PDP-11 , Электроника-60, СМ-1420 и им подобные
Главная | Каталог статей | Регистрация | Вход
Приветствую Вас Гость | RSS

Дорогие посетители.Убедительно просим Вас Загружать имеющуюся у Вас информацию (Файлы образов дисков, Исходники, Книги и т.д.) касающуюся ПЭВМ с архитектурой PDP-11.

Меню сайта
  • Создание сайтов в Усть-каменогорске
  • Люди на сайте
    Друзья Нравится
    Google
    Форма входа
    Категории раздела
    Contrib [88]
    Статьи, взятые из разных источников.
    My atticle [2]
    Статьи,написаные лично мною
    Forex [22]
    Всё о заработке на рынке FOREX
    Поиск
    Облако тэгов
    RT-11 LSI-11 PDP-11 Последняя версия RT-11 БК ДВК бк-0010 приставка Бортника сопроцессор Кроскомпилятор схема ANDOS БК-0011 КНГМД RADIX-50 Rtty БК0010 радиосвязь Сеть УКНЦ 11/70 11/83 11/93 RSX-11
    Прогноз Forex
    Котировки от Forex4You
    Открыть реальный счет
    Статистика
    Счетчик тИЦ, PR и обратных ссылок Resurs.kz: сайты Казахстана и раскрутка сайта
    Онлайн всего: 1
    Гостей: 1
    Пользователей: 0
    Поиск
    Главная » Статьи » Contrib

    The PDP-11 Maintenance Panel and Front Panel

    The PDP-11 Maintenance Panel and Front Panel

    The PDP-11 Maintenance Panel was introduced with the 11/20. It comprised of two single height boards, a lamp driver and a lamp/switch board. Through the use of the four switches, the processor clock could be interrupted and single stepped through the ISR and BSR timing states of the 11/20. Timing state numbers, Condition Codes and certain other flags were displayed on the maintenance panels 16(?) lamps. Of course, having the machine frozen in suspended animation also facilitated the troubleshooting of the internal circuits by oscilloscope.

    On the 11/40 there were two slots into which a maintenance panel could be plugged. In the first slot you could single step the processor clock and cycle through the microcode program. The processor could be made to stop at a specific microcode address (MPC) set in the front panel switches. The panel could be plugged into a different slot and perform the same function for the Floating Point Processor. If you thought that the PDP-11 had wide and complex micro-instructions, you should have seen the FPP!

    The panel also performed the same function on the 11/05 (no floating point) and was also used on the 11/70. A good technician would have been able to determine a fault on an 11/40 and 11/70 using nothing more than the maintenance panel, front panel, and knowledge of the inter-relationships of the processor's data paths.

    [I am unclear whether the Maint Panel was used in later PDP-11s. I seem to remember that certain of the micro-stepping and micro-trapping functions were incorporated into the calculator-style front panels of the 11/04, /34, /44 and /60. And I draw a total blank re: the later Micro-PDP-11's. Perhaps someone can fill in some detail here? -MMcC]

    The front panel on the PDP-11 went through some major revisions. The original 11/20, 11/15, and later the 11/05 and 11/10, 11/40 and 11/35, 11/45, 11/50, 11/55, 11/70 and 11/74 were equipped with what the purists regard as the "real" front panels. Subsequent machines (mostly the LSI-11s) had either chicklet-calculator octal panels, or just a trio of switches of various purposes (always including Enable/ Halt). These latter had Console ODT, instead.

