I'm 23, and my first computer wasn't an Apple, Trash-80, or that ilk -- it was called CARDIAC, which stood for CARDboard mumble mumble Computer. It was made of cardboard, and measured about 6 inches by 12. The left half was the "CPU," with a little flow diagram of instructions, a sliding strip of cardboard in the input slot, and another in the output slot. There was a four digit accumulator and two three digit registers. The right half had 100 memory cells (numbered 00 - 99), and the memory pointer was a ladybug-like thing that you stuck in the hole of the pointed-to cell. Each cell could hold a three digit word. (Everything was in decimal). The leading digit determined the instruction, the other two were usually a memory pointer or a constant. As I recall, the instructions were something like: Read a number from the input into memory location xx, Write a number from memory location xx to the output, Read the (bottom three digits of) the accumulator into xx, Write from memory location xx to the accumulator, Add the number in location xx to the accumulator, Add the constant xx to the accumulator, Subtract the contents of xx from the accumulator, Subtract xx from the accumulator, Shift the accumulator x bits to the right and y to the left, Jump to location xx (possibly conditionally..) Whatever they were, there were ten. The only opcodes I recall are zero and eight. Zero was the read from input to memory, which I remember because memory cell 00 was hard-wired (i.e., printed in ink) with 001 -- with this one could actually write a bootstrap program! Eight was the shift instruction. Come to think of it, I think 9 was the jump instruction, because memory cell 99 had its first digit hard-wired to be a 9, to avoid segmentation violations. I have a vague feeling this teaching aid was put out by somebody respectable, like Bell Labs. Anyway, my fifth/sixth grade teacher used this to prepare us for the future.. >> My first computer was a mechanical thing I think came from sears. >> It was a three-bit machine I think - I remember it had three little >> windows where either a 0 or a 1 popped up. You programmed it by >> sticking little white plastic tubes on prongs that stuck out the back, >> and you executed cycles by grabbing the end of a plastic slide >You bet. Mine came from Edmund Scientific, I think. I mentioned it here >a few months ago, and someone overseas asked me in email how it worked... >but to tell the truth, I couldn't quite recall. I remember it had vertical >metal bars also, and rubber bands. Mine was from Edmund Scientific, too--made of red and white plastic (mostly) and called something like "Digicomp I". I remember the manual had a bunch of nifty programs in it (complete with coding forms) for exciting things you could do with one 3-bit register (or 3 1-bit registers, if you prefer). The most exciting thing it could do was play Nim. Maybe it could do a serial (1-bit-at-at-time with carry) binary addition, too. (Well, I'm sure it *could*, and that would be one of the obvious things to include, so I think it was there.) I remember being very frustrated--I must have been about 10 years old at the time--in that the manual did not explain the logical principles by which the machine operated. The manual just gave the "programs" and some high-sounding verbiage about how each of them related to an Important Concept in the operation of "real" digital computers (the kind with vacuum tubes). I had not the wits to figure out that staring at the thing until it was clear how it worked was the central point of the entire exercise. So I stared for a long time, but never quite figured it out before the machine broke or got lost in a household move... But from what I do remember, I think that the way it must have worked was this: There were three horizontal sliders, each of which could be in either of 2 positions (set or cleared), indicated by a big digit 1 or 0 that showed through a window on the "front panel". The three digits lined up one over another, so you could read them as a binary number. (I remember learning to count up to seven in binary using this machine.) On the "backplane" each slider had 6 sets of flat, rectangular frobs (labeled T and F in the original, but I have labeled them 0 and 1 in the diagram below). They protruded about 5 mm perpendicular to the direction of motion of the slider: if you looked at the back of the machine with the slider running from left to right, the frobs stuck out toward you. You programmed the machine by putting short bits of plastic tubing on some subset of the frobs. The "coding sheets" in