首页> 外国专利> Binary-coded decimal adder with radix correction

Binary-coded decimal adder with radix correction

机译:具有基数校正的二进制编码的十进制加法器

摘要

1,103,384. Electronic computers. ING. C. OLIVETTI & C. S.p.A. 1 March, 1965 [2 March, 1964], No. 8683/65. Headings G4A and G4C An electronic computer has a cyclic serial memory in the form of a delay line and adapted to contain the contents of a plurality of registers arranged in bit interleaved fashion. Counting and addressing operations within each memory register are performed by utilizing tag bits associated with the stored data characters instead of by conventional counters and address registers. The computer operates in a sequence of statuses, the progression in the sequence being determined by the operation being performed. General arrangement.-The computer comprises a memory LDR, Fig. la, in the form of a magnetostrictive delay line and including ten 22-character registers I, J, M, N, R, Q, U, Z, D, E, each character comprising eight bits. The registers are interleaved by bit in the memory. Registers I and J are instruction registers each capable of storing 22 instructions, each instruction comprising a 4-bit operation portion B5-B8 and a 4-bit operand address portion B1-B4. The remaining registers are addressable by instructions and store numbers, each 8-bit character comprising four bits B5-B8 for a binary-coded-decimal digit, bits B1-B4 storing various tag bits: B4 for indicating the decimal point, B3 a sign bit, B2 indicating the presence of a significant decimal digit; B1 is a control bit. Thus the B1 bits in the first and last characters of registers R, E, respectively start and stop a pulse generator 44, Fig. 1b. Bits B1 in the register N indicate the location of the next instruction in registers I or J. Bits B1 in register M indicate the operative location in an input or print-out operation, and in an addition operation that digits in register N require correction by adding a filler digit. Bits B1 in register U indicate the location of an instruction interrupted for a subroutine. The computer includes a serial binary adder 72 (Fig. 4, not shown) having a flip-flop to store a decimal carry when the sum of two binary coded digits lies between ten and fifteen. A shift register K may be connected to the delay line to provide a 1-character delay in the circulation loop. The shift register K can also act as a counter by connecting it to the adder 72; as a buffer memory for an output printer 21; and as a parallel-to-serial converter for input data from the computer keyboard 22. The keyboard 22 comprises a ten-key numeric keyboard 65, an address keyboard 68 and a function keyboard 69, the three keyboards controlling a mechanical code-bar decoder co-operating with electric switches for producing the binary signals. The computer can execute instructions for the four rules, transfer from the register M, into the register N, print-out, programme stop, conditional and unconditional jump. The computer can operate in three modes: "manual," " automatic " and " entering programme " according to the position of a switch 23. In automatic operation, instruction-extract and instruction - execute phases alternate automatically. In manual operation, the input register M is automatically addressed and a selected arithmetic operation may be performed on a number