Its because this was before electronic computers and software, so it was the only game in town. Experience beats theory all day long, so experienced practitioners of mechanical engineering can easily explain it to you, meanwhile theoretical mech engineers in your modern schools say "this is how I learned it secondhand". Its not really that we've gotten dumber. We just don't have people that are so experiential.
@@neglectfulsausage7689 That and folks were paid to have pride in their work, not this "get it done" attitude of the modern day. It was about getting it done *right*.
Because people took it seriously enough and they were allowed to be more cross skilled. Nowadays people learn a whole lot and then get put on a laser focused role, get disenfranchised and then the whole thing goes to shit.
My grandfather served on the USS Wisconsin in the 1950s in the secondary battery control room. We watched this video together and he's been telling me nonstop stories and memories about the MK. 1 Fire Control Computer, as well as his time in the Navy. Thank you!
My dad worked for Arma corporation during WWII who built many of the fire control computers. His specialty was gear train design. When I was younger I couldn't grasp the concept he would tell me of a mechanical computer, TRS-80s were just becoming popular. I really wish this video was available back in the 80s so I could appreciate his work when he was still with us.
Well, I'd bet the guy who wrote the script and narrating it are paid way more than your average teacher. There's also the fact that correct understanding of the material affected the lives of hundreds if not thousands of sailors directly and the war effort of hundreds of millions people indirectly, so there's more effort.
Designing this wouldn't really be difficult conceptually, but the effort put into machining, tuning, and calibrating the machine to give any semblance of a useful answer is admirable
Designing a complicated mechanical machine with just a drafting table and a pencil also deserves a great deal of respect, even today with kinematically linked 3D MCAD software this would have been a pain in the ass to engineer.
They used a lot more than slide rules. They not only used computers to plan the trip, but there was a computer on the ship itself. en.wikipedia.org/wiki/Apollo_Guidance_Computer
What I said was: "Hell, NASA put a man on the Moon using little MORE than sliderules." And that is the onboard computer. That's only DURING the mission. There was a LOT more that was done to engineer the hardware[which was, uh, alot], mission profile, etc. www.npr.org/sections/ed/2014/10/22/356937347/the-slide-rule-a-computing-device-that-put-a-man-on-the-moon "That seemingly simple tool has a serious resume. NASA engineers used slide rules to build the rockets and plan the mission that landed Apollo 11 on the moon. It's said that Buzz Aldrin needed his pocket slide rule for last-minute calculations before landing." ""The slide rule is an instrument that was used to design virtually everything," says Deborah Douglas, the director of collections and curator of science and technology at the MIT Museum in Cambridge, Mass. The museum just ended a three-year exhibit on slide rules. "The size of a sewer pipe, the weight-bearing ability of a cardboard box, even rocket ships and cars.""
Note that in the 1960s and 1970s the words "computing" and "computers" changed their meanings as stored-program digital computers came into widespread use. What used to be called computers, used for fire control (as here), astronomical prediction, gasoline consumption, and other applications, are now called "calculators", to indicate that they calculate some continuous function of the inputs, rather than evaluating a general series of logical steps (an algorithm) involving the inputs. This is why we now call these mechanical devices (including the Antikythera mechanism) calculators rather than computers. The principles of modern calculators are generally different from the principles of modern computers. We do have examples of pneumatic and other mechanical computers, used as toys or in industrial process control, but their performance does not even remotely approach that of electronic logic gates, flip-flops, memory devices and other such specialized components packed onto integrated circuits.
Well that's sort of a comparison between a special purpose or general purpose computer. We can also observe the transition in cars as we shifted from mechanical mechanisms to sensors and circuits to analog ECUs to general purpose computers running control software.
Before WWII, actually a computer was a person, during the late 19th and early 20th centuries, usually a woman with a degree in mathematics (though not always, many of the tables used in astro-navigation were computed by previously out-of-work accountants and engineers as part of a federal jobs program), who would do the heavy number-crunching for scientific endeavors, such as calculating the stress on a dam, wind tunnel fluid dynamics problems, or astronomy. The world's first professional computer programmers, for instance, were the computers who converted problems from human-readable math into machine readable form.
@Johnny When my father was conscripted in the early 1960's to the big field artillery they still did the 'computing' by hand. Two soldiers did the maths with a table and a slide ruler. If both had the same outcome it was a go. But when one of the couples in that line of canons had a difference my father's job (as an officer with a PhD) was to have done the calculation in his head already and decide. His tremendous speed didn't prove to be very inspiring later helping me with homework. So I copied his tactics: just imagine the computing and the tools in your head. It really works and even better: guys serving in the artillery regularly become deaf to one ear, the one closest to the BOOM. Since he was sitting right in the middle of a battery of 5-6, he is deaf as a post now. Good times...