the manual were simply a diagrammatic representation of the back of the machine (or was it the front?) with the placement of the bits of tubing marked. | | | | | | ---0-1------0-1------0-1------0-1------0-1------0-1----- Numbers ---0-1------0-1------0-1------0-1------0-1------0-1------ attached here ---0-1------0-1------0-1------0-1------0-1------0-1----- show in windows | | | | | | set clear set clear set clear bit 0 bit 0 bit 1 bit 1 bit 2 bit 2 A thin vertical metal rod (indicated by | in the diagram -- WARNING: NEW USERS SHOULD NOT ATTEMPT TO READ THIS USING SCANDINAVIAN ASCII) ran across the middle of the range of motion of the frobs--i.e. the places where either a 0 frob or a 1 frob could be, depending on the state of each slider. When you generated a clock pulse by moving the Big Red (or was it white?) Slider that lay beneath this "logic" section, the metal rod, actuated by the clock slider and by rubber bands, "looked for" the ABSENCE of little plastic pieces of tubing protruding outward from the frobs: if it was not blocked by plastic pieces on any of the frobs lying beneath it, the rod engaged a mechanism (located above the three "logic" sliders) which set or cleared bits by moving the three horizontal sliders. A compact description of the arithmetic and Boolean functions that this setup could compute is left as an exercise for the reader. I THINK that what each vertical wire did was "hardwired" into the machine, as indicated on the diagram (not nececessarily in the order given). I could be wrong. Maybe you could program which bits got set or cleared by a particular vertical wire by putting more bits of tubing on other frobs. (There was a T column and a F column on the coding form, but I think there was a third column, too.) I do remember that there were some extra plastic bits--sort of Y-shaped--that you could install to OR together two of the vertical wires and make either of them trigger the setting (or clearing, as the case might be) of the bit that one of them normally controlled. To do this you had to give up the setting or clearing function of the other wire. Edmund Scientific also offered an analog computer (the usual 3-potentiometer electrical slide-rule type, I think--I never could afford one!) and a fancier mechanical computer (Digicomp II?) that worked by dropping marbles down an incline full of mechanical flip-flops. (Well, I don't really know if it was fancier, since I never owned one. I wanted to, though--it sure looked like more fun than the one I had.) I don't recall ever seeing any of these toy computers at Sears (where the original poster thought his might have come from) or in the Sears catalog. But they were all in the Edmund Scientific catalog for many years--and may still be, for all I know! Newsgroups: alt.folklore.computers From: lauren@vortex.com (Lauren Weinstein) Subject: Re: Plastic mechanical computers circa 1970? Date: Sat, 30 Apr 1994 06:37:08 GMT In article <1994Apr29.193907.9609@midway.uchicago.edu>, Eric Fischer wrote: >In article <2pqee8$ptm@euas20.eua.ericsson.se> etxelv@eua.ericsson.se writes: >> Does anybody remember little mechanical computers that you could >>build about 20 years or so ago? .... ... >It was called the Digi-Comp, I think... Indeed. The unit was a kit made of white and red plastic, and consisted of three horizontal levers (one per bit) and a number of vertical levers held by little metal springs. It was called the DigiComp I. A completely different unit, called the DigiComp II, used marbles on a slanted board. I got a DigiComp I. The big problem with the DigiComp I was that the little metal springs would constantly fly loose and get lost. I ended up replacing them all with tiny rubber bands (I had bags of them from the orthodontic braces I had at the time!). They worked, but caused the levers to be much harder to operate than with the springs (I ended up experimenting with a range of different rubber band types that came in different colored envelopes at the orthodontist's office). While I haven't seen it for years, it is almost certain that I still have it around somewhere. --Lauren-- Newsgroups: alt.folklore.computers Subject: Re: Plastic mechanical computers circa 1970? From: mshapiro@netlink.nix.com (Michael Shapiro) Date: Sat, 30 Apr 94 21:31:29 PDT etxelv@eua.ericsson.se (Eric Valentine) writes: > Does anybody remember little mechanical computers that you could > build about 20 years or so ago? My uncle gave me one that was about > half finished along with such landmark books as _Electronic Digital > and Hybrid Computers_. No instructions. I aped what he had