entered via the keyboard into the register M. Bistable circuits in a " condition staticizer " circuit 25 (Fig. 6, not shown) are arranged to indications of various conditions in the computer. Thus a flip-flop A0 is controlled by the B2 bit positions of register M to be energized during the time significant digits are read out of the register M. Flip-flops A1, A2 perform the same function for registers N"Y where Y is a register being currently addressed. The computer operates in one of a plurality of statuses P1-Pn, the current status being indicated by the condition of a flip-flop P1- Pn, Fig. lb, the next following status being determined by a logical network 27. Printer.-The printer 21 comprises a rotatable type drum co-operating with a hammer movable stepwise parallel to the axis of the drum, a timing disc controlling the character to be printed. Starting computer operation.-Initially, a general reset button AG is depressed. Next, a start button AV is depressed which causes the computer to enter status P21 in which the shift register K is connected with the adder 72 and acts as a counter, thereby causing the control circuits to enter a start bit in the register R and a stop bit in the register E, these bits being effective to control the pulse generator 44 during subsequent operations. Number entry. Status P21 is followed by status P0 in which data may be entered into the memory register M from the keyboard, the register M being connected to the shift register K to form a closed loop. If a negative number is to be entered, depression of a minus sign key causes a negative sign bit B3 = " 1 " to be written in all the decimal denominations of the register M. The required numeric key is then depressed, and subsequently the binary coded value of the digit is entered in that denomination of register M which is first available after the operation of the numeric key, together with a tag bit in the B1 position of that character. Subsequent digits are entered in a similar way, each new digit being entered in the decimal denomination preceding the denomination of the last entered digit, under the control of the B1 bit whose denomination is shifted with each entry. A decimal point is entered by depressing decimal point key 67 after entering the units digit whereby a decimal point is written in bit position B4 of the units digit character. In manual operation, in the status P0, a number is first entered, followed by the address of a register, followed by the entry of a function on the function keyboard 69, the address and function being transferred to an instruction register 16 to effect the desired operation. Transfer between registers.-This is effected during a complete memory cycle during which the computer is in status P2. A switching network 36 effecting the desired connections between the registers. Alignment of numbers in the registers.- Alignment to bring the units digit into the first decimal position is effected during status P3, the switching network 36 connecting the output and input of the register to be aligned to the input and output of the shift register K so that during each memory cycle the content of the register K is delayed by one digit period, until the decimal point indicator is detected in the first position of the register. Similarly, in status P14, a number may be shifted until its most significant digit is in the first position. Preparatory to print out, a number may be shifted to bring its least significant digit into the first decimal denomination, thereby eliminating non-significant zeros. Sign comparison.