View unanswered posts | View active topics It is currently Wed Jul 17, 2019 10:37 pm



Reply to topic  [ 12 posts ] 
 WITCH-E - recreation of Harwell Dekatron 
Author Message

Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1202
David Anders is running the WITCH-E project, to recreate the 1951 Harwell Dekatron machine, also known as the WITCH, described as the oldest running computer and on show at TNMoC in Bletchley Park. The Witch is a decimal machine, using relays and a gas-discharge circuit element called a dekatron - WITCH-E recreates a dekatron on modular boards. See the wiki, the repo and a couple of videos:

http://witch-e.org
https://github.com/prpplague/witch_dev

dekatron tests
Image

WITCH-E Project Introduction
Image

The WITCH bootstrap is encoded with relays. Here's a test of the recreation of that circuit:
WITCH-E Project - Relay Set J
Image


Mon Mar 28, 2016 2:29 pm
Profile

Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1202
Over at 6502.org, BDD wondered about the sound of a WITCH. So here are some videos of the WITCH in action - I think it's demonstrated most days, at The National Museum of Computing on Bletchley Park.

Two-tonne Witch computer gets a reboot [2012] [BBC]
My video from a July 2014 visit
My video from a September 2015 visit


Mon May 23, 2016 7:31 pm
Profile

Joined: Wed Apr 24, 2013 9:40 pm
Posts: 160
Location: Huntsville, AL
Ed:

That was a great demonstration. Thanks for taking the time to dig up the links and post them here. I very much enjoyed the short demonstration; there's nothing like a relay-based computer. There were a number of other decimal-based machines, but I certainly enjoyed seeing the dekatron's in action in this machine.

_________________
Michael A.


Mon May 23, 2016 10:56 pm
Profile

Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1202
A good way to follow Dave's progress on dekatron computing is his hackaday.io presence, for example:

(I say it's a good way - I think you have to keep visiting to check on updates. I'd prefer to have an RSS feed. Maybe stalking Dave on YouTube is also a good way. I do that too!)


Sun Sep 30, 2018 11:14 am
Profile

Joined: Thu Apr 04, 2019 6:20 pm
Posts: 22
BigEd wrote:
A good way to follow Dave's progress on dekatron computing is his hackaday.io presence, for example:

(I say it's a good way - I think you have to keep visiting to check on updates. I'd prefer to have an RSS feed. Maybe stalking Dave on YouTube is also a good way. I do that too!)


just fyi, i haven't really done anything with this project since the fall of 2017. there was very little interest in the overall project of documentation and replica, even from The National Museum of Computing (TNMOC). the design is fairly complete with only the paper tape reader still needing to be implemented...


Thu Apr 04, 2019 6:23 pm
Profile

Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1202
Good to see you here, Dave! I'm sure you'd get some interest for your projects here, although not necessarily enough to justify a production run. A paper tape reader is a nice idea: physically realised tangible storage.


Thu Apr 04, 2019 8:37 pm
Profile

Joined: Thu Apr 04, 2019 6:20 pm
Posts: 22
BigEd wrote:
Good to see you here, Dave! I'm sure you'd get some interest for your projects here, although not necessarily enough to justify a production run. A paper tape reader is a nice idea: physically realised tangible storage.


Thanks!

just fyi for anyone interested, i've collected probably the biggest set of public available documents and pictures outside of TNMOC about the WITCH. my github account includes a wide range of pictures of various parts of the WITCH, high resolution scans of the schematics, redrawn functional diagrams, programming manual, example programs, and transcribed copies of original engineering documents. everything can be found here: https://github.com/prpplague/witch_dev

in addition there is a javascript emulator: https://github.com/prpplague/witch-e

feel free to contact me for questions!


Tue Apr 09, 2019 11:48 pm
Profile

Joined: Thu Apr 04, 2019 6:20 pm
Posts: 22
BigEd wrote:
I'm sure you'd get some interest for your projects here, although not necessarily enough to justify a production run.

i had considered scaling the design back considerably to just 2 registers that are 4 digits plus sign wide, and the accumulator at 8 digits. this would make it small enough to do some basic examples of computer based decimal math, and small enough to be class room accessible. i pitched this idea to the TNMOC and got a lot of negative feedback...