done and > finished building the thing which seemed to operate with a system of > interlocking levers. Unfortunately, my levers didn't lever. Didn't > budge at all. Thus began my career as bug-builder, I suppose. > Can anybody ID this thing? I don't know if it's the same unit you have, but I'm looking at a little plastic, rubber band, and wire pin computer called a DIGI-COMP1. It was produced (in kit form, of course) by a company named ESR, located in Montclair, New Jersey. The box and manual show a copyright date of 1963, so this 30+ year old thing may not be the one you're thinking of. It has three plastic flip-flops that slide back and forth, a clear and set clock, a reset clock, and some other pieces. The manual shows 15 experiments, including an automatic elevator, a bank lock, the game of NIM (a computer game!), etc. -- INTERNET: mshapiro@netlink.nix.com (Michael Shapiro) UUCP: ...!ryptyde!netlink!mshapiro Network Information eXchange * Public Access in San Diego, CA (619) 453-1115 From: bgrant@umcc.umcc.umich.edu (Bruce Grant) Newsgroups: alt.folklore.computers Subject: Re: Plastic Mechanical Computers circa 1970 Date: 1 May 1994 03:22:15 -0400 In article <7eqmLc3w165w@netlink.nix.com>, Michael Shapiro wrote: >etxelv@eua.ericsson.se (Eric Valentine) writes: > >> Does anybody remember little mechanical computers that you could >> build about 20 years or so ago? My uncle gave me one that was about >> half finished along with such landmark books as _Electronic Digital >> and Hybrid Computers_. No instructions. I aped what he had done and >> finished building the thing which seemed to operate with a system of >> interlocking levers. Unfortunately, my levers didn't lever. Didn't >> budge at all. Thus began my career as bug-builder, I suppose. >> Can anybody ID this thing? > >I don't know if it's the same unit you have, but I'm looking at a little >plastic, rubber band, and wire pin computer called a DIGI-COMP1. It was >produced (in kit form, of course) by a company named ESR, located in >Montclair, New Jersey. The box and manual show a copyright date of 1963, >so this 30+ year old thing may not be the one you're thinking of. It has >three plastic flip-flops that slide back and forth, a clear and set >clock, a reset clock, and some other pieces. The manual shows 15 >experiments, including an automatic elevator, a bank lock, the game of >NIM (a computer game!), etc. > >-- This sounds like a shameless knock-off of the Geniac, a Boolean algebra machine which you programmed not in machine language but in hardware (in this case, wires, light bulbs, jumpers and circular pieces of pegboard for switches, etc.) to solve a variety of simple problems. Back in 1956 you could buy one of these babies for $17.95 from Edmund Berkley & Associates, who described themselves even then as "makers and exhibitors of small robots". Bruce Grant (bgrant@umcc.ais.org) Newsgroups: alt.folklore.computers Subject: Re: Plastic Mechanical Computers circa 1970 From: mshapiro@netlink.nix.com (Michael Shapiro) Date: Sun, 01 May 94 21:35:35 PDT bgrant@umcc.umcc.umich.edu (Bruce Grant) writes: > > In article <7eqmLc3w165w@netlink.nix.com>, > Michael Shapiro wrote: ... > >I don't know if it's the same unit you have, but I'm looking at a little > >plastic, rubber band, and wire pin computer called a DIGI-COMP1. It was > >produced (in kit form, of course) by a company named ESR, located in > >Montclair, New Jersey. The box and manual show a copyright date of 1963, > >so this 30+ year old thing may not be the one you're thinking of. It has > >three plastic flip-flops that slide back and forth, a clear and set > >clock, a reset clock, and some other pieces. The manual shows 15 > >experiments, including an automatic elevator, a bank lock, the game of > >NIM (a computer game!), etc. > > > >-- > This sounds like a shameless knock-off of the Geniac, a Boolean algebra > machine which you programmed not in machine language but in hardware > (in this case, wires, light bulbs, jumpers and circular pieces of > pegboard for switches, etc.) to solve a variety of simple problems. > > Back in 1956 you could buy one of these babies for $17.95 from Edmund > Berkley & Associates, who described themselves even then as "makers and > exhibitors of small robots". I don't think the DIGI-COMP1 was a "knock-off" of anything. It was a strictly mechanical implementation of digital logic. For more information on Edmund C. Berkeley's work, see his 1949 book "Giant Brains, or Machines That Think." It was published by John Wiley and Sons and should be available in any good computer science