That's exactly what they are. The full fire control problem was quite literally solving differential equations, and these computers could do it just as quickly and reliably as later electronic computers--actually slightly more accurately than modern digital computers (the analog computing did not have to include approximations). The one advantage the electronic computer has over the electromechanical computer is the ease with which it can be reprogrammed to do different sorts of calculations; these fire control computers had to be literally built for the ship that they were to be used on. While the basic computer itself could be used, any change in the guns that it was controlling required an entirely separate set of cams and gears to represent the different ballistic performance of the gun--so a computer for the medium-velocity 5"/38-caliber gun couldn't be used with the higher-velocity 5"/54-caliber gun introduced just a few years later, without completely replacing the cams and gears used in it.
I'd say that in some parts of the US, education is still taken very seriously. Also, if you personally take education seriously, you will get ahead of the pack regardless of country.
When I was a kid you could get surplus B-52 ranging sights for about ten bucks. They were also mechanical computers with lots of nice gears, ball bearings, optics (two very good 2" achromats suitable for small telescopes) and a great air-driven gyroscope. Cost for taxpayers in 1942 supposedly around $7000. Worth at least that much to me and many other nerdy kids.
@@288gto7 Oops. Thanks for the correction. I did a little research and found out what I had was not a B-52 bombsight, but a B-29 drift meter. Lots of ball bearings and lenses.
The U.S. Navy also used these during the Korean War. Very clever stuff. The existence of analog computers like these allowed the creation of the "CIC" (combat information center). We stand on the shoulders of giants.
I like how blatant the design of the barrel cam is so blatant. It's hard to describe. You would never know what it is for unless someone told you, and you never would be able to come up with it yourself, but when you see it you realise how simple it is for such a complex problem.
+Peacock486 Sort of - but moreso linear functions. If they were truly lookup tables, they would be non-linear and discrete - so there would be discontinuities, "sharp edges" or even holes/gaps that would probably prevent any rollers from working correctly on them.
Yeap, as long as they're defining continuous functions. Common examples include reciprocals, square roots, superelevation angle for the guns vs, target range, time of flight vs. range, etc. Also remember it will engage the runout at the edge of the cam if the input gets out of bounds. My grandfather was stationed in a gun director on a destroyer during the war and was well acquainted with this equipment.
I wonder if people of the future will look back on our electronics and say "that is fucking insane, how did they even manage to come up with such a crazy contraption?" lol
Leuel48Fan Oh, Im sorry, I incorrectly remembered what I originally posted and didnt bother to read it when I replied :P Yes, modern electronics are impressively intimate, definitely happy with my degree choice
Love seeing how computing was accomplished mechanically, love these old educational films. Sometimes I like to imagine digital programming like a machine.
Math never changes. The mechanical means of computing essentials like trig functions & products has a simplicity that really demonstrates the theory and how it manifests in the real world This content would be great for students who are mechanically minded but not very good or interested in maths
This is really helpful for understanding the fundamentals of what computing actually is. I'm currently on course to move towards a software engineering career path, and I sometimes find it difficult to conceptualize how code interacts with the physical world, and this really does help give me an understanding. To me, it feels like every "program" back in those days wasn't written, but actually BUILT. That might be the basis behind phrases like "building code", and maybe not. Either way, this was extremely interesting and insightful. Thank you very much for this upload!
Electronic fire conrol computers were invented in the late 40s and early 50s. But they weren't in widespread use in the fleet until much later, with the exception of newer ships. The US navy had plenty of WW2 surplus and thus most still used traditional mechanical computers, unless they were modernized along the way.
Fascinating! As far as I know, mechanical computers were in use well into the 60s, before miniaturisation, reliability (and response time) of digital computers had advanced enought to replace them. It's one thing to computer ballistic curves by room-large computers ('45 - '55) and another thing to constantly control a moving object (and the moving objects on it).
As a machinist, I'm watching this and feeling really puzzled as to how they managed to produce some of these cam shapes in the days before NC machines.
+Sylderon plotting the values on a 2D chart using tables and good old fashioned pen and paper and then milling until the radius on each point is correct. After that a copier milling machine will copy the came shape of the master for mass production.
+Sylderon Very easy my friend, hard work and not depending on google or smart phones like today's guys who want to copy and paste! hahaha! We used our own heads! If you saw the beauties I and my buddies created with files, lathes and milling machines and special tools WE created.
+Peter Diaz I've done some similar work myself...turning and then polishing bearing bores to +/- a couple microns on an engine lathe, turning out shafting on a 1970 LeBlonde Regal with badly-worn ways and still getting it to +0 -0.0005 inch and a Ra32 finish with no taper as a daily routine. Granted, it took longer than my supervisor wanted, but that's because I didn't half-ass everything like he did. I often feel that perhaps I should've been born around the turn of the century than in the 1985. Also, I don't even have a smart phone; never cared for touch screens.
