Bendix G-15: Alan Turing

Part 7 of a series: Bendix G-15 Restoration

There is nothing I can add to the volumes that have already been written about Alan Turing, so I’ll use this post to collect interesting links, especially with references to the Bendix G-15.

Wikipedia: Alan Turing

Alan Mathison Turing was an English mathematician, computer scientist, logician, cryptanalyst, philosopher, and theoretical biologist. Turing was highly influential in the development of theoretical computer science, providing a formalization of the concepts of algorithm and computation with the Turing machine, which can be considered a model of a general-purpose computer. He is widely considered to be the father of theoretical computer science and artificial intelligence.

I recently read “Alan Turing’s Electronic Brain” and it is a very good read for anyone interested in the Bendix G-15, as it details the evolution of the instruction set, drum memory, etc.

Amazon: Alan Turing’s Electronic Brain

The mathematical genius Alan Turing, now well known for his crucial wartime role in breaking the ENIGMA code, was the first to conceive of the fundamental principle of the modern computer-the idea of controlling a computing machine’s operations by means of a program of coded instructions, stored in the machine’s ‘memory’. In 1945 Turing drew up his revolutionary design for an electronic computing machine-his Automatic Computing Engine (‘ACE’). A pilot model of the ACE ran its first program in 1950 and the production version, the ‘DEUCE’, went on to become a cornerstone of the fledgling British computer industry. The first ‘personal’ computer was based on Turing’s ACE.

I recently watched “The Imitation Game” and highly recommend it.

Amazon: The Imitation Game

During the winter of 1952, British authorities entered the home of mathematician, cryptanalyst and war hero Alan Turing (Benedict Cumberbatch) to investigate a reported burglary. They instead ended up arresting Turing himself on charges of ‘gross indecency’, an accusation that would lead to his devastating conviction for the criminal offense of homosexuality – little did officials know, they were actually incriminating the pioneer of modern-day computing. Famously leading a motley group of scholars, linguists, chess champions and intelligence officers, he was credited with cracking the so-called unbreakable codes of Germany’s World War II Enigma machine. An intense and haunting portrayal of a brilliant, complicated man, The Imitation Game a genius who under nail-biting pressure helped to shorten the war and, in turn, save thousands of lives.

Next: TBD

Bendix G-15: Harry Huskey

Part 6 of a series: Bendix G-15 Restoration

The chief designer of the Bendix G-15 was computer pioneer Harry Huskey who based it on the Automatic Computing Engine (ACE) designed by Alan Turing at the National Physical Laboratories in Britain. Huskey had worked with Turing there for a year. Much of this post is taken from the Harry Huskey website. There is also some good information at Wikipedia: Harry Huskey and CHM: Harry Huskey. His obituary is here: Computer pioneer Harry Huskey dies at age 101.

Dr. Harry Huskey was born in the Smoky Mountains region of North Carolina and grew up in Idaho. He gained his master’s degree and then his PhD in 1943 from the Ohio State University on Contributions to the Problem of Geocze. Harry Huskey taught mathematics at the University of Pennsylvania and then worked part-time on the early ENIAC computer in 1945.

Harry Huskey designed and managed the construction of the Standards Western Automatic Computer (SWAC) at the National Bureau of Standards in Los Angeles (1949 to 1953). He also designed the G15 computer for Bendix Aviation Corporation, which could perhaps be considered as the first “personal” computer in the world. He had one at his home that is now in the Smithsonian Institution in Washington, D.C.

SWAC Computer
Harry Huskey with Bendix G-15 Donated to Smithsonian
Harry Huskey at the wheel of his creation!

I’ll add more information about Harry Huskey as I find it.

Next: Bendix G-15: Alan Turing

Bendix G-15: Hexadecimal, Sexadecimal, Shmexadecimal

Part 5 of a series: Bendix G-15 Restoration

One of the first things you discover when you start exploring the Bendix G-15 is that numbers are represented in base 16, but the hexadecimal (Greek root) A-F “digits” are not used for values 10-15. Instead, they use sexadecimal (Latin root) “digits”, where the values 10-15 are represented by the letters U-Z. Here’s more than you ever wanted to know about sexadecimal.

The Bendix Programmer’s Reference manual offers this (hilarious) definition:

“Hexadecimal” became a standard with IBM 360 Fortran

The now-current notation using the letters A to F establishes itself as the de facto standard beginning in 1966, in the wake of the publication of the Fortran IV manual for IBM System/360, which (unlike earlier variants of Fortran) recognizes a standard for entering hexadecimal constants. As noted above, alternative notations were used by NEC (1960) and The Pacific Data Systems 1020 (1964). The standard adopted by IBM seems to have become widely adopted by 1968, when Bruce Alan Martin in his letter to the editor of the CACM complains that with the ridiculous choice of letters A, B, C, D, E, F as hexadecimal number symbols adding to already troublesome problems of distinguishing octal (or hex) numbers from decimal numbers (or variable names), the time is overripe for reconsideration of our number symbols. This should have been done before poor choices gelled into a de facto standard!”

