Burroughs L7000

[billdeg]

This week I was contacted to pick up an old 60's mainframe I was told...turned out to be a Burroughs L7000 mini computer. I am going to help deliver it to the MARCH (midatlanticretro.org) museum this fall.

http://www.vintagecomputer.net/burroughs/L7000/

[MikeS]

Originally Posted by billdeg

This week I was contacted to pick up an old 60's mainframe I was told...turned out to be a Burroughs L7000 mini computer. I am going to help deliver it to the MARCH (midatlanticretro.org) museum this fall.

http://www.vintagecomputer.net/burroughs/L7000/

---
Hardly a mainframe, and minicomputer is stretching it a little...
Unless it comes with a complete set of tapes the chances of getting it up and running are pretty slim. There should be a set of batteries and a DC300 tape drive and tape inside the right end cover; if that tape can be read then you might be able to bring it back to life.

If you get it up and it has at least one cassette drive, I may have some software for it.

Good luck!

[EvanK]

Originally Posted by MikeS

minicomputer is stretching it a little...

It's a business machine from (at least) 1972; there are some details on pages 33-34 of the document here (pages 47-48 of the total pdf): http://www.bitsavers.org/pdf/auerbac...uters_1972.pdf .... so if it's not a minicomputer then what would you call it?

We also have a later (1982) Burroughs B-80.

[MikeS]

Originally Posted by EvanK

It's a business machine from (at least) 1972; there are some details on pages 33-34 of the document here (pages 47-48 of the total pdf): http://www.bitsavers.org/pdf/auerbac...uters_1972.pdf .... so if it's not a minicomputer then what would you call it?

Good question
Burrpoghs (sic ) called them 'Commercial' mini-computers; they're really a category of their own with the integrated keyboard and dual printer and various (or no) peripherals.
Programmable fancy typewriter? Electronic Accounting Machine?
I didn't say it wasn't a mini-computer, just that it stretches the definition a little; definitely single-user, 8KW max, ledger cards, no video display, no disk I/O, only programmable in assembler (although there were cross-compilers that ran on the 'real' mini-computers), comm available but rarely installed, etc.

We also have a later (1982) Burroughs B-80.

That's a little closer to a mini; disk drives, video display, programmable in BASIC, COBOL etc.

[Chuck(G)]

Olivetti had something like this too. "Accounting machine" might be a better term for it:

http://www.technikum29.de/en/computer/commercial.shtm

[billdeg]

By itself, the workstation is not a mini or a general purpose computer. It has a card reader built in. The separate "Magnetic System" unit is I think where the storage is happening. I am not sure if multiple workstations plugged into a single L7000 or not. The L7000 looks like a *huge* A/C unit, with logic cards kind of like the PDP 8 we have at MARCH. Ultimately this is a business computer like later 70's Raytheon VT302 and Lanier Model-103. It was marketed for accountants.

Once I get back to the system and take some good pictures I will get a better "picture" of what I am dealing with. I did not find any kind of user manual, just a service record checklist doc.



Bill

[MikeS]

Originally Posted by Chuck(G)

Olivetti had something like this too. "Accounting machine" might be a better term for it:

http://www.technikum29.de/en/computer/commercial.shtm

------
Well, although mini-computer is kinda stretching it, "Accounting machine" is selling them short; although they are certainly descended from the electro-mechanical and hybrid adding machines on steroids that are usually called accounting machines, they are certainly true computers in their own right, although obviously configured differently from what we think of today as "computers".

They're an interesting almost completely forgotten and ignored branch of computing with very little info or even interest out there, illustrated by the fact that even knowledegeable vintage hackers like Bill and Evan don't really know what they actually are.

Although Olivetti, NCR, Monroe and others were also major players, Burroughs was for a long time the leader in accounting machines, found in every bank branch and the offices of most medium sized businesses. While IBM was focussed on punched card batch processing machines Burroughs et al were providing equipment to do the same sort of thing in real time with direct operator input.

Originally the machines were completely mechanical except for an electric motor to drive and turn the hundreds of wheels, cams and levers that did the calculation and printing on ledger cards and journal rolls; this was the Burrough 'F' series, programmed with metal pins of different lengths and locations in exchangeable "program panels."