They explain things so well. I like electric computers, but i love mechanical computers. Plus it's easier to learn. Example: spend hours trying to figure out on your own how to tell a computer to turn a electric motor 60 times in a minute or you could do something easier by using gears to make a wheel turn 60 times in a minute. I would make a mechanical computer.
+Patrick Bartosz Watching this video makes me wonder if they would've missed their targets at all if their guns weren't shooting a little inconsistently.
They kept the computer calibrated to the battery's average muzzle velocity (which changed over time due to powder temperature and other factors), and each gun could also apply a correction if its muzzle velocity strayed significantly from the mean (due to individual barrel wear, etc.). The computer also calculated for the effects of wind, ship roll/pitch, parallax due to distance between director and guns, and, most importantly, relative motion between ship and target. Spot corrections could be dialed in after observation of fall of shot (shell splashes). Computers for larger guns also corrected for Coriolis effect. Since a number of approximations had to be made in the fire control problem (including the assumption that the target's motion would remain in a straight line at constant speed during the shell's flight) and the computer's inputs relied on human operation in the director, its accuracy was by no means 100%, but this technology was state of the art in the 1940s.
Sometimes computational problems are better solved by more simple devices. Power companies still use "Analog Computers" which are nothing more than huge networks of Resistors, Capacitors, and Inductors, because it is easier to model their networks and the complicated differential equations associated with them by actually building out the circuits and testing them than with simulation by a transistor based computational device. Food for thought
Don't know how to tell you this but that's electronics. And I really doubt facility that cost millions uses only few electronic components for calculations
Near 11:55 the barrel cam is revealed to transcend all known laws of physics at the time, amazing that we do not use the technology anymore! The barrel cam is infinite and eternal.
I had the same question. How the hell do you figure it out? The answer is calculus, figuring many many many rectangles, small enough to get the resolution you need. The displacement of ships is figured the same way. It's like slices of bread. Slice that cam into 20,000 slices, figure the area/profile of each, then add it all together. Then for machining, you do that 1 'slice' at a time. Does that make sense?
It does not break any physical laws. Infinity does not equal to larger numbers or lenght. Infinity means there is no finite ammount of something, as you can always find a new variant that was never listed yet. That means infinity happends when it comes to decimals for example. What the narrator stated, is that it can output any number between x and y. And since that number includes decimals, then the ammount of possible outputs in infinite In a practical scenario, the ammount of possibilities is limited by the precision of the reading person or machine. But in theorical scenarios, the number is infinite, provided that your reading precision is infinite. This is similar to how at school in physics you do not consider the drag from air, or you assume a spherical cow.
@@darylmorning Have you ever seen how woodworkers use a simple turning lathe and cut various shaped and swooping design elements into a shaft of wood? This isnt rocket sceince.
The only problem with mechanical computers is that due to the mechanisms that drive them, they can succumb to wear and tear over its life time. That is especially true for parts that are small, making the use of mechanical equipment in microchips and systems on chips limited.
+Bungis Albondigas LOL as if electronics don't succumb to their own form of wear and tear? The big advantage of electronics is generally they work or do not where mechanical systems often work improperly before they cease to function.
+Bungis Albondigas The real disadvantage is that mechanical systems move at finite speed. You can switch ultra-low currents and voltages much faster than you can close a physical switch, or rotate a cam, or whatever. As ingenious as these old solutions were before the age of transistors, there's a reason we don't use them anymore.
@@Nevir202 Electronics, aka electron current switches, do wear out due to metal deposition drift. Eventually the microscopic distances between the elements used for transistors and their junctions allows the metal they're made from to be literally pulled apart, similar to galvanic behavior between two different metals. However, if you properly cool the part, this takes far, far longer, and most electronic microprocessors have redundancy built in. You may notice over the course of years your laptop getting slower. Thats because it doesnt cool properly. And its losing some transistors in its processor. But its a consumable item for a specific job. Given real good air flow, I've managed to keep various parts of computers running for probably 1 million real hours over 10 years. The first thing to fail was actually a pair of power transistors on the motherboard. two literally exploded, causing some of the housing to break and be ejected. Ive seen this happen to a guitar amplifier board, which I replaced. Basically its usually a triangular widget from near the center of the chip/transistor that flies off and it looks like someone gouged it with a philips screwdriver. But I've got a now 12 year old processor still running. Its a bit slower than before, but I over-fan the chassis rather than put the minimal required, so it stays basically around 98-100 degrees. They last a long time when you treat them well. Meanwhile, mechanical devices like this can also last a long time, provided you oil/grease them, and CONSTANTLY mediate them with new oil and grease. The raeson car engines break down after a few hundred thousand is because we're striking a balance between the cost of maintenance, the plan of obsolescence, and the value of keep ing it going. IF you had a constant flow of new oil and all the used oil was taken away, engines in the real world last millions of hours, as I think BMW or that other german car company (maybe it was swedish) showed. They put an engine to work for basically an entire year, thats 365 days running 24 hours a day, and kept the oil constantly flowing from new to engine to used waste oil that they got rid of, and the engine showed almost no signs of wear. So again, the issue of wearing down mechanically is an issue of lubrication and compromise between how much maintenance you want to do, and how much product you want to use, vs its expected end of life for replacement with newer and better and its overall value.