library. In chapter 3 of that book, Berkeley (correct spelling) describes the design of Simple Simon, "a very simple mechanical brain." A friend and I tried to build one in the mid 1950s. Somehow it never worked. Wish we could have had a kit. Anyway, it was nice to get a strictly mechanical unit a few years later. When I was teaching computer science, I used it as a demonstration of a finite-state machine. -- INTERNET: mshapiro@netlink.nix.com (Michael Shapiro) UUCP: ...!ryptyde!netlink!mshapiro Network Information eXchange * Public Access in San Diego, CA (619) 453-1115 Newsgroups: alt.folklore.computers From: armb@setanta.demon.co.uk (Alan Braggins) Subject: Re: Plastic mechanical computers circa 1970? Date: Tue, 3 May 1994 08:43:47 GMT In article dp@world.std.com (Jeff DelPapa) writes: > than the standard IBM issue sense of humor. It included a description > of digital logic using things that looked like concertina bellows with > hoses at one end, and boxing gloves at the other. I understand that for a while "fluidics" was considered a more promising technology for computing than electrics, because it was possible to build hydraulic switching elements smaller, and faster, and/or with lower power requirements than using relays. In particular, you can make a chamber with one inlet and two outlets shaped such that a flow to either outlet is stable, but can be "knocked across" to the other outlet by a brief flow from auxiliary inputs at right angles. -- Alan Braggins armb@setanta.demon.co.uk abraggins@cix.compulink.co.uk "Any technology distinguishable from magic is insufficiently advanced" GCS/E d-(!) p c++ !l+ u@ e++(*) m+ s+/ n+* h---(fiance - wedding 21/5/94) f g-(+) w+ t@ r+ y(see h) Newsgroups: alt.folklore.computers From: jerome.yuzyk@freddy.supernet.ab.ca (Jerome Yuzyk) Subject: Re: Plastic mechanical computers circa 1970? Date: Tue, 3 May 94 12:03:00 -0700 -> I have a pendulum and weight driven clock that is made mostly from the -> (paperboard) pages of the book in which the parts and instructions are -> printed. Make your own Working Paper Clock James Smith Rudoplh, preface by Isaac Asimov Harper & Row ISBN 0-06-091066-6 I've been working on mine in my spare time for the last year or so. .-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-.-^-. Jerome Yuzyk jerome.yuzyk@freddy.supernet.ab.ca BRIDGE Scientific Services Edmonton Alberta Canada * RM 1.3 00857 * I can quit anytime I want; I just don't want to! Newsgroups: alt.folklore.computers From: rsf@mother.idx.com (Rob Freundlich) Subject: mechanical Turing Machine Date: Mon, 22 Aug 1994 10:20:11 UNDEFINED The thread on mechanical computers caught my eye, as I just read an article in Scientific American (in the mathematical diversions section, I forget the title) about building a Turing machine from a toy train set! To summarize: Using three types of switches and a combination of straight and curved track, you can build a Turing machine. You build an infinite :-) number of identical sections of track. There are n tracks connecting each section, where n is the number of states for the TM. Each section of track consists of two parts. The shell allows a train on any state-track to enter and exit the core (next paragraph) identically from either direction. The core has the logic in it. It consist of a "read/write" head for each state-track (this will hold the "digit" of the cell), a line of track that will flip the value of each r/w head, and track to direct the train to various paths depending on the state rules. The three types of switches are all variants on a Y-shaped track, ie: \ / \ / \ / \/ || || || || || || The train can do three things: 1. Enter from the "upright" and go left. Call this "0" 2. Enter from the upright and go right. Call this "1" These are your two digits. The third thing it can do is enter from the left or right and go down the upright. In a "simple" switch, entering from the left or right sets the switch so that the next time the train comes up the upright, it will exit the same direction. That is, if the train comes in from the left, next time it enters from below it will exit left. This a simple switch can be written by entering from left (0) or right (1) and read be entering from the bottom. In a "sprung" switch, one of the left/right branches is blocked so that the train can enter from that branch but not exit. Thus, if the left side is blocked, the train will *always* go right when entering from the bottom. In a "flip-flop" switch, each time the switch is passed through, it flips value. So if it's originally 0 and a train enters from the