As solid state technology became available some of those wheels, cams and levers were replaced with transistors and core memory, and then integrated circuits; this was the Burroughs 'E' series, programmed with metal pins and wired patch board programming panels.

Then Burroughs brought out the 'L' series, programmed in 'SL3' and 'SL5' assembler (System Languages), which ranged from the original L2000 to the L9000 before the ledger cards were finally abandoned and the B80 replaced them with disk-based systems programmed in high-level languages, mostly Cobol.

About the only thing that the L series had in common other than appearance was the fact that they all had an integrated paper tape reader beside the keyboard for loading firmware and application and utility programs. Up to the L5000 the system memory was actually a small fixed disk while from the L6000 onward it consisted of 2KB memory cards; also the keyboard and PPT reader were mechanical through the 5000 and electronic after that. Although not all models did, they were mainly intended to use ledger cards and had split platen printers with a separate ledger card feeder so that the printer was actually three separate printers in a way, using a Selectric-style golf ball until the L9000 finally went to a dot-matrix printer (and sacrificed the red/black dual color capability). The ledger cards could have magnetic stripes on the back which stored account names etc. and account balances and allowed for automatic insertion and alignment (although the mechanical machines could also align the cards by punching little notches).

A full-blown L could have magnetic stripe ledger I/O on the console with or without an optional auto-feeder/stacker, and a free-standing separate auto-reader that could process a stack (probably what Bill is thinking of), up to four digital cassette drives, a paper tape and edge-punched card reader and punch, an 80 or 96 column card reader and punch, and even datacomm and a video display on the latest models. No workstations plugged into an L, it *was* the workstation, although some models could connect to a 'real' mini or mainframe.
I have a picture here of an L with all the bells and whistles; I'll try to find time to scan it.

Unfortunately there is almost no firmware or software around except for what is sometimes found in the back of the machine with the print set, or the unfortunately rare DC300 cartridge that can still be read. Ironically the older hard-disk based units are more likely to be operational while the solid-state based units are of course huge bricks without the firmware to boot them up.

There are a few others out there that I know of, although AFAIK the only one that actually works (or did at one time) is a restored L5000 at Bletchley Park.

The odds are against it, but I'll certainly cross my fingers for ya that you get it going when you get it. The most crucial part is that DC300 if there is one; if the band hasn't rotted, the oxide crumbled or stuck together etc. and if the L actually did completely and successfully dump a memory image to it the last time it was shut down, then you just might be in business. If there actually is one, then remove the cart and inspect/repair/clean it and the drive as much as possible before you turn on the machine; on power-up it will automatically try to read that cartridge and could quite possibly destroy it if it's in bad shape.

Good luck!

[Chuck(G)]

The NCR 446 "accounting machine" shown is programmable via paper tape (program resides in core) and the Nixdorf 820 is certainly a full-blown computer.

And IBM called the plugboard-programmed 407 an "Accounting Machine", even though it probably has far less power than the Olivetti P203.

I got to witness the crossover and back between punched-card unit-record accounting systems and the real thing back in the 60's. For most of the 60's, this Chicago-area manufacturer had been doing just fine with the standard unit-record setup. (407+sorter+reproducing punch+interpreter). The CEO somehow crossed paths with an IBM Sales Engineer who told him about the wonders of computing. The CEO signed up for a 360/20 with accompanying peripherals, hired a programmer and an operator and waited for the savings to roll in. Every day, the CEO received a nice thick printout of production figures, account balances, sales, etc. that he never read.

Eventually said CEO slipped into an IBM conference and posed as representing an outfit with no information automation whatsoever. After analyzing his needs, the sales engineer recommended--you-guessed it--a standard unit-record setup.

The 360/20 and extra personnel were gone at the end of the month. Corporate America had reason to be skeptical of computerization back then. An "accounting machine" was probably perfect for many businesses.

Did you note that the entire type basket in the Olivetti was in a cassette that could be changed out? Pretty cool.

[billdeg]

Hopefully next week we can move this system to the MARCH computer museum in Wall, NJ.

I took some new pictures. The new pics start with P100*.jpg

http://www.vintagecomputer.net/burroughs/l7000/

[Ksarul]

Definitely a nice find there, Bill!