Wed Apr 10, 2019 12:04 am
Profile

Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1202
It's a tricky one, to get buy-in, as people always have their own preconceptions of what would be ideal. I was good at maths, as a child, and so the arithmetic approach to computing had an easy appeal to me. But I like it when educators are able to find non-numeric ways to demonstrate and teach computing. Ideally we need more people who skilled in both education and technology.

IIRC, WITCH can execute programs from (very limited) register space, or from tape. I guess a mini-machine can only execute from tape? I think a physically looped tape might even be sufficient for any program, but it's not quite as clear as a conventional machine with a PC and branch instructions. So, would the mini-machine be more comparable to a calculator with keystroke memory?

It still must be worthwhile to cook up some example programming challenges, to be able to illustrate how the machine could be used in teaching.


Wed Apr 10, 2019 10:54 am
Profile

Joined: Thu Apr 04, 2019 6:20 pm
Posts: 22
BigEd wrote:
It's a tricky one, to get buy-in, as people always have their own preconceptions of what would be ideal. I was good at maths, as a child, and so the arithmetic approach to computing had an easy appeal to me. But I like it when educators are able to find non-numeric ways to demonstrate and teach computing. Ideally we need more people who skilled in both education and technology.

IIRC, WITCH can execute programs from (very limited) register space, or from tape. I guess a mini-machine can only execute from tape? I think a physically looped tape might even be sufficient for any program, but it's not quite as clear as a conventional machine with a PC and branch instructions. So, would the mini-machine be more comparable to a calculator with keystroke memory?

It still must be worthwhile to cook up some example programming challenges, to be able to illustrate how the machine could be used in teaching.


yea, the execute from memory would be difficult to do on the mini-machine version. i've struggled with coming up with a version of the design that could be easily used and manufactured. i certainly could making it a lot more compact by consolidating the design several ways. one way is to ditch the 5-bit baudot format used for the data storage, and replace it with standard 4-bit BCD. the other is dropping the modular dekatron simulator, and consolidating them to smaller PCBs that only display the stored value.

it could certainly be done... the real question is: does it have any educational value?
to which i don't have an answer...


Fri Apr 26, 2019 12:10 am
Profile

Joined: Wed Jan 09, 2013 6:54 pm
Posts: 1202
Another tricky one - what might be great for educational value might not be so good for historical value, and vice-versa. There could be an argument that something decimal-based is more accessible, it removes a possible barrier to thinking about computing at a low level.

One of the thoughts I keep returning to is that computation is really about symbols, not about numbers. But many of us in computing are highly numerate and don't find numbers difficult, we even find them attractive, so we don't see this so naturally. Having ideas for teaching - projects, lessons - which are not based on numeracy could be an advantage in appealing to more of the audience. Even if numbers come into it, they are not the foremost consideration, but a means to an end - I'm thinking of games, of turtles, of robots, of graphics.

But I'm not sure how this helps with the idea of building up a simple machine with electronics, or the idea of revisiting a specific historical architecture with an educational aim.


Fri Apr 26, 2019 5:20 am
Profile

Joined: Thu Apr 04, 2019 6:20 pm
Posts: 22
BigEd wrote:
Another tricky one - what might be great for educational value might not be so good for historical value, and vice-versa. There could be an argument that something decimal-based is more accessible, it removes a possible barrier to thinking about computing at a low level.

One of the thoughts I keep returning to is that computation is really about symbols, not about numbers. But many of us in computing are highly numerate and don't find numbers difficult, we even find them attractive, so we don't see this so naturally. Having ideas for teaching - projects, lessons - which are not based on numeracy could be an advantage in appealing to more of the audience. Even if numbers come into it, they are not the foremost consideration, but a means to an end - I'm thinking of games, of turtles, of robots, of graphics.

But I'm not sure how this helps with the idea of building up a simple machine with electronics, or the idea of revisiting a specific historical architecture with an educational aim.


yea i'd love to do a demo of the current design to some education professionals to get some feed and ideas, but that has never materialized. i did a demo at TNMOC, but only got some "oh that's cool" responses. no real feedback on moving forward with it, which is why put the project on hiatus...


Fri Apr 26, 2019 5:53 pm
Profile
Display posts from previous:  Sort by  
Reply to topic   [ 12 posts ] 

Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group
Designed by ST Software