bottom, it exits left and the value flips to 1. Next time it'll exit right and the value flips to 0. The read/write head consist of a sprung switch to hold the digit, and a "write" line that somehow flips the sprung switch. I forget the exact structure. The article goes into much more detail, with pictures and an example. I highly recommend it. --- Rob Freundlich, Senior Software Engineer | Some folks you don't have IDX Systems Corporation | to satirize, you just quote 'em. | - Tom Paxton, from "A Folk Amiga 4000/040. deal with it. | Singer's Guide to Usenet" Newsgroups: alt.folklore.computers From: jh@cadre.com (Joe Hartley) Subject: CARDIAC call! (was: Ancient home computers was: Apple Lisa) Date: Wed, 24 Aug 1994 20:31:51 GMT >>andy@shell.portal.com (andy) writes: >>Somewhere in my apartment, I have a Cardiac. Wasn't it put out by >>Bell Labs, as a freebie for phone companies to give to science >>teachers? > >I got mine by sending a letter an a dollar for postage to Bell Labs; >I'd read about it in a magazine. > I had a CARDIAC years ago, which Mom consigned to the trash heap sometime after I moved out but before I went back to reclaim my "stuff". It indeed was put out by Bell Labs and distributed to schools; I got mine from a teacher who was cleaning out a storage space and thought I'd like some of the stuff in there. (I also got a couple of slide rules and a 4' diameter planetarium!) CARDIAC gave me my first lesson in computers, and I was hooked. Let's see, what do I remember of it? It opened like a book, and had the "CPU" on the right, and "storage" on the left. Storage was 100 little boxes that had a hole punched in each one. You'd write the instruction or data (3 decimal digits only!) into each cell. The current cell was indicated by a little cardboard ladybug with a long nose that you'd put into the hole. The CPU had 3 sliding pieces of cardboard, one for each of the 3 digits in CARDIAC's "byte". The first digit was the opcode, (yep, only 10 instructions in its set!) with instructions like jump, add, subtract, store and read. The last 2 digits represented the memory location in the storage. The book that came with it was extremely well done; I remember the epiphany of understanding how this thing multiplied numbers larger than its storage space. If anyone out there still has one, I'd like to get my hands on one, either permanently (I'll entertain the concept of buying it!) or temporarily so I may copy it and make my own working model. Hmm, maybe I'll write a simulator for my Sun. THAT would be seriously retro! --- ========================================================================== Joe Hartley - jh@cadre.com - Cadre Technologies, Inc. Without deviation from the norm, "progress" is not possible. - Frank Zappa ... and in sports, Shoemaker-Levy 9, Jupiter 10. From: domiller@ualr.edu (Dale Miller) Newsgroups: alt.folklore.computers Subject: Re: CARDIAC call! (was: Ancient home computers was: Apple Lisa) Date: 24 Aug 94 16:35:47 CST > If anyone out there still has one, I'd like to get my hands on one, > either permanently (I'll entertain the concept of buying it!) or > temporarily so I may copy it and make my own working model. Hmm, maybe > I'll write a simulator for my Sun. THAT would be seriously retro! I bought 1 (NEW!) a couple of years ago from: Comspace Corp. 243 Dixon Ave. Amityville, NY 11701 (516) 789-0700 (516) 789-0980 Fax I have no affiliation with Comspace other than as a customer. Sorry, I don't remember the price, other than too much. Dale P.S. If you get the simulator working, I want a copy. There used to be one for an IBM OS/MVT system that I used (too many) years ago in an engineering course. P.P.S. For anyone interested, I could probably make a scan of the parts. -- Dale O. Miller - domiller@ualr.edu | University of Arkansas at Little Rock Systems Programmer | 2801 S. University Ave. Voice: +1 501 569 8714 | Little Rock, AR 72204-1099 USA From: m-jb3281@granny.cs.nyu.edu (Jonathan Baker) Newsgroups: alt.folklore.computers Subject: Re: LEGO computer Date: 7 Sep 1994 20:53:44 -0500 In alt.folklore.computers you write: > >should be represented by swinging levers, rotating gears, etc... , and I've >worked out enough theory that once I solve the final problem I'll be well on >my way. The final problem is this: > I can't build an AND gate. >Lever System - For an AND gate, the output lever moves a little when only one >input is activated. I can't figure out how to build a "rectifier" that will >convert this movement into no movement at all without resorting to an >inelegant solution like rubber bands. > I