@@neglectfulsausage7689 Maybe I'm reading you wrong, but feels like you think you're contradicting me, when nothing you said contradicts anything I said?
Pay attention boys and girls, this is how the guns of Battlestations Pacific find their targets without your help from Battlestations Midway. Also, see the gunnery problem taken up to eleven from this point in this anime Last Exile - 2 - Lut Vanship (SUB) 9:06
Great film but I wonder why they didn't put the torque amplifier in it? Maybe because it doesn't do any computation per se but allows many integrators, etc. to be connected without the whole thing grinding to a halt.
cubesinanutshell could be. Maybe this is a guide for the users of the system to understand how it operates. But someone that repairs it must also know things like the electric motors that drive it, the torque amplifies, and likely a hundred other things.
I agree that this video is great but mechanical computers are limited by "backlash" inherent in all mating surfaces. The resulting error adds, multiplies or squares depending upon the mechanism. As a result, a complex chain of mechanisms will hamstring the supposed infinite resolution of the analog systems. In practice, even simple relay driven digital computers have better resolution in fire control systems.
It would have to be re-digitized from the original film, which is presumably somewhere in the government archives. I assume that this exists because it was converted at some point by the government itself, which would have used standard NTSC 640x480 (which it is), the resolution of most American TVs well into the 90s.
That is a Mark 1A Fire Control Computer. It's an analog computer and believe it or not a analog computer is faster than a digital computer. Your solution is present at the output as soon as you finish entering the problem. No clock cycles, just answers.
A large X-class solar flare is more than the equivalent of an emp attack, and could take out every computer on earth, as well as every power grid, and it would be at least a decade before it was restored, if it ever was. It would not take out mechanical computers.
@@jsat5609 Exactly. Btw...a German engineer (Konrad Zuse) made the first mechanical computer. A remarkable task as he was literally building this thing in his parent's living room. He even wrote his own programming language called "Plankalkül" for the Z3 computer. Imagine how far we could have got back then if not for a psychopathic Austrian with a ridiculous beard. Neural networking would possibly already be a reality?
I wonder who was the intended audience of this film back in the 50s. It seams a lot of work was put in it's realization (or were the images generated with the help of a mechanical computer as well? :-))
Does anyone know a good resource to learn more about this kind of stuff? A video series or book perhaps? Really interesting to see how everything fits together!
Tbh these were probably some of the more advanced mechanical computers before everything gradually turned into being electronic. I know analog computers might be something that will return as they are apparently really efficient at simulating AI, however these are also running on electricity.
Having a bit of programming background, I wonder if it's at all possible to have endless precision. But, it all comes to naught when you can only fit a couple digits in output, no matter how precise the calculation was
Look on Amazon for Norman Friedman's "Naval Firepower: Guns and Gunnery in the Battleship Age." Despite the title, it concentrates almost entirely on describing the development of fire control systems in most of the major navies of the world from about 1890 to 1950.
A question that comes to mind, say I'm a country with a less developed industry than the US, would it be economical not to go for the continuous calculation that the US went for but to go for step calculations dismissed by the narrator as not how the world works. My thinking is that although range doesn't change in steps, but loading your guns does, so if your average reload time is say 90 seconds, you would only need to be able to have a fire solution for each 90 second interval. Would building a less able computers and modifying your doctrine to suit be workable?
I swear these videos from the 30s-50s do the best job at explaining things
Its because this was before electronic computers and software, so it was the only game in town. Experience beats theory all day long, so experienced practitioners of mechanical engineering can easily explain it to you, meanwhile theoretical mech engineers in your modern schools say "this is how I learned it secondhand". Its not really that we've gotten dumber. We just don't have people that are so experiential.
@@neglectfulsausage7689 That and folks were paid to have pride in their work, not this "get it done" attitude of the modern day. It was about getting it done *right*.
This was back when the world didn’t move at breakneck pace.
Because people took it seriously enough and they were allowed to be more cross skilled. Nowadays people learn a whole lot and then get put on a laser focused role, get disenfranchised and then the whole thing goes to shit.
