The Mini-Mini Computer

Thanks to Doug Coward for scanning and making these available.

Doug Coward wrote:

About 8 years ago I was in my favorite Bay Area electronic surplus store - Mike Quinn Electronics. At the time, they were still in the Quonset hut at Oakland Airport. I bought a pile of books and manuals. And when I got home I discovered among the books 3 handwritten and xeroxed, double-sided sheets stapled together entitled "MINI-MINI COMPUTER". The six pages described the design of a computer built using four core planes from an IBM 1401. At the time it was written, these core planes were available from Mike Quinn's for less than $9 each.

The pages had no date on them, just the author's name and address. In 1999 I decided to track down this Hal A. Chamberlin Jr.

Hal (I learned) is the author of the well regarded book "Musical Applications of Microprocessors".

Hal Chamberlain wrote:

The "MINI-MINI COMPUTER" is a paper design I did shortly after graduating in response to requests from hobbyists around the country about how to make a core memory unit function. It was a slimmed down version of the computer I had successfully built and run a year earlier called the HAL-4096. MINI-MINI was described rather completely on 2-4 (don't remember exactly) hand-printed letter-size pages which I reproduced via blueprint machine. I filled about a dozen requests for copies but never heard of anybody actually building one. I have at least one set of the original blueprints in my archives in America.

MINI-MINI was strictly digital with 4000 (not 4K) 4-bit words of core memory. Each word could hold one octal digit plus a "flag" bit which was used to indicate the boundaries between words and whether a word was negative or positive. Thus it was a variable word length machine like an IBM 1620. Instructions were 10 digits long and consisted of 2-digit opcode and two 4-digit addresses. There was no accumulator; everything was memory-to-memory so one instruction could do a lot.

Although MINI-MINI was probably never built (at least not by me), HAL-4096 WAS built and ran from around 1970 until around 1979 when I moved from New Hampshire back to Raleigh (it was retired then and didn't make the move because wires in the core memory were corroding and breaking). I have some photos, a newspaper clipping, some of the homemade boards it used, perhaps a memory plane sample, and some software listings. It was a pretty complete system with full console, Selectric typewriter (and later a line printer), card reader, card punch, paper tape reader and punch, and modem (300 baud). Software included a full-featured assembler and BASIC interpreter, a version of which was remotely accessible via auto-answer modem. There were tons of utilities and some experimental music synthesis software (the real reason it was built). I've still got much of the software on paper tapes, card decks, and printed listings.

I graduated with a BS in the Spring of 1970 so the design of the Mini-Minithat would have been the Summer of that year. I believe that Mike Quinn had suggested that I write some plans for using IBM 1401 core memory planes which he had in abundance at the time and that was the result. It was really based on the then running HAL 4096 which used a larger IBM 1620 memory unit.

The HAL 4096 was begun in 1968 (beginning of my Junior year) when I got the core memory unit from Mike Quinn and some logic boards from IBM where I had worked the preceeding Summer. An early version of the core memory and ALU doing "something useful" (don't remember what it was) was shown at the 1968 NC State University Engineer's Fair. The CPU didn't execute the full instruction set and have any useful software until 1969-1970.

The original scans by Doug are greysacle jpg each is between 2 and 3 Mbytes :

I cleaned them up a little and saved them as black and white png images which are between 200 and 300 Kbytes each:

Hans B Pufal
ACONIT
7 Jan 2002