Motorola 6800
From Wikipedia, the free encyclopedia
The 6800 is an 8-bit microprocessor produced by Motorola and released shortly after the Intel 8080 in late 1974.[1] It had 78 instructions, including the (in)famous, undocumented Halt and Catch Fire (HCF) bus test instruction.[2] It may have been the first microprocessor with an index register.
It was usually packaged in a 40 pin DIP (dual-inline package).
Several first-generation microcomputers of the 1970s, available by mail order as kits or in assembled form, used the 6800 as their CPU; examples are the MEK6800D2 development board, the SWTPC 6800 (the first computer to use the 6800), the MITS Altair 680 range (MITS offered these as alternatives to its Altair 8800 which used the Intel 8080), several of the Ohio Scientific designs, Gimix, Smoke Signal Broadcasting, Midwest Scientific, and the Newbear 77/68.
The 4051, a professional grade desktop graphical system intended for user programming in BASIC, was manufactured and sold by Tektronix. This integrated a 6800 processor, memory card, storage display tube, keyboard, and magnetic tape cassette in a single unit and employed an external thermal imaging printer for hard copy.
The 6800 'fathered' several descendants, the pinnacle being the greatly extended and semi-compatible 6809, which was used in the Vectrex video game console and the TRS-80 Color Computer, among several others. There are also many microcontrollers descended from the 6800 architecture, such as the Motorola 6801/6803, 6805, RS08, 68HC08, 68HC11 and 68HC12.
Hitachi, Ltd. acted as a second source for many of Motorola's CPUs, and also produced its own derivatives including the 6301 and 6303, which could run 6800 code. These microprocessors also had a couple of extra instructions added to their instruction sets.
Competitor MOS Technology came up with an architectural relative of the 6800, with its 6502 ('lawsuit compatible' MPU) and its successors. The 6502 did not have the 16 bit registers of the 6800, but had more addressing modes and was substantially cheaper. The 6502 was used in many computers and game consoles during the late 1970s and early-to-mid-1980s (most notably the Atari 2600, Apple II, the Commodore PET, VIC-20 and Commodore 64, the Acorn Electron/BBC Microcomputer, and the Nintendo Entertainment System/NES).
The 6800 was supplanted by the Motorola 68000, used in large numbers in the Apple Macintosh family before the introduction of the PowerPC, a RISC technology developed by IBM and produced jointly with Motorola.
Contents |
[edit] Programming model
|
||||||||
|
||||||||
|
||||||||
|
||||||||
|
A - Accumulator A
B - Accumulator B
X - Index register
PC - Program Counter
SP - Stack Pointer
CCR - Conditional Code Register: Half carry, Interrupt mask, Negative, Zero, oVerflow and Carry
The 6800 is a standard "8-bit" 2's complement microprocessor, like the Intel 8080. It supports 8-bit bytes, 16-bit addresses and 64KB of memory. Like the 8080, code can be placed in ROM. This is an important feature compared to some of the minicomputers of the time, especially the PDP-8, which used a linked-list embedded in the code for subroutine linkage.
The 6800 had a 16-bit stack pointer, so the stack can be located anywhere and can grow to any length up to the size of the memory. Compare this with the 6502 which had only an 8-bit stack pointer and which mapped the stack into the second page (addresses 256 - 511).
Like the 8080, the 6800 supports multi-precision arithmetic using a carry flag and the ADC (add with carry) and SBC (subtract with carry) instructions. Decimal arithmetic is supported in a fashion similar to the 8080: there is a DAA (decimal adjust accumulator) instruction, which corrects the result of the 2's complement ADD on two packed decimal digits. Unlike the 6502, the 6800 also provided add and subtract without carry (ADD and SUB instructions).