Or this is just survivorship bias
My grandfather served on the USS Wisconsin in the 1950s in the secondary battery control room. We watched this video together and he's been telling me nonstop stories and memories about the MK. 1 Fire Control Computer, as well as his time in the Navy. Thank you!
You should mention a story or two about this! Great Stuff
A team of FC's and ET's from my ship (USS Stout DDG 55) got her antennae turning and burning again.
My dad worked for Arma corporation during WWII who built many of the fire control computers. His specialty was gear train design. When I was younger I couldn't grasp the concept he would tell me of a mechanical computer, TRS-80s were just becoming popular. I really wish this video was available back in the 80s so I could appreciate his work when he was still with us.
Now that's how things should be explained in school... So clear..
Well, I'd bet the guy who wrote the script and narrating it are paid way more than your average teacher. There's also the fact that correct understanding of the material affected the lives of hundreds if not thousands of sailors directly and the war effort of hundreds of millions people indirectly, so there's more effort.
given a school wants you to understand in the first place
How people designed this just absolutely blows my mind
Designing this wouldn't really be difficult conceptually, but the effort put into machining, tuning, and calibrating the machine to give any semblance of a useful answer is admirable
Designing a complicated mechanical machine with just a drafting table and a pencil also deserves a great deal of respect, even today with kinematically linked 3D MCAD software this would have been a pain in the ass to engineer.
It really, really makes you respect the engineers before computers were mainstay. Hell, NASA put a man on the Moon using little more than sliderules.
They used a lot more than slide rules. They not only used computers to plan the trip, but there was a computer on the ship itself.
en.wikipedia.org/wiki/Apollo_Guidance_Computer
What I said was:
"Hell, NASA put a man on the Moon using little MORE than sliderules."
And that is the onboard computer. That's only DURING the mission. There was a LOT more that was done to engineer the hardware[which was, uh, alot], mission profile, etc.
www.npr.org/sections/ed/2014/10/22/356937347/the-slide-rule-a-computing-device-that-put-a-man-on-the-moon
"That seemingly simple tool has a serious resume. NASA engineers used slide rules to build the rockets and plan the mission that landed Apollo 11 on the moon. It's said that Buzz Aldrin needed his pocket slide rule for last-minute calculations before landing."
""The slide rule is an instrument that was used to design virtually everything," says Deborah Douglas, the director of collections and curator of science and technology at the MIT Museum in Cambridge, Mass. The museum just ended a three-year exhibit on slide rules. "The size of a sewer pipe, the weight-bearing ability of a cardboard box, even rocket ships and cars.""
Videos like these are pure gold. It seems that was a golden era for US navy (military) when it comes to educational videos ..
After seeing this film on the elegance of mechanical computers I really want to build one myself :)
So? Six year out... Did you build one?
Wow, if you really did build one, could you please share it on RUclips?
Yes me too
ME TOO!
Well? Is it built?
Note that in the 1960s and 1970s the words "computing" and "computers" changed their meanings as stored-program digital computers came into widespread use. What used to be called computers, used for fire control (as here), astronomical prediction, gasoline consumption, and other applications, are now called "calculators", to indicate that they calculate some continuous function of the inputs, rather than evaluating a general series of logical steps (an algorithm) involving the inputs. This is why we now call these mechanical devices (including the Antikythera mechanism) calculators rather than computers. The principles of modern calculators are generally different from the principles of modern computers. We do have examples of pneumatic and other mechanical computers, used as toys or in industrial process control, but their performance does not even remotely approach that of electronic logic gates, flip-flops, memory devices and other such specialized components packed onto integrated circuits.
Well that's sort of a comparison between a special purpose or general purpose computer. We can also observe the transition in cars as we shifted from mechanical mechanisms to sensors and circuits to analog ECUs to general purpose computers running control software.
Before WWII, actually a computer was a person, during the late 19th and early 20th centuries, usually a woman with a degree in mathematics (though not always, many of the tables used in astro-navigation were computed by previously out-of-work accountants and engineers as part of a federal jobs program), who would do the heavy number-crunching for scientific endeavors, such as calculating the stress on a dam, wind tunnel fluid dynamics problems, or astronomy. The world's first professional computer programmers, for instance, were the computers who converted problems from human-readable math into machine readable form.
@Johnny When my father was conscripted in the early 1960's to the big field artillery they still did the 'computing' by hand. Two soldiers did the maths with a table and a slide ruler. If both had the same outcome it was a go. But when one of the couples in that line of canons had a difference my father's job (as an officer with a PhD) was to have done the calculation in his head already and decide. His tremendous speed didn't prove to be very inspiring later helping me with homework. So I copied his tactics: just imagine the computing and the tools in your head. It really works and even better: guys serving in the artillery regularly become deaf to one ear, the one closest to the BOOM. Since he was sitting right in the middle of a battery of 5-6, he is deaf as a post now. Good times...