    The 11/20 panel is laid out roughly as (not to scale):

     +-------------------------------------------------------+
    | ADDRESS REGISTER RUN BUS FETCH EXEC |
    | +---+---+---+---+---+---+ +---+---+ +-----+----+ |
    | |ooo|ooo|ooo|ooo|ooo|ooo| | o | o | | o | o | |
    | +---+---+---+---+---+---+ +---+---+ +-----+----+ |
    | DATA SRC DST ADDR |
    | +-+---+---+---+---+---+ +---+---+ +----+ |
    | |o|ooo|ooo|ooo|ooo|ooo| |ooo|ooo| | ooo| |
    | +-+---+---+---+---+---+ +---+---+ +----+ |
    | SWITCH REGISTER L E C H S T D |
    |\|/ +---+---+---+---+---+---+ +-+-+-+-+-+-+ +-+ |
    | O |ttt|ttt|ttt|ttt|ttt|ttt| |m|m|m|t|t|m| |m| |
    | +---+---+---+---+---+---+ +-+-+-+-+-+-+ +-+ |
    +-------------------------------------------------------+

    The small circles are incandescent lamps (i.e., prone to burning out). The odd thing at the lower-left is the keyswitch. The 10 o'clock position is OFF, straight up is POWER, and 2 o'clock is PANEL LOCK.

    The lower-case `t's are toggle switches. The lower-case `m's are momentary-contact switches. They are all large paddles, coloured alternately magenta and orange. The left bank of switches are used to set the address and data registers. The action switches on the right are:

    • L LOAD ADDR, copies the switch register to the address register
    • E EXAM, loads the data register from the memory location indicated in the address register
    • C CONT, continues the processor from a HALT
    • H ENABLE/HALT, enables the processor to run if up, stops it if down
    • S S-INST/S-CYCLE, selects the style of single-stepping
    • T START, resets the system and starts the processor
    • D DEP, deposits the contents of the switch register in the memory location indicated by the address register. This switch must be pushed up to operate, unlike any of the others.

    The way ENABLE/HALT, CONT, and START work are (this is true of an 11/70 as well, and presumably of everything above an 11/40):

    • ENABLE/HALT: If in the HALT position the machine will HALT at the proper moment, depending on the S-INST/S-CYCLE. If in the ENABLE position, the S-INST/S-CYCLE is ignored. The 11/40 also halts the bus arbitrator, which has the side-effect of killing off any in-progress DMA transfers. The 11/45 and 11/70 do the right thing, other models have not been reported on as yet.
    • CONT will just start the processor at where the PC points.
    • START will issue a bus reset before starting.

    If you want to do a bus reset without running the machine, you switch to HALT, and depress START. I don't think it executes anything at that time, which would imply that START can do less than CONT, which will single-step in this case.

    The 11/05 economized on the front panel. LEDs replaced lamps and there was only a single row of them. They were multiplexed, and whether they displayed data or address depended on the function being performed. Action switches remained similar to the 11/20 with the possible omission of the single step. [Help needed here ...]. The switches were small and made of white plastic.

    The other machines mentioned above had serious front panels. Great big triangular wedge shaped keys; two rows of data and address lines; and function lights. The 11/40 had:

     Run, Bus, User,
    Processor, Console, Virtual

    The 11/45 and 11/55 had:

     User I Super I Kernel I Prog Phy
    User D Super D Kernel D Cons Phy

    Data Paths u-Address FPP/CPU
    BUS Register Display Register

    ...and a pair of rotary switches to select between the two sets of displays!

    The 11/70 display was similar if slightly different in detail. Of course it had 22 of those nice big Data/Address switches!

    With the advent of the 11/04, 11/34, and 11/60, two consoles were offered. One could go the minimalist approach and purchase the cheapie and get a Boot and Halt switch along with a Run lamp. Not exactly loved by those that needed to interact with the machines intimately! Alternatively one could purchase the more fully featured (KY11-B ?) panel with the chicklet keys and calculator face. Many technicians had their own personal chicklet panels for troubleshooting those machines ordered with the minimal panel.

    The 11/60 was shipped with only the calculator-style panel. I seem to recall that DEC might have relented and done likewise with later 11/34 models.

    More recent PDP-11s seem to have nothing more than a power switch, although I believe a button or two and a couple of LED's might lurk under the covers if you know where to look for them.

    Категория: Contrib | Добавил: un7jks (25.11.2009)
    Просмотров: 4353 | Рейтинг: 0.0/0 |

    Хостинг от uCoz