
operates as a software Stack Pointer (SP) for interrupts and
calls.
The dsPIC33FJXXXGPX06/X08/X10 instruction set has
two classes of instructions: MCU and DSP. These two
instruction classes are seamlessly integrated into a single
CPU. The instruction set includes many addressing modes
and is designed for optimum C compiler efficiency. For most
instructions, the dsPIC33FJXXXGPX06/X08/X10 is capa-
ble of executing a data (or program data) memory read, a
working register (data) read, a data memory write and a
program (instruction) memory read per instruction cycle. As
a result, three parameter instructions can be supported,
allowing A + B = C operations to be executed in a single
cycle.
A block diagram of the CPU is shown in
Figure 3-1. The
programmer’s
model
for
the
3.1
Data Addressing Overview
The data space can be addressed as 32K words or
64 Kbytes and is split into two blocks, referred to as X and
Y data memory. Each memory block has its own indepen-
dent Address Generation Unit (AGU). The MCU class of
instructions operates solely through the X memory AGU,
which accesses the entire memory map as one linear data
space. Certain DSP instructions operate through the X and
Y AGUs to support dual operand reads, which splits the
data address space into two parts. The X and Y data space
boundary is device-specific.
Overhead-free circular buffers (Modulo Addressing mode)
are supported in both X and Y address spaces. The Modulo
Addressing removes the software boundary checking over-
head for DSP algorithms. Furthermore, the X AGU circular
addressing can be used with any of the MCU class of
instructions. The X AGU also supports Bit-Reversed
Addressing to greatly simplify input or output data
reordering for radix-2 FFT algorithms.
The upper 32 Kbytes of the data space memory map can
optionally be mapped into program space at any 16K pro-
gram word boundary defined by the 8-bit Program Space
Visibility Page (PSVPAG) register. The program to data
space mapping feature lets any instruction access program
space as if it were data space. The data space also includes
2 Kbytes of DMA RAM, which is primarily used for DMA
data transfers, but may be used as general purpose RAM.
3.2
DSP Engine Overview
The DSP engine features a high-speed, 17-bit by 17-bit
multiplier, a 40-bit ALU, two 40-bit saturating accumula-
tors and a 40-bit bidirectional barrel shifter. The barrel
shifter is capable of shifting a 40-bit value, up to 16 bits
right or left, in a single cycle. The DSP instructions operate
seamlessly with all other instructions and have been
designed for optimal real-time performance. The MAC
instruction and other associated instructions can concur-
rently fetch two data operands from memory while multi-
plying two W registers and accumulating and optionally
saturating the result in the same cycle. This instruction
functionality requires that the RAM memory data space be
split for these instructions and linear for all others. Data
space partitioning is achieved in a transparent and flexible
manner through dedicating certain working registers to
each address space.
3.3
Special MCU Features
The dsPIC33FJXXXGPX06/X08/X10 features a 17-bit by
17-bit, single-cycle multiplier that is shared by both the
MCU ALU and DSP engine. The multiplier can perform
signed, unsigned and mixed-sign multiplication. Using a
17-bit by 17-bit multiplier for 16-bit by 16-bit multiplication
not only allows you to perform mixed-sign multiplication, it
also achieves accurate results for special operations,
such as (-1.0) x (-1.0).
The dsPIC33FJXXXGPX06/X08/X10 supports 16/16 and
32/16 divide operations, both fractional and integer. All
divide instructions are iterative operations. They must be
executed within a REPEAT loop, resulting in a total execu-
tion time of 19 instruction cycles. The divide operation can
be interrupted during any of those 19 cycles without loss
of data.
A 40-bit barrel shifter is used to perform up to a 16-bit, left
or right shift in a single cycle. The barrel shifter can be used
by both MCU and DSP instructions.
Note:
This data sheet summarizes the features
of the dsPIC33FJXXXGPX06/X08/X10
family of devices. However, it is not
intended to be a comprehensive reference
source. To complement the information in
this data sheet, refer to Section 2. “CPU”
(DS70204) in the “dsPIC33F Family Ref-
erence Manual”, which is available from
the
Microchip
web
site
(www.microchip.com).