Arithmetic instructions set the usual 2's complement flags: Sign, Zero, oVerflow and Carry. Unlike the 8080 and the 6502, the 6800 had a complete set of branch instructions, basically identical to those provided on the PDP-11, for comparing both signed and unsigned quantities. For example, if you want to compare i with j, then load i into an accumulator (LDAA i), compare or subtract with j (CMPA j or SUBA j) and then branch:
- BEQ - Branch if i equals j
- BNE - Branch if i does not equal j
- BHI - Branch if i is greater than j (unsigned)
- BHS - Branch if i is higher or same as j (unsigned)
- BLO - Branch if i is lower than j (unsigned)
- BLS - Branch if i is lower than or equal to j (unsigned)
- BGT - Branch if i is greater than j (signed)
- BGE - Branch if i is greater than or equal to j (signed)
- BLT - Branch if i is less than j (signed)
- BLE - Branch if i is less than or equal to j (signed)
- BPL - Branch if plus (positive or zero)
- BMI - Branch if minus (negative)
- BCC - Branch if Carry clear (same as BHS)
- BCS - Branch if Carry set (same as BLO)
- BVS - Branch if oVerflow set
- BVC - Branch if oVerflow clear
- BRA - Branch always
Unlike the PDP-11, 6502 and 8080 the 6800 was big endian, as was the IBM 360.
Unlike the PDP-11 and 6502, but like the 8080, the 6800 used borrow carry: the carry flag is set on a borrow during subtraction instead of clear.
Unlike the 8080 and especially the PDP-8, but like the PDP-11, the 6800 had no special purpose I/O instructions. Instead, all I/O devices share the same address space as memory.
Unlike the 8080, the 6800 had indexing, which allowed it to directly support data structures. The base address of a data structure is loaded into the index register, and then an 8-bit unsigned offset from the second byte of indexing instructions is added to it to read different items from the structure.
The 6800 had four addressing modes: immediate, indexed, extended and direct (or zero-page). The zero-page mode allows faster direct access to the first 256 bytes of memory and is similar to the zero-page mode of the PDP-8. The zero-page can be thought of as an extension of the register set, an idea taken much further with the zero-page indirect modes of the 6502.
The instruction set mnemonics owe a great deal to the PDP-11, and are burned into the brains of many assembly language programmers and computer architects. Many CPUs, including the ARM remained close to the form of the 6800 mnemonics.
[edit] 6800 Bus
Like the 8080, the 6800 defines a bus and was the center of a family of peripheral ICs, see the table below.
The bus was particularly simple: peripherals generally needed just three control signals in additional to the address and data lines:
- Φ2 or E (from MC6802): a clock cycle is defined as two successive falling edges of E.
- R/W a read cycle if high, or a write cycle if low.
- CS chip select if low (decoded from address lines and VMA).
The bus was clocked: setup and hold times are relative to falling edge of clock E. The 8080 bus was strobe signal based: setup and hold times are specified relative to the rising edges of MEM_RD, MEM_WR, IO_RD and IO_WR, which are signals generated by the 8228 system controller chip.
[edit] Clocking, Power
There were three speed grades: 6800 ran at 1 MHz, 68A00 ran at 1.5 MHz and 68B00 ran at 2 MHz. Instructions took between 2 (for LDAA immediate) to 12 (for SWI) cycles.
Like the 4004 and 8008, the 6800 required two non-overlapping clock inputs: Φ1 and Φ2. The clocks must not overlap to prevent races in transparent latch-based designs: latches which feed each other must not both be transparent at the same time. The 6800 used dynamic logic: the maximum high time for Φ1 and Φ2 was 9.5 μs. Several companion chips were available to generate these clocks: the MC6870 (built-in crystal), MC6871 (built-in crystal with clock stretching logic for slow devices) and MC6875 (external crystal, clock stretching and reset synchronizer).
The MC6802 was a 6800 with a built-in clock generator. A crystal could be directly connected to the chip. The MC6802 also had 128 bytes of RAM built-in and was intended to be used with an MC6846 for a complete two-chip microcomputer.
The 6800 required just a single 5V supply. This made it more appealing than the 8080, which required +12V, +5V and -5V.
[edit] Criticisms
There are a number of problems which were remedied in later devices, such as the 6809 and 6811:
- Only one pointer register. Although the stack pointer can be used in tight copy loops, it is awkward and interrupts have to be disabled for this trick. The 6800's most important competitor, the 8080 had three pointers (but no indexing).