have studied A. K. Dewdeney's pulley computer, but I would like to >keep strings and rubberbands out of this. Have you contacted Danny Hillis, founder of Thinking Machines? While he was at MIT, he built a Tinkertoy tic-tac-toe machine, which is on display at the Computer Museum in Boston. How are you doing your AND-gate? I thought about it for a few minutes, in terms of IC design. If I remember correctly, there are two input lines which can trigger part of the output, but the output is held at 1 by the "pullup" transistor; only when both inputs are 1 is there enough current flow to pull the output down. Seems to me you might have two levers for input, with their far ends leaning on the output lever. In the meantime, the output lever is counterweighted with the "pullup" weight. Only when both levers are pushed over, is there enough force on the output lever to counteract the "pullup" weight. Of course, balancing the counterweights would be one difficult point; another would be making some kind of rectifiers so that the un-pushed output lever on one input does not affect the preceding gates in the logic path. Say, something like this: "P" Weight \--/ \/ ------------0------> out ^ ^ | | A ---0--+ | B ---0-----+ where 0 is a fulcrum wheel. Pull down A or B to raise them as inputs. They would exert upward force. Well, I suppose that's what you did, and there would be some small movement of the output lever. Hmm, thought, if the output lever only moves a small amount, as a rectifier, you could mount a second lever after the output lever as an inverter, but offset a bit vertically from the output lever: the small motion on one high input would not touch this lever, but the big motion would touch and move it. Jonathan Baker baker@sacco.nyu.edu Newsgroups: alt.folklore.computers From: glen@beca.ece.jcu.edu.au (Glen Harris) Subject: Re: LEGO computer Date: Thu, 8 Sep 94 02:45:47 GMT In alt.folklore.computers you write: >should be represented by swinging levers, rotating gears, etc... , and I've >worked out enough theory that once I solve the final problem I'll be well on >my way. The final problem is this: > I can't build an AND gate. EASY! Radio controlled systems use it. The diagram is thus: ----> | ---->/ | / O--------> O--------> | / ----> | ---->/ The arrows represent the two inputs and the output, the "O" is a pivot. When one input is pushed, the pivot swings, when BOTH are pushed, the output moves. Of course, there may have to be some secondary levers in the inputs and outputs to set up the throw distances right, since the end that is not pushed will move back toward it's input a certain distance before contact. The right hand diagram represents the case where the top input is pushed, but the botton one isn't. -- * * * * * * * * * * * * * * * * * * * "Thats fifty green fires and hot * * * * * * * Glen Harris * * * * * * * * * * leads to go, with a side order * * * * * * * * 4harrig@wench.ece.jcu.edu.au * for blisters and scorpions. * * * * * * * * * * * * * * * * * * * * * * * * * Hold the mercy." - Conina (Sourcery) * * From: lgroebe@iadfw.net Newsgroups: alt.folklore.computers Subject: DIGICOMP Toy Computer (was Screwy things...) Date: 20 Jul 1995 23:10:40 GMT David K. Cornutt, Residentially Engineered, Huntsville, AL USA, writes: > >1964? --Back in my preschool >days, my father bought me a computer, sort of. See, there >was this company that made various kinds of educational >toys that were supposed to demonstrate principles of >mechanics and electronics. One of the things they sold was a >sort of mechanical computer. It was programmed by putting >short lengths of plastic tubing on various pegs; these >were sensed by metal rods. It had a 3-bit register that could >be set by the user, and was displayed using plastic cards >with "0" and "1" stickers, that slid back and forth behind >windows. Power came from a hand crank. I guess it was >about 6" high, 12" long, and 6" deep. It had what >seemed like a zillion plastic parts that took me two >weeks to put together. Of course, it didn't do a lot, >but I do recall that it could do basic logic operations >and some arithmetic. Being only 5 years old, I never >did grasp the programming principle (I just put the >tube bits where the examples in the book said to), and >I managed to break the thing after a few weeks. --- I know that machine! That was the DIGICOMP, an oddball toy computer indeed. I was given mine in 1967. Your description is a pretty good match with my own memory...except you left out the little rubber bands which made sure that the metal rods would be pressed against the plastic posts. I was able to completely confuse my fifth grade class by trying to explain binary numbers back in 1968 - even the teacher had no comprehension. (Of course, later that same day another student stood up to claim Haley's comet was going to crash into the Earth in 1968...) The book of "programs" that came with it included routines for counting, one to play NIM, and something that (I swear) claimed to predict the weather. Even at a young age I could see the latter was gonna be tough with a display that counted only up to 7 (in binary). But I became a whiz at Nim. :-) I would KILL to get my hands on a DIGICOMP again...does anyone have one? I want to build an emulator...