CNC in 1952?

[kvom01]

In the opening chapter of his 1952 novel "Player Piano", Kurt Vonnegut describes a factory where 50 lathes are all simultaneously producing motor shafts. the lathes are controlled by a magnet tape loop which has a record of the motions of a single machinist making a shaft. I was wondering, when reading this recently, whether Vonnegut dreamed this up himself or whether there were similar types of automated machining being done at that time.

He describes the work being dropped between the chuck and tailstock, so evidently bar pullers might not have been around then.

 

[chucker]

it is possible that he seen this for him self check the link below and it says "The first commercial NC machines were built in the 1950's, and ran from punched tape"

CNC Machine and Computer Numerical Control History

 

[Dcrouch]

Vonnegut worked for GE until 1951 or so. He may have see them at one of the GE plants.

 

[Heinz R. Putz]

I worked for Swedish Machine Tool starting in 1971, we produced and siold the forerunner of CNC lathes, it was called an S10NC, had a 10" chuck and you actually made a part manually and recorded all motions, speeds, feeds ,etc, by punching holes into a card.
To CNC: In 1973 we introduced our first CNC lathe control at the Cincinnatti Machine Tool Show.
This was, to the best of my knowledge, prior to Fanuc or anybody else.
At the 1974 Chicago Show, we actually produced and gave away 10 pound dumbbells, it seemed every japanese visitor took pictures of our machine and carried a dumbbell.
Those were the days, we had a great market until Mori Seiki started showing up with the SL3 and SL4 for about half our price and put us out of the market.
I quickly learned I better know Fanuc really well and I have been teaching it ever since.
Heinz, doccnc.com

 

[Milacron]

In the opening chapter of his 1952 novel "Player Piano", Kurt Vonnegut describes a factory where 50 lathes are all simultaneously producing motor shafts. the lathes are controlled by a magnet tape loop which has a record of the motions of a single machinist making a shaft. I was wondering, when reading this recently, whether Vonnegut dreamed this up himself or whether there were similar types of automated machining being done at that time.

You are describing NC, not CNC of course, but the first electronic NC machine was made by John T. Parsons in 1948....apparently a small mill to produce 2D templates from data on punched cards. The Air Force was impressed enough with this it led to a contract in 1949 which ultimately led to the development work done at the Servomechanism Lab at MIT to create a monstrous Cincinnati Hydro-Tel vertical mill for complex 3D milling via punched card data fed into a bank of numerical control and servo drive towers that filled a room. Assembled in 1951, but not until 1953 was MIT satisfied with the results enough to publish.

In the meantime a company known as Arma had produced a little punch card driven NC lathe that was given public demonstration in 1950 but when the Korean War broke out they suspended work on it ! Far as I can tell the first NC machine using magnetic tape was the Numericord system by Giddings & Lewis, offered in 1955.

-----------------------------------------------

Speaking of punched cards, I can remember even in the mid 70's there were what seemed like never ending job advertisements for "keypunch operators" in Raleigh, NC. I presume these were used for storing mainframe data for various businesses but I wonder when those fell out of favor for that application ? I recall a girlfriend getting a job doing that and her reporting back it was so mind numbing she had to quit after a couple of weeks.

 

[Limy Sami]

On the subject of punch tapes - were they ever made with steel tapes? ...... must be 1970 - 72 I saw an old machine in a works office,....... ''something - Vickers'' and was told it was ''control tape puncher for the auto shop'' or has my memory finally gone.

 

[Milacron]

On the subject of punch tapes - were they ever made with steel tapes? ...... must be 1970 - 72 I saw an old machine in a works office,....... ''something - Vickers'' and was told it was ''control tape puncher for the auto shop'' or has my memory finally gone.

I recall tapes made of some sort of silver vinyl material that might look like steel at casual glance...maybe that's what you saw.

 

[dlankhaar]

I lived in Seattle in 1971 and wrote and debugged assembler code on punch cards. After it was proven we transfer the punch cards to IBM tape drives.

Are choices of beer were Budweiser or Rainier...

 

[mister honey]

I recall tapes made of some sort of silver vinyl material that might look like steel at casual glance...maybe that's what you saw.

That was mylar NC punch tape, as I recall.

Mike

 

[Milacron]

That was mylar NC punch tape, as I recall.

Yeah, now that you mention it, mylar sounds more precise. Suppose same stuff as mylar recording tape but thicker, wider and not magnetic.

 

[Bruce Nelson]

In about 1968 there was a enterprising fellow in western Wisconsin that fashioned a method of recording signals ordinary on magnetic tape on a tape recorder to drive the x and y locations of a automatic drill for circuit boards. As I recall, he tried to sell the idea to 3M, but alas, technology had already passed him by. He did, however actually have contracts for drilling circuit boards. By the way, punched paper tape was used for everyday purposes for N/C work. Mylar was used for long runs.

Lord Byron

 

[bryan_machine]

When I went to college at the university of Iowa in 1978 we programmed with punched cards, terminals phased in over the next few years, it was considered behind the curve. When I did census pre-list work later, they still talked about a "keypunch operator" who would be entering the data into a computer.

Data entry (people typing stuff in all day) was likely called "keypunch" for a long time. It really only disappeared with the rise of the web - where the trick is to get the end user, or their computer/phone/camera, to just enter the data directly. It would indeed have been mind numbing. Postal processing and paper check processing probably still both have lots of people doing similar tasks.