In a way these geared calculators are like a physical algorithm, no?
That's exactly what they are. The full fire control problem was quite literally solving differential equations, and these computers could do it just as quickly and reliably as later electronic computers--actually slightly more accurately than modern digital computers (the analog computing did not have to include approximations). The one advantage the electronic computer has over the electromechanical computer is the ease with which it can be reprogrammed to do different sorts of calculations; these fire control computers had to be literally built for the ship that they were to be used on. While the basic computer itself could be used, any change in the guns that it was controlling required an entirely separate set of cams and gears to represent the different ballistic performance of the gun--so a computer for the medium-velocity 5"/38-caliber gun couldn't be used with the higher-velocity 5"/54-caliber gun introduced just a few years later, without completely replacing the cams and gears used in it.
This thing explains computing better than most modern textbooks...
back when education was taken more seriously ?
+sickleandsuckle ..
hmm?
sickleandsuckle it's taken more serious now(unless you live in usa)
Try opening your textbook.
I'd say that in some parts of the US, education is still taken very seriously. Also, if you personally take education seriously, you will get ahead of the pack regardless of country.
The brilliance to design such machines is inspiring.
your user name is awesome does it have a meaning or is it a fuck it number pad tantrum and a j for fun type of thing?
This is awesome! Not only is it just a really neat subject but it is also presented in a way that's simple, but not dumbed down.
When I was a kid you could get surplus B-52 ranging sights for about ten bucks. They were also mechanical computers with lots of nice gears, ball bearings, optics (two very good 2" achromats suitable for small telescopes) and a great air-driven gyroscope.
Cost for taxpayers in 1942 supposedly around $7000. Worth at least that much to me and many other nerdy kids.
Nice, i wish i had one too. Also do you mean 1952 and not 1942? Because b52 entered service first in 1952
@@288gto7 Oops. Thanks for the correction. I did a little research and found out what I had was not a B-52 bombsight, but a B-29 drift meter. Lots of ball bearings and lenses.
The U.S. Navy also used these during the Korean War. Very clever stuff. The existence of analog computers like these allowed the creation of the "CIC" (combat information center). We stand on the shoulders of giants.
Where were you when I was in college? This guy explained it better than any professor I had.
How clear it is… very didactic. Could still be shown to young students at school …
I like how blatant the design of the barrel cam is so blatant. It's hard to describe. You would never know what it is for unless someone told you, and you never would be able to come up with it yourself, but when you see it you realise how simple it is for such a complex problem.
The multiplication 36:00 was the most interesting part. I would not have had that simple idea.
Yeah, it's basically a y=ax+b function
hey gurl u into math cause we could divide your legs, add my digits, and then multiply
So cams are basically physical lookup tables?
+Peacock486 Sort of - but moreso linear functions. If they were truly lookup tables, they would be non-linear and discrete - so there would be discontinuities, "sharp edges" or even holes/gaps that would probably prevent any rollers from working correctly on them.
Yeap, as long as they're defining continuous functions. Common examples include reciprocals, square roots, superelevation angle for the guns vs, target range, time of flight vs. range, etc. Also remember it will engage the runout at the edge of the cam if the input gets out of bounds. My grandfather was stationed in a gun director on a destroyer during the war and was well acquainted with this equipment.
Absolutely amazing! :] As a M.E. student, this fascinates me a lot. Thank you for sharing :)
I wonder if people of the future will look back on our electronics and say "that is fucking insane, how did they even manage to come up with such a crazy contraption?" lol
If you understand even slightly how they work (logic gates, binary counters, transistors), you've already said that a long time ago!
Leuel48Fan As an electrical engineering senior, I have a pretty good understanding, and I cannot say I agree with you in the slightest.
Kakunapod Why not? How a modern CPU works is just short of miraculous. Incredible engineering and testing the limit of physics going on there.
Leuel48Fan Oh, Im sorry, I incorrectly remembered what I originally posted and didnt bother to read it when I replied :P Yes, modern electronics are impressively intimate, definitely happy with my degree choice
Kakunapod lol ok. I myself are going to be studying Computer Science, should be great. Very interesting and cool field.
Love seeing how computing was accomplished mechanically, love these old educational films.
Sometimes I like to imagine digital programming like a machine.
Very useful for Venus probe design
Math never changes. The mechanical means of computing essentials like trig functions & products has a simplicity that really demonstrates the theory and how it manifests in the real world
This content would be great for students who are mechanically minded but not very good or interested in maths
This is really helpful for understanding the fundamentals of what computing actually is. I'm currently on course to move towards a software engineering career path, and I sometimes find it difficult to conceptualize how code interacts with the physical world, and this really does help give me an understanding.