(!) * * * There were a number of interesting Toy computers in the 50s and 60s -- the Geniac, cardia, and Think-atron are three that come to mind. And then all those curious Radioshack analog computer kits and the like... From: pearl@spectacle.sw.stratus.com (Dan Pearl) Newsgroups: alt.folklore.computers Subject: Re: DIGICOMP Toy Computer (was Screwy things...) Date: 21 Jul 1995 17:47:30 GMT ESR made several toy computers. I've got them all somewhere... Digicomp - It wasn't as hard to program as the other poster implied, nor as tough to assemble (it took about 5 minutes). The little rubber bands to hold the metal logic sense rods against the logic tubes were replaced by springs in a later version. Digicomp II - This one was about two feet long, 14" wide, and 3" deep, and stood on an incline, because it used gravity power to send marbles along the course. This was quite an engineering accomplishment, but unfortunately, it never quite worked for me. The nastiest part was that there was a marble-routing deck suspended BELOW the main deck. The marbles would fall into holes onto the deck below, and then exit, via a flimsy ramp, onto the main deck. Flip flops would bounce or jam, the marbles would hang up in the below-deck, requiring endless disassembly and reassembly. Think-a-Dot - A colorful blue/yellow display of 8 dots on the front, and three entrance chutes for marbles on the top. This was the most bullet proof of all the ESR products. It was cute. Dr. Nim - A red inclined field with white moving plastic parts. It played nim flawlessly. A simple machine, but it taught me about inductive logic and game theory when I was 9. -- ------------------------------------------------------------------------------ Dan Pearl ** Stratus Computer, Inc. ** pearl@sw.stratus.com From: david@pangloss.micro.ti.com (David Thomas) Newsgroups: alt.folklore.computers Subject: Re: DIGICOMP Toy Computer (was Screwy things...) Date: 22 Jul 1995 17:04:01 GMT Dan Pearl (pearl@spectacle.sw.stratus.com) wrote: > ESR made several toy computers. I've got them all somewhere... > Digicomp - It wasn't as hard to program as the other poster implied, nor > as tough to assemble (it took about 5 minutes). The little rubber > bands to hold the metal logic sense rods against the logic tubes > were replaced by springs in a later version. > Digicomp II - This one was about two feet long, 14" wide, and 3" deep, and > stood on an incline, because it used gravity power to send > marbles along the course. This was quite an engineering > accomplishment, but unfortunately, it never quite worked for me. > The nastiest part was that there was a marble-routing deck > suspended BELOW the main deck. The marbles would fall into > holes onto the deck below, and then exit, via a flimsy ramp, > onto the main deck. Flip flops would bounce or jam, the marbles > would hang up in the below-deck, requiring endless disassembly > and reassembly. I found the Digicomp II to be a lot more fun. If I recall, it had a four-bit register, a six-bit register/counter, and something like a nine- or eleven-bit accumulator, compared to just one three-bit register in the Digicomp I. It was easier to program (flipping levers rather than pushing little tubes onto pegs), and it could perform multiplication and division. The analogy to a real digital computer was better, too. Whether you found these toys frustrating or fun depended a lot on the precision of the parts. Some units worked a lot better than others. -- David Thomas (david@micro.ti.com) Texas Instruments, Houston (713)-274-2347 From: mwandel@bnr.ca (Markus Wandel) Newsgroups: alt.folklore.computers Subject: Neat old digital logic toy Date: 3 Jan 1996 17:56:51 GMT I was reading a newspaper article the other day, which described a digital "computer" toy made by the German toy manufacturer Kosmos in 1968 or so. I've never seen this toy, lost the article, and forgotten the name of