Bendix tried to make it easy

Next: Bendix G-15: Harry Huskey

DEC PDP-11/23+: New Old Chassis!

Part 19 of a series: DEC PDP-11/23+ Restoration

A few months ago, I located and acquired a proper BA11 chassis and front panel from a fellow in Kansas and was able to pick it up on my way from Chicago to Denver. The BA23 chassis I’ve been using works fine, but it’s not typical of a PDP-11/23+ system from that era, so this is a huge leap in authenticity.

The chassis required quite a bit of clean-up (lots of Kansas-style dust) and the fabrication of some missing parts, but now looks great. Some of the tasks completed:

– Cleaned corrosion from the front bezel and chassis
– Replaced missing card guides on backplane
– Fabricated and painted metal back brackets

BA11 Front Bezel (outside view)
BA11 Front Bezel (inside view)
New Metal Back Brackets

The power supply is dead and, instead of fixing it, I decided to replace it with a modern ATX power supply. I feel much better about putting a more reliable power supply into a museum setting. It’s a reversible modification, so no PDP-11s were harmed in the process and I can always go back. Some of the tasks completed:

– Removed original power supply
– Fabricated aluminum plate to match original dimensions
– Fabricated power cable with Molex connector to match original
– Installed terminal strips and replaced ATX connector with crimp connectors
– Replaced wiring from power supply to backplane

Original Power Supply
New Power Supply and Plate
New Power Supply (early prototype) in BA11

The PDP-11 power supply is responsible for the initial startup timing and signals for the QBUS and CPU, so I used a custom Arduino shield and some firmware to handle this. I’m hardly the first to do this, so there were plenty of examples to follow. Most of these examples are designed to replace the front panel, but I have the original front panel and needed something to replace the logic provided by the original power supply. The shield I designed is currently built on a proto board (ugly wires and bad soldering), but I’m thinking about spinning a custom PCB so it will look more professional. Some of the tasks completed:

– Built custom Arduino shield to interface switches, backplane, and power supply
– Wrote custom Arduino firmware to generate the correct timing and signals
– Mounted Arduino and shield behind the front bezel

Arduino
Prototype Shield (not pretty)
Enclosure

One of the best sources I’ve seen for information on emulating the QBUS power supply timing and signals is a blog post by Malcolm Macleod at QBUS Front Panel Project. His state-machine, C code, and clear explanations were invaluable.

ATX Power Supply Wiring
QBUS Power Up Sequence

It was a lot of work bringing up the new chassis, but it’s now installed in the rack and working well. I still have a few hardware projects to complete, but I can see the light at the end of the tunnel now.

The New Chassis

Next: TBD

Bendix G-15: Typewriter Console

Part 4 of a series: Bendix G-15 Restoration

I’m starting with what I’m calling the typewriter console, also known as the “Master Writer”, since it’s easily transported, fairly light and relatively robust compared to the fragile vacuum-tube CPU and magnetic tape mechanisms. I live in the mountains at the end of a long, bumpy dirt road and the thought of all those tubes and mechanisms bouncing around led me to find some spare office space down on the flatlands for most of the restoration, but I can work on the typewriter console at home.

Students using the Bendix G-15 computer

The typewriter console is an interesting part of the system in that it’s based on an early IBM electric typewriter. Electric typewriters were often modified for use as input / output devices for early mainframes and minicomputers. The most commonly modified typewriters were made by IBM. In the 1950s, modified standard versions of the IBM Electric A, B, and C models were used as console typewriters or terminals on many early computers (e.g., GE 225, JOHNNIAC, IBM 650IBM 1620PDP-1). The IBM Selectric typewriter, introduced in 1961, was easier to interface to a computer and was favored in new designs, such as the IBM 1130 computer and the IBM 1050 terminal.

The IBM Standard Electric Typewriter

IBM eventually sold a modified version of the type called the “Input-Output Writer”. It was resold under a number of computer manufacturer names:

Here’s the console at the pickup, complete with dust. Now the fun begins!

Back at the lab and ready to begin

Covers off!

It looks like I’m missing one of the toggle-switch caps on the typewriter base. I’ll need to find a matching replacement or fabricate one.

Next: Bendix G-15: Hexadecimal, Sexadecimal, Shmexadecimal

Bendix G-15: Garage Archaeology

Part 3 of a series: Bendix G-15 Restoration

I finally got a good look at the Bendix today. It’s been in my friend’s garage for many years and has a fair bit of dust on it but it’s in great condition overall. I live in the mountains at the end of a very bumpy dirt road, so we’re going to transport the large, fragile components (CPU, 2 tape drives) to some spare office space I’ve located down on the flatlands to avoid any damage. I hate to think of 350 irreplaceable vacuum tubes rattling around in the back of a truck. I’ll bring the console, which is relatively small and light, up to my house and work on it here.

