240 year old CNC machine

OK so cnc could be cams nuts and cranks :-)

Take a look at this http://www.chonday.com/Videos/the-writer-automaton

It's only 212 years old but at least the Jacquard Loom is programmable.

The loom is also very impressive, but not as cute :-)

“It’s a Babbage & Lady Ada’s machine =D”


The japanese have Karakuri tradition going back a long time, http://www.karakuri.info/origins/index.html

Also, PBS recently had a show about the Antikythera machine, looked at it with x-rays - 30 bronze gear wheels computing motion of the moon and planets. Click on the video. http://www.pbs.org/wgbh/nova/ancient/ancient-computer.html

Marc Isambard Brunel designed automatic machinery for carving ships blocks over two hundred years ago. Not exactly CNC, but very impressive.

Thanks AWOL, I only knew of his sons inventions, what a family !

To a certain level, this:


...could be considered an "ultimate successor" to The Writer and similar automata of the period. It and other "fire control computers" used many cams and other mechanisms to calculate with. In fact, one of the machines (perhaps the Mark 1 itself) used not only 2D profile cams for calculation, but also had carefully machined 3D profile cams for certain calculation needs.

This would basically be an extrapolation of the 2D cam "stack" used in The Writer (which was moved into position for each letter by the lower cam letter selection wheel). Such an extrapolation (a discrete set of cams in a stack to a continuous singular 3D cam) might have even occurred to designers of automata in the 18th century, but the realization using the technology of the time may not have been possible, or perhaps it wasn't even seen as necessary.

Then again, I am not an expert on automata of the period - so maybe such a system was realized and used at some point...?

I find it amazing that people had the patience and foresight to build these ingenious devices.

The Boeing 707s in the 70s still used a ball resolver for calculating position from the Doppler Navigator ( I think ) There is a picture of one on http://terry-kidd.blogspot.com/2009/12/analog-computers.html .

I recall a workshop legend of one flight crew ( who are trained to plot their position by a dozen different methods ) turning back to London when half way to JFK because the doppler packed up , they had grown to rely on it.

I struggled to believe a greasy metal ball turned by metal rollers, with all the vibration on an aircraft, could be so accurate !

Boffin1: I struggled to believe a greasy metal ball turned by metal rollers, with all the vibration on an aircraft, could be so accurate !

Actually, such resolvers (a similar one was used in early analog computers that used a knife-edge disc and a metal or glass wheel) would likely be very clean, and run "dry" (probably sealed as well) - with a bit of pressure between all the moving parts. I would imagine that vibration could be an issue; maybe they mounted them on some kind of shock absorbing system (with input/output via syncros or similar?)...

I remember our guys in the doppler section in the avionics workshop ( I was in the multiplex section - groundbreaking compared to the ball resolver :-) ) would put some form of grease on the ball before assembling in a sealed case.

But if it works don't fix it !

Also applies to the horizontal antenna jutting forward from the top of the tail of the 707, they found it radiated vertically .

And the antenna tuning unit for it that had servos tuning variable caps and inductors for optimum at the same time, with with no consideration of what the other servo was doing, but it worked.

Our instructor said nobody could actually explain it, and that had to do thousands of tests before getting it approved. He said they had hooked it to a metal crutch that some guy was using, and it whirred around for a couple of seconds and locked on.