Computers in general used both paper and mylar tape until the late '70s (at least) as well.

 

[Screwmachine]

Here's a great thread from a while back with ancient -but still running- (C)NCs. Some nightmarish looking control cabinets crammed full of boards and wires.

Oldest running CNC

 

[706jim]

The tape was available in various formats. Plain paper was cheapest and some paper tapes were embedded with oil.

Stinky.

We used mylar tape, but there was also a mylar aluminum foil laminate that I suppose some people preferred.

I recall that mylar cost about 15 times as much as paper.

 

[kvom01]

Some non-machining recollections from the punched card era.

I worked my way through university as a programmer 1967-71, with all programs on punched cards. The keypunch machines were quite well-designed machines that fed blank cards and stacked finished ones. They had a "format" card on a cylindrical holder that allowed tabbing to fixed columns. The 027 machines had round metal keys and were a bit awkward to type on, and the newer 029 has square plastic keys like a terminal. After punching the "deck" for a program, you'd use a Magic Marker to draw a diagonal line across the top of the cards. This allowed you to reassemble the program more easily in order in case you dropped the deck, and showed out of order cards.

I was used to punching up my own programs, but on my first job past college I was forbidden to do so, as I was too "expensive" for manual key entry. So I had to write the program on coding sheets and send them down to the keypunch operators. Afterwards I could use the keypunch for corrections or changes.

In my second job there was no keypunch available, so again the coding sheets were sent out to a service. Since these programs were fairly large they were also "verified". Verification meant that the punched cards were fed through the keypunch a second time and rekeyed. The format card told the keypunch to reject cards where the keystroke didn't match the punched holes. Obviously verifying cost twice as much. On that job I had to take my program to NYC to use on the only computer in the US that could compile the code. I had to work the midnight shift to use the machine, which had a paper tape reader and no card reader. So the first task was to convert the cards to tape. A smaller computer did this. As it read the cards the punched paper spewed out rapidly forming a tangled pile on the floor. Afterwards I had to re-spool the tape using a manual winder, taking care not to tear it. Then the tape was fed into the main computer, the reader making another pile on the floor. The output was a cassette tape that could be used for testing on the target machine. Since the cassettes were error prone I generally ran the tape through 3 times to ensure getting a good cassette. This was in 1975-76.

The other use we had for paper tape was the teletype. Before internet/email, this was the "best" way to send a typed message. However, the machine was very awkward to use, and once a letter was keyed and the tape punched you couldn't backspace and correct it. So teletype messages usually had a lot of typos. Long distance was expensive, so we used the machine offline to punch the tape, and then fed it back once the call was dialed.

 

[adama]

I think part of the mylar thing came about because later tape readers used light transmission to see the punches. Oily paper and a bad day and enough light would get through to trick the reader and create hell.

 

[DDoug]

I think part of the mylar thing came about because later tape readers used light transmission to see the punches. Oily paper and a bad day and enough light would get through to trick the reader and create hell.

Mylar is still being used, it's for machines with no memory, they back up the tape every part, and re-run it.

I have seen a monarch lathe that still run's production to this day that way.

 

[DDoug]

I have a book in my library "oxygen cutting" put out by airco, 1951.

In it they describe a 2 axis machine torch, not magnetic spindle tracer, not electric eye.

It has a drum with "tick marks" much like standard paper tape, an eye reads the
"ticks" and drives the machine- 2 axis bridge machine like common nowadays....1945.

There are pix of the actual machine working, so at least one got built.

Bibliography leads to "steel may 28,1945", but google don't seem to turn anything up.

 

[Bruce Nelson]

I thought that the chad produced from those tape punches could be marketed as a additive to breakfast cereal to provide bulk and combat diarrhea.

Lord Byron

 

[Clive603]

On the subject of punch tapes - were they ever made with steel tapes?

Dunno about steel punched tape but the Deihl Combitron desktop "computer" I had to learn for my first job had its set-up and boot up data on a steel magnetic tape. No static memory in 1971 so it had to run that every time it was turned on. Nearly 5 minutes. Thing was about the size of a big electric typewriter with a whopping great paper tape puncher / reader strapped on the side. I started in mid October 1971 and I think it arrived a couple or three weeks before me so I guess they needed a convenient victim. Not knowing anything about computers didn't seem to be a handicap. "Oh you will soon pick it up." Which quote pretty much sums up my working life! Whopping 100 steps of program memory and the ability to do things section by section via the punched paper tape. Only one reader for both program and data tho'. Which got interesting until I twigged that the 35 mm film splicing deck would do the deed on paper tape too. That part of the job was mostly accuracy statistics for Swingfire ATGW and other missile systems. Deihl programming manual in one hand, Schaums Outline Series Statistics in the other and pages of numbers to go to the typist. No wonder I've never believed any government statistics since.

Splicing up a "bloopers 'n bangs" film from the 35 mm format trials records before binning the rest was fun tho'. Especially the bit where a missile was loaded into the box upside down so it turned right instead of left when it popped over the hill, heading straight for the VIP box instead of gathering into the remote sight field of view like it was supposed to. I guess that counts as a CNC failure as the gather controller was a little computer which worked out where the missile had to go to get from the launcher box into the controllers sight field of view given the separation and relative bearing. Not quite that simple when the launch vehicle is on t'other side of the hill. Worth the effort I guess. Parking a FV438 where an MBT can see you isn't exactly a safe way of going about things.

Clive