Some of the equipment has been used as a home by unwanted creatures. A contact at one of the large computer museums recommended that the contaminated equipment be disinfected by a professional pest control service. That’s what we’ll do!

My friend and Bendix G-15 donor
Front view of the CPU
Side view of the CPU
The “Master Writer” typewriter console

Next: Bendix G-15: Typewriter Console

Bendix G-15: The Project

Part 1 of a series: Bendix G-15 Restoration 

T

I have a long-time friend who recently mentioned to me that he has an old computer in his garage that he’d like to find a good home for, preferably a museum.

“What type of computer is it?”, I asked.
“A Bendix G-15”, he answers.
“Oh cool”, I answered, having no clue what he was talking about.

I had to look it up and what I found is absolutely amazing.

The Bendix G-15 is a very early (1956) vacuum tube computer, based on Alan Turing’s Pilot ACE design, used primarily for military, scientific, and engineering. It was hailed as the first “personal computer” (meaning it only need one person to operate it) and was one of the very last vacuum tube computers produced. The base system, without peripherals, cost $49,500. A working model cost around $60,000 (over $500,000 by today’s standards).

It’s an odd beast, unlike any other computer I know of. It has a 29-bit word length, a whopping 2,176 words of magnetic drum memory, and an almost indecipherable instruction set. All operations are performed serially, 1-bit-at-a-time. It was gradually discontinued when Control Data Corporation took over the Bendix computer division in 1963.

Bendix celebrates the 100th G-15 computer

There were only about 400 produced and they are quite rare now (this model is #350). One is in the Smithsonian, several are in museums around the world, and one is in my friend’s garage. There are several functioning G15s that I know of, one of which is a restoration commissioned some years ago by Paul Allen at the Living Computer Museum in Seattle.

(source unknown)

The Bendix Corporation sold its computer division to Data General in 1963, so some of the last few G-15s produced (including this one) were badged as Control Data, but nothing is different underneath the hood.

The G-15 was well-known by many of the pioneers in our industry, including Donald Knuth, Ken Thompson, Paul Allen, Alan Perlis, Thomas Kurtz, David Evans, Ivan Sutherland, Elliot Organick, Mitch Kapor, David Ahl, Niklaus Wirth, John Backus, Vint Cerf, Seymour Cray, Max Palevsky, and Jeff Duntemann. For many of them, it was the first computer they used.

Students using the Bendix G-15 computer (source unknown)

My friend’s G15 is in nearly mint condition, appearance-wise, but hasn’t been powered-up since 1979. Apparently, most of the 450 vacuum tubes have died an early death and most of the 3,000 germanium diodes have degraded to a direct short. Restoring it to full functionality would be extremely expensive, impractical, and I’m not even sure the fire marshal would allow it indoors. Some early users even relied on it to heat their offices. It would, however, make a magnificent museum display. It’s not as large as it looks in the photos. There are 3 units measuring 3′ x 3′ x 5′ and a teletype-style terminal. It is heavy, totaling over 1,500 pounds, and requires great care when transporting it.

To make a long story short, I have found a museum that wants the Bendix G-15 and I’ll be starting the long restoration (mostly clean-up) process soon! The goal is to clean it up, evaluate its condition, transport it to the museum, and put together a display.

Here’s a list of museums, no doubt incomplete, that have G-15s:

Vintage Computer Federation Museum (Wall Township, NJ, USA)
System Source Computer Museum (Hunt Valley, MD, USA)
Computer History Museum (Mountain View, CA, USA)
Living Computer Museum (Seattle, WA, USA)
Smithsonian National Museum of American History (Washington, DC, USA)
Ridai Museum of Modern Science (Tokyo, Japan)
Australian Computer Museum (Western Australia)

I have often forgotten to include before and after photos of my restorations, so I’ll make an extra effort to do so for this one, especially since it’s largely a cleanup operation.

Bonus! Here’s a fun video in which the Bendix G-15 saves the day!

Science Fiction Theatre: Dr. Robot (1956, Season 2, Episode 29)

Science Fiction Theatre

The G15 appears in these time frames:

02:05 – 02:11
11:23 – 12:10
20:52 – 21:40

Another bonus: The Bendix G-15 plays a Christmas carol!

I’ve Got a Secret – Game Show

I’ve Got a Secret Game Show

This is complex project, even if it is just cleaning and evaluation, so I’ve reached out to a few vintage computer groups:

VCF Minis and Mainframes Forum

Facebook Bendix G-15 Forum

Classic Computer Forum

Next: Bendix G-15: A Brief History

DEC PDP-11/23+: Back to Work!