- Stack instructions use post-decrement on push and pre-increment on pop instead of the more natural post-increment on pop and pre-decrement on push. This shows up when the stack pointer is transferred to the index register: the first thing on the stack is at offset 1.
- The index register can not be directly pushed or popped from the stack.
- The accumulators and index registers occupy different spaces and thus there are no instructions to transfer or operate between the two. For example there is no way to directly add an accumulator or even a constant to the index register. Later processors remedied this by adding the "ABX" (add B to X) and "LEA" (load effective address) instructions.
- The CPX (compare X) instruction does not affect the Carry flag, so direct magnitude comparisons with the index register are not possible.
- The DAA (decimal adjust) instruction only worked after addition, and not subtraction. To subtract packed decimal, one had to use 9's complement followed by addition. Other microprocessors solved this problem in various ways: The 6502 had a decimal mode bit: when set both the normal addition and subtraction instructions operated on packed decimal. The Z80 had a subtraction bit in its status flags. If the last operation was a subtraction, this bit was set to let the decimal adjust instruction know to operate differently. The 8086 simply had a DAS (decimal adjust for subtraction) instruction.
[edit] Designers
It's not clear if there was a chief architect. Some of the 6800 designers went on to join MOS Technology where they created the 6502. These included Chuck Peddle and Bill Mensch.[3]
Bill Mensch designed the MC6820.
[edit] Development Systems
- The Motorola Exorciser was a development system available for the 6800. It used plug in cards with 86-pin edge connectors and ran MDOS (Motorola Disk Operating System) off of 8 in. floppy disks.
- MIKBUG was an early debugger in ROM. It allowed loading of memory from paper tape using the S19 format.
- RASM was the relocatable assembler available on MDOS.
- MPL "Motorola Programming Language" was a PL/M like PL/I clone.
[edit] Peripheral IC's
- MC6810 128 bytes of RAM
- MC6818 Real Time Clock (used in IBM PC/AT)
- MC6820/6821 "PIA" Peripheral Interface Adapter (GPIO / parallel port)
- MC6828 "PIC" Priority Interrupt Controller
- MC6830 1024 bytes of ROM
- MC6840 Counter/Timer
- MC6843 Floppy disk controller
- MC6844 DMA controller
- MC6845 CRT Controller (use in VGA)
- MC6846 ROM + Counter/Timer + GPIO
- MC6847 Video Display Generator
- MC6850 "ACIA" Asynchronous Communications Interface Adapter ("async" serial port)
- MC6852 "SSDA" Synchronous Serial Data Adapter (can be used for floppy disk serialization)
- MC6854 "ADLC" Advanced Data Link Controller (HDLC/SDLC serial communications)
- MC6860 Low speed FSK modem
- MC6883 "SAM" Synchronous Address Multiplexer (DMA and DRAM controller for MC6847)
[edit] See also
[edit] References
- ^ Motorola (1976), Motorola Annual Report 1975, pp. 13, <http://www.motorola.com/mot/doc/6/6496_MotDoc.pdf> "We introduced our M6800 microprocessor system late in 1974 and it gained broad industry acceptance in 1975."
- ^ Wheeler, Gerry (December 1977). "Undocumented M6800 Instructions". BYTE 2 (12): 46-47. Halt and Catch Fire instruction.
- ^ Transcript of interview made on 9 October 1995 by Rob Walker for Stanford University's Silicon Genesis project Interview with Bill Mensch (html).
[edit] External links
- MC6800 applications manual from 1975- lots of information
- MDOS User's Manual
- Motorol Exorciser Emulator
- MIKBUG
- 680x images and descriptions at cpu-collection.de
- Instruction set summary
- Java Applet Simulator of a simplified M6800 Microprocessor
|
||||||||||||||||||||||||||
|
|||||
This article was originally based on material from the Free On-line Dictionary of Computing, which is licensed under the GFDL.