-In status P9, a circuit 64 (Fig. 4, not shown) is effective to compare the signs of the numbers in two registers, a bi-stable device A8, energized at the beginning of the comparison, being de-energized should disagreement occur. Addition and subtraction.-Addition is effected in a plurality of memory cycles. In a first cycle, the computer is in status P5, a carry being transmitted if the sum of two decimal digits is greater than 9. A tag bit is recorded in each decimal denomination in bit positions B1 of the sum register M to indicate any necessary correction to the sum digit. In the second memory cycle, the computer is in status P6, a filler digit 6 is added to each sum digit which has produced a decimal carry. In the case of subtraction, which is effected by addition of complements, the appropriate filler digit corrections are also effected during the second memory cycle under the control of tag bits B1, except that if the minuend is less than the subtrahend, a third and fourth memory cycle are required for the addition of unity and correction of this new result. The addition is performed by first aligning the two numbers in registers M and N with respect to their decimal points, the switching network 36 being then effective in status P5 to connect the registers M and N to the adder 72. On conclusion of the addition or subtraction operation, the next following status is either P17 (extract the next instruction) if the computer is in automatic mode or P18 (print out the first addend) if the computer is in the manual mode. In a modified arrangement for addition and subtraction, in a first memory cycle in which the computer is in status P40, the number M is added to the complemented number N to determine whether N is greater than M or not. In a second memory cycle, M is added to N, the greater of the two numbers being complemented to 15 if a subtraction is required. In a third memory cycle, in which the computer is in status P60, digit correction is performed by adding the filler digit + 6 to each sum digit which has produced a final binary carry. The digits of the result are recomplemented if a subtraction is required. Multiplication and division.-These are performed by repeated addition and shift; and repeated subtraction and shift, respectively, the sequence of statuses being as in Fig. 8b (not shown). Print-out.-A print-out operation requires a sequence of three computer statuses P18, P19, P22. In status P18, the shift register K is connected to the adder 72 so as to act as a counter for counting pulses derived from the printer timing disc, a bi-stable device A7 (Fig. 6, not shown) compensating for the phase difference between the signals from the disc and the clock pulses from the generator 44. The shift register K is initially preset with the number to be printed, and when the value 16 is reached, the printing hammer is actuated. The contents of the instruction staticizer 16 are first printed out followed by the contents of the addressed register, which is connected by the switching network 36 to the shift register K. A tag bit B1 in register M is employed to identify the next digit to be printed. A