To me, it feels like every "program" back in those days wasn't written, but actually BUILT. That might be the basis behind phrases like "building code", and maybe not. Either way, this was extremely interesting and insightful.
Thank you very much for this upload!
WE still had that when I was in Navy in the 60's
Electronic fire conrol computers were invented in the late 40s and early 50s. But they weren't in widespread use in the fleet until much later, with the exception of newer ships. The US navy had plenty of WW2 surplus and thus most still used traditional mechanical computers, unless they were modernized along the way.
Fascinating! As far as I know, mechanical computers were in use well into the 60s, before miniaturisation, reliability (and response time) of digital computers had advanced enought to replace them. It's one thing to computer ballistic curves by room-large computers ('45 - '55) and another thing to constantly control a moving object (and the moving objects on it).
As a machinist, I'm watching this and feeling really puzzled as to how they managed to produce some of these cam shapes in the days before NC machines.
+Sylderon plotting the values on a 2D chart using tables and good old fashioned pen and paper and then milling until the radius on each point is correct.
After that a copier milling machine will copy the came shape of the master for mass production.
+Sylderon Very easy my friend, hard work and not depending on google or smart phones like today's guys who want to copy and paste! hahaha! We used our own heads! If you saw the beauties I and my buddies created with files, lathes and milling machines and special tools WE created.
+Peter Diaz I've done some similar work myself...turning and then polishing bearing bores to +/- a couple microns on an engine lathe, turning out shafting on a 1970 LeBlonde Regal with badly-worn ways and still getting it to +0 -0.0005 inch and a Ra32 finish with no taper as a daily routine. Granted, it took longer than my supervisor wanted, but that's because I didn't half-ass everything like he did.
I often feel that perhaps I should've been born around the turn of the century than in the 1985.
Also, I don't even have a smart phone; never cared for touch screens.
It's called "Craftsmanship".
@@SynchroScore Have you ever used a mechanical calculator, the kind where you turn a crank handle to add the numbers?
Those designers were f-ing brilliant
Oh man, did they make any more videos like this? I could watch this stuff all day.
I always wondered how mechanical fire control computers were built. I would like to see one for the B-29's remote turrent systems.
It makes me happy to know that mechanical computers are now making a comeback.
They explain things so well. I like electric computers, but i love mechanical computers. Plus it's easier to learn. Example: spend hours trying to figure out on your own how to tell a computer to turn a electric motor 60 times in a minute or you could do something easier by using gears to make a wheel turn 60 times in a minute. I would make a mechanical computer.
By the 4th minute of this video i am enlightened with how mechanical computers works, the basic concept that is .
this is amazing and very clever use of mechanical to do computer
Hopefully this technology doesn't fall into the hands of North Korea.
+Bobbie Bees ahahahaha you son of a bitch!
+Bobbie Bees You would actually be shocked at how effective this technology was.
+Patrick Bartosz Watching this video makes me wonder if they would've missed their targets at all if their guns weren't shooting a little inconsistently.
+Bobbie Bees they have known that already......for your information.....
They kept the computer calibrated to the battery's average muzzle velocity (which changed over time due to powder temperature and other factors), and each gun could also apply a correction if its muzzle velocity strayed significantly from the mean (due to individual barrel wear, etc.). The computer also calculated for the effects of wind, ship roll/pitch, parallax due to distance between director and guns, and, most importantly, relative motion between ship and target. Spot corrections could be dialed in after observation of fall of shot (shell splashes). Computers for larger guns also corrected for Coriolis effect. Since a number of approximations had to be made in the fire control problem (including the assumption that the target's motion would remain in a straight line at constant speed during the shell's flight) and the computer's inputs relied on human operation in the director, its accuracy was by no means 100%, but this technology was state of the art in the 1940s.
Wow - so simple but exquisitely complicated! Wonderful to watch.
Absolutely beautiful. Mind = blown.
In the last year and a half I must be 200 of these views. So easy to fall asleep to
This video is simply great, analog precision is the only endless precision, even today
This is the funniest and the most awesome video I've ever seen on "computers".
just awesome ,i wondered to see those , how intelligently use those mechanical components to calculating ,,,,,,,,hats off
Wow 63 years later... ten times more computing can be on a chip the size of a grain of salt. Technology at its best.
in another 63 years the technology now will look like this video in comparison
I will never forget "range change" ever again
This will become an excellent laser math learning puzzle for children : addition, subtraction, multiplication and algebra. Thanks for sharing. ron
I love this. All this knowledge is being lost as machinists and mechanical engineers die off. Thanks for posting.
i now understand the c++ code even better
So simple yet so complex
Some humans are simply brilliant.