the toy, but I found it fascinating. It consists of 10 slide switches, 10 lamps, jumper wires, and a power supply. The switches are elaborate. They contain 10 SPDT sections. Five are on in one position, the other five in the other. That is, the switch is either O O O O O O O O O O | | | | | O O O O O O O O O O or O O O O O O O O O O | | | | | O O O O O O O O O O where "O" is a terminal and "|" is a connection made by the switch. Each connection point "O" is brought out as several (it looked like 3) holes into which wires can be plugged. The device is "programmed" with jumper wires. For example, to program the logic function "lamp 1 lights when (switch 1 is on or switch 2 is on) and switch 3 is off" one merely routes power to the lamp through appropriate contacts. Limited in complexity by the number of switch contacts and jumper wires, an arbitrary combinatorial function of 10 inputs and 10 outputs can be built up! Without any transistors or diodes or any other electronics. The picture showed that each switch and each lamp had room for a label (transparent for the lamps) to name the inputs and outputs. The article mentioned, among other things, addition and subtraction circuits (of two 5-bit binary numbers, one assumes.) Way cool. Just reading the article makes me want to play with one, and that despite more elaborate toys like the Sun Sparc10 I'm typing this on being available (heck: I could use it to make a logic compiler for this toy.) I wish I'd had this as a kid. Anything else out on the net about this thing? Sample "programs", more detailed descriptions, reminiscences and such? -- Markus Wandel Ottawa Ont. Canada (613) 592-1225 markus@pinetree.org <-- NOT 'mwandel@bnr.ca' (that's for work only) *** DISCLAIMER *** Not speaking for or representing my employer in any way. From: ard12@eng.cam.ac.uk (A.R. Duell) Newsgroups: alt.folklore.computers Subject: Re: Neat old digital logic toy Date: 9 Jan 1996 16:33:30 GMT mwandel@bnr.ca (Markus Wandel) writes: > >I was reading a newspaper article the other day, which described a >digital "computer" toy made by the German toy manufacturer Kosmos in >1968 or so. I've never seen this toy, lost the article, and forgotten >the name of the toy, but I found it fascinating. Hey, I had one of those years ago, and almost certainly still have it. No, it's not for sale. I got it long after 1968, though - I guess about 1975.. >It consists of 10 slide switches, 10 lamps, jumper wires, and a power >supply. The power supply was 3 C-cells, although I added a jack socket to mine to use a little mains adaptor. The batteries quickly became expensive... It came in kit form - you had to assemble the switch contacts and sliders before you could do anything. The fixed contacts of the switches formed terminals like those in solderless breadboards, and you pushed bare ends of wire in to 'program' it. >Each connection point "O" is brought out as several (it looked like 3) >holes into which wires can be plugged. Exactly right. It was 3 holes per terminal. >The picture showed that each switch and each lamp had room for a label >(transparent for the lamps) to name the inputs and outputs. On mine, you could fit pieces of translucent paper in the lamp cover to label the outputs, and feed a strip of paper along the top of the sliders to label them. There were cut-out sheets in the manual for the 'programs' it provided, and some blanks for your own hacks. >Way cool. Just reading the article makes me want to play with one, and It was fun... I hacked mine arround a bit - In the end I added some small relays from the local electronics shop to act as 'memory' and (I think) in the end I used a uniselector-type stepping switch to provide a sequencer. By that point the thing was mostly used for I/O :-) >that despite more elaborate toys like the Sun Sparc10 I'm typing this >on being available (heck: I could use it to make a logic compiler for >this toy.) I wish I'd had this as a kid. Another 'toy' I had at about that time was made by Philips. This was a digital logic system based on a set of identical modules. Each module had 3 inputs, A,B,C and an output F (and a lamp connected to this output). There were 8 'programming' sockets - 0, 1, C, C*as outputs and A*B*, AB*, A*B, AB as inputs. By patching those together you could make any 3-input combinatorial function, and by linking modules together, you could make sequential functions. I seem to recall I only had 2 such modules, along with the battery/input switch unit. That was great fun as well... -- -tony ard12@eng.cam.ac.uk The gates in my computer are AND,OR and NOT, not Bill