机译:1,103,384。电子计算机。 ING。 C.OLIVETTI&C.S.p.A. 1965年3月1日[1964年3月2日],第8683/65号。标题G4A和G4C电子计算机具有延迟线形式的循环串行存储器,并适于包含以比特交错方式排列的多个寄存器的内容。每个存储寄存器中的计数和寻址操作是通过利用与存储的数据字符关联的标记位来执行的,而不是通过传统的计数器和地址寄存器来执行的。计算机以状态序列进行操作,该序列的进程由所执行的操作确定。总体布置-计算机包括一个图1a所示的存储器LDR,其形式为磁致伸缩延迟线,并包含10个22个字符的寄存器I,J,M,N,R,Q,U,Z,D,E,每个字符包括八位。寄存器按位在存储器中交错。寄存器I和J是指令寄存器,每个指令寄存器能够存储22个指令,每个指令包括4位操作部分B5-B8和4位操作数地址部分B1-B4。其余寄存器可通过指令寻址并存储数字,每个8位字符包括4位B5-B8用于二进制编码的十进制数字,位B1-B4存储各种标记位:B4用于指示小数点,B3是符号位,B2指示存在有效的十进制数字; B1是控制位。因此,寄存器R,E的第一个和最后一个字符中的B1位分别启动和停止脉冲发生器44,图1b。寄存器N中的B1位指示下一条指令在寄存器I或J中的位置。寄存器M中的B1位指示输入或打印输出操作中的操作位置,以及在加法运算中,寄存器N中的数字需要通过以下方式进行校正:添加一个填充数字。寄存器U中的位B1指示为子例程中断的指令的位置。该计算机包括一个串行二进制加法器72(图4,未示出),该串行二进制加法器72具有一个触发器,用于在两个二进制编码数字之和在十与十五之间时存储一个十进制进位。移位寄存器K可以连接到延迟线以在循环回路中提供1个字符的延迟。通过将移位寄存器K连接到加法器72,它也可以用作计数器;作为输出打印机21的缓冲存储器;键盘22包括一个十键数字键盘65,一个地址键盘68和一个功能键盘69,这三个键盘控制一个机械编码解码器。与电子开关配合产生二进制信号。计算机可以执行以下四个规则的指令:从寄存器M传输到寄存器N,打印输出,程序停止,有条件和无条件跳转。计算机可以根据开关23的位置以三种模式进行操作:“手动”,“自动”和“进入程序”。在自动操作中,指令提取和指令-执行阶段会自动交替进行。在手动操作中,输入寄存器M被自动寻址,并且可以对通过键盘输入到寄存器M中的数字执行选择的算术运算。“条件静态器”电路25中的双稳态电路(图6,未示出)是可操作的。可以指示计算机中各种状况。因此,触发器A0由寄存器M的B2位位置控制,以便在从寄存器M中读出有效数字的期间通电。触发器A1,A2对寄存器N“ Y执行相同的功能,其中Y为计算机在多个状态P1-Pn之一中操作,当前状态由触发器P1-Pn的状态表示,图1b,下一个随后的状态由逻辑确定。网络27.打印机。-打印机21包括一个可旋转的鼓,与一个可平行于鼓的轴线逐步移动的锤子配合,一个定时盘控制着要打印的字符。接着,按下启动按钮AV,使计算机进入状态寄存器P21,在状态P21中,移位寄存器K与加法器72连接并用作计数器,从而使控制电路进入启动位。寄存器R和as寄存器E的高位,这些位在随后的操作中有效地控制脉冲发生器44。号码输入。状态P21之后是状态P0,在状态P0中,数据可以从键盘输入到存储寄存器M中,寄存器M连接到移位寄存器K形成闭环。如果要输入一个负数,则按下减号键会导致负号位B3 =“ 1”写入寄存器M的所有十进制面额。然后按下所需的数字键,然后将数字的二进制编码值输入到寄存器M的该名称中,该值首先在操作数字键之后与该字符的B1位置中的标记位一起使用。随后的数字以类似的方式输入,每个新数字都在B1位的控制下以最后输入的数字的形式在十进制形式中输入,而B1的名称随每次输入而移动。在输入单位数字后,通过按小数点键67来输入小数点,从而在单位数字字符的位位置B4中写入小数点。在手动操作中,在状态P0中,首先输入数字,然后输入寄存器的地址,然后在功能键盘69上输入功能,该地址和功能被传送到指令寄存器16以实现该功能。所需的操作。寄存器之间的传输。-这是在计算机处于状态P2的完整存储周期内完成的。交换网络36实现寄存器之间的期望的连接。寄存器中数字的对齐。-在状态P3期间进行对齐以使单位数字进入第一个小数位,交换网络36连接寄存器的输出和输入,以使其与移位寄存器K的输入和输出对齐因此,在每个存储周期中,寄存器K的内容将延迟一个数字周期,直到在寄存器的第一个位置检测到小数点指示符为止。类似地,在状态P14中,数字可以移位直到其最高有效数字在第一位置。准备打印时,可以将数字移位以将其最低有效数字带入第一个十进制面额,从而消除不重要的零。符号比较。-在状态P9中,电路64(图4,未示出)有效地比较两个寄存器中的数字的符号,在比较开始时通电的双稳态装置A8被去耦。如果发生分歧,请充满电。加法和减法-加法在多个存储周期中进行。在第一个循环中,计算机处于状态P5,如果两个十进制数字的总和大于9,则传送进位。在总和寄存器M的位位置B1中的每个十进制单位中记录一个标记位,以指示任何必要的更正总和数字。在第二个存储周期中,计算机处于状态P6,填充数字6加到每个产生十进制进位的总和数字上。在减法的情况下,通过加补码来实现,在第二个存储周期中,在标记位B1的控制下,也会进行适当的填充数字校正,除非减数小于减数,则第三和第四为了增加统一性和更正此新结果,需要存储周期。通过首先将寄存器M和N中的两个数字相对于它们的小数点对齐来执行加法,然后交换网络36在状态P5中有效以将寄存器M和N连接到加法器72。减法运算时,如果计算机处于自动模式,则下一个状态为P17(提取下一条指令),如果计算机处于手动模式,则下一个状态为P18(打印第一个加数)。在用于加法和减法的修改布置中,在计算机处于状态P40的第一存储周期中,将数字M添加到补数N中,以确定N是否大于M。在第二个存储周期中,将M加到N,如果需要减法,则将两个数字中的较大者补码为15。在计算机处于状态P60的第三个存储周期中,通过将填充数字+ 6加到已产生最终二进制进位的每个和数字上来执行数字校正。如果需要减法,则对结果的数字进行补码。乘法和除法-这些是通过重复加法和移位来执行的;分别进行重复减法和移位,状态顺序如图8b所示(未显示)。打印输出-打印输出操作需要一系列三个计算机状态P18,P19,P22。在状态P18中,移位寄存器K连接到加法器72,以用作对从打印机定时盘导出的脉冲进行计数的计数器,双稳态装置A7(图6,未示出)补偿相位差。在盘上的信号和发生器44的时钟脉冲之间进行转换。移位寄存器K最初预先设置有要打印的数字,并且当达到值16时,激活打印锤。首先打印指令静态化器16的内容,然后打印寻址寄存器的内容通过交换网络36连接到移位寄存器K。寄存器M中的标记位B1用于标识要打印的下一位。一种

著录项

  • 公开/公告号US3304418A

    专利类型

  • 公开/公告日1967-02-14

    原文格式PDF

  • 申请/专利权人 ING. C. OLIVETTI & C. S. P. A.;

    申请/专利号US19650435813

  • 发明设计人 SANDRE GIOVANNI DE;PEROTTO PIER GIORGIO;

    申请日1965-03-01

  • 分类号G06F9/00;G06F3/02;G06F3/06;G06F7/495;G06F7/50;G06F9/40;G06F15/02;G06F15/78;G11C21/00;

  • 国家 US

  • 入库时间 2022-08-23 13:48:11

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号