Part 18 of a series: DEC PDP-11/23+ Restoration

Progress has been very slow but I’m back at it now, trying to finish up this restoration as soon as I can. A number of things have happened since the last post:

Took delivery of two RL02 disk drives and a controller. They required a bit of cleanup and, so far, it looks like they’re not in working condition. I’ll go ahead and install them in the rack and try to get them working later.

Acquired a proper BA11-N chassis and front panel. The chassis I’ve been using (BA23) works fine, but it’s not typical of a PDP-11/23+ system from that era. This will require quite a bit of clean-up and I’m not sure of the status of the power supply, but it’ll look great. I’ve started a thread in one of the VCFed forums to solicit help on how to deal with some of the corrosion on the chassis: Classic Repair #1: White Rust on Metal Chassis.

  • Acquired some system documentation and software. The software is on 8″ floppy disks, so I may try to get an RX01 or RX02 floppy drive working. If that doesn’t work out, I’ll copy some of the software (using Kermit at 9600 baud) onto one of the RL02 drives. I’m still on the lookout for a complete set of RT-11 or RSX-11M+ documentation.
  • The QBone board I’ve been waiting for has finally shipped and I’ll be ordering one soon. This board fits in the QBUS backplane and can emulate a variety of devices, including RAM, serial I/O, disk drives, etc. For the museum installation, it will provide disk services without putting wear and tear on fragile vintage disk drives and media. Looking forward to checking this out: QBone – The premiere PDP-11 add-in card.
  • Tear down and clean-up of a VT100 terminal. The video drive board seems to be wacky, and the keyboard is missing, so it’ll take some work to get it running. Any VT100 keyboards out there? These things are ridiculously difficult to find!
  • Tear down and clean up of an LA120 printer. The status of this machine is unknown, so once I get it put back together I’ll see what remains to be done. Should be fun hearing the sound of a dot-matrix printer again!

Next: DEC PDP-11/23+: New Old Chassis!

DEC PDP-11/23+: Disk Drives Found!

Part 17 of a series: DEC PDP-11/23+ Restoration

Progress has been slow lately, as I’ve been trying to locate a pair of RL02 drives to complete the restoration. I acquired two RL02 drives with the initial eBay purchase and documented their restoration in earlier posts, but they were earmarked for another project so I’ve been scouring eBay and other forums since then to find someone to sell or donate another pair. These drives are very difficult to find and, due to their weight and fragility, can be very expensive to ship.

The situation was looking pretty grim until a few weeks ago, when I was contacted by Kevin Kolasa, a vintage computer enthusiast from Edmonton, Alberta in Canada. He recently rescued some DEC equipment from the merciless jaws of a recycler and very generously offered to part with two of his RL02 drives as well as some other DEC equipment. The catch, of course, is that the gear is about 1,300 miles away. He’s currently loading the bounty onto a shipping pallet and we’re about to begin an adventure in transporting vintage computer equipment from Canada to the United States in winter during the Coronavirus pandemic. Should be interesting! I’ll keep you posted as we make progress.

This adventure is a bit too far for a road trip, and we can even go to Canada due to the pandemic, so we’re looking at LTL (less-than-truckload) freight shipping.

The RL02 drives appear to be in great condition and the donation includes some disk cartridges, cables, controllers, etc. as well as a TS05 tape drive in a DEC “corporate cabinet”. Kevin is building a bomb-proof pallet! I’m guessing it’s about 500 lbs., but we’ll know for sure when the forklift puts it on the truck.

Update 2021.05.01: Getting closer!



Update 2/2/2021: The drives are on their way!

Woo hoo! The drives were just loaded onto a truck for transport to the Edmonton freight terminal, where they’ll be loaded onto another truck bound for Denver! Thanks to Kevin for all his hard work – it’s been epic.

I’ve learned a little bit about shipping from this experience; it’s complicated and there seems to be a fair bit of voodoo involved in calculating rates. The best rate I could find was for shipping from the Edmonton terminal to the Denver terminal. Door-to-door was several times more expensive. Kevin is arranging to transport the crate from his house to the terminal (a 12-minute drive) through an independent carrier and I’ll arrange to pick up the crate from the freight terminal in Denver with a U-Haul trailer and bring it back to Boulder. Another option is to disassemble the crate at the terminal in Denver and use two cars to bring it back to Boulder. Lots of complications, but it’s still cheaper than door-to-door.

Update 2/4/2021: Won’t be long now!

Update 2/8/2021: Arrived in Denver!

Update 2/9/2021: Drives made it!

Now it’s time to clean up the gear and test it out.

Next: DEC PDP-11/23+: Back to Work!