A "computer" at that time was a huge sophisticated slide rule. Or or hundreds of slide rules working together
Amazing how we engineered this. The insides look so ridiculously complicated.
Sometimes computational problems are better solved by more simple devices. Power companies still use "Analog Computers" which are nothing more than huge networks of Resistors, Capacitors, and Inductors, because it is easier to model their networks and the complicated differential equations associated with them by actually building out the circuits and testing them than with simulation by a transistor based computational device. Food for thought
Kameron Moon no they don't. en.wikipedia.org/wiki/Network_analyzer_(AC_power)#Decline_and_obsolescence
Don't know how to tell you this but that's electronics. And I really doubt facility that cost millions uses only few electronic components for calculations
Near 11:55 the barrel cam is revealed to transcend all known laws of physics at the time, amazing that we do not use the technology anymore! The barrel cam is infinite and eternal.
While it's cool and was useful for the time, it's also difficult to manufacture (can't simply turn one out, as it isn't cylindrical).
How the hell do you machine that? I got the linear and the rotation wrapped but how the hell do you combine the two with machine tools of the era?
I had the same question. How the hell do you figure it out? The answer is calculus, figuring many many many rectangles, small enough to get the resolution you need. The displacement of ships is figured the same way. It's like slices of bread. Slice that cam into 20,000 slices, figure the area/profile of each, then add it all together. Then for machining, you do that 1 'slice' at a time. Does that make sense?
It does not break any physical laws. Infinity does not equal to larger numbers or lenght. Infinity means there is no finite ammount of something, as you can always find a new variant that was never listed yet. That means infinity happends when it comes to decimals for example. What the narrator stated, is that it can output any number between x and y. And since that number includes decimals, then the ammount of possible outputs in infinite
In a practical scenario, the ammount of possibilities is limited by the precision of the reading person or machine. But in theorical scenarios, the number is infinite, provided that your reading precision is infinite. This is similar to how at school in physics you do not consider the drag from air, or you assume a spherical cow.
@@darylmorning Have you ever seen how woodworkers use a simple turning lathe and cut various shaped and swooping design elements into a shaft of wood? This isnt rocket sceince.
Holy crap. This is super handy! I want to 3d print one.
This has a very good explanation of a differential.
This is amazing how does it only have 8 comments ?
This is oddly therapeutic.
such a great video, i can really loose myself in this stuff.
The only problem with mechanical computers is that due to the mechanisms that drive them, they can succumb to wear and tear over its life time. That is especially true for parts that are small, making the use of mechanical equipment in microchips and systems on chips limited.
+Bungis Albondigas LOL as if electronics don't succumb to their own form of wear and tear? The big advantage of electronics is generally they work or do not where mechanical systems often work improperly before they cease to function.
+Bungis Albondigas The real disadvantage is that mechanical systems move at finite speed. You can switch ultra-low currents and voltages much faster than you can close a physical switch, or rotate a cam, or whatever. As ingenious as these old solutions were before the age of transistors, there's a reason we don't use them anymore.
@@fendermarxist that depends on the application.
@@Nevir202 Electronics, aka electron current switches, do wear out due to metal deposition drift. Eventually the microscopic distances between the elements used for transistors and their junctions allows the metal they're made from to be literally pulled apart, similar to galvanic behavior between two different metals. However, if you properly cool the part, this takes far, far longer, and most electronic microprocessors have redundancy built in. You may notice over the course of years your laptop getting slower. Thats because it doesnt cool properly. And its losing some transistors in its processor. But its a consumable item for a specific job. Given real good air flow, I've managed to keep various parts of computers running for probably 1 million real hours over 10 years. The first thing to fail was actually a pair of power transistors on the motherboard. two literally exploded, causing some of the housing to break and be ejected. Ive seen this happen to a guitar amplifier board, which I replaced. Basically its usually a triangular widget from near the center of the chip/transistor that flies off and it looks like someone gouged it with a philips screwdriver.
But I've got a now 12 year old processor still running. Its a bit slower than before, but I over-fan the chassis rather than put the minimal required, so it stays basically around 98-100 degrees. They last a long time when you treat them well.
Meanwhile, mechanical devices like this can also last a long time, provided you oil/grease them, and CONSTANTLY mediate them with new oil and grease. The raeson car engines break down after a few hundred thousand is because we're striking a balance between the cost of maintenance, the plan of obsolescence, and the value of keep ing it going. IF you had a constant flow of new oil and all the used oil was taken away, engines in the real world last millions of hours, as I think BMW or that other german car company (maybe it was swedish) showed. They put an engine to work for basically an entire year, thats 365 days running 24 hours a day, and kept the oil constantly flowing from new to engine to used waste oil that they got rid of, and the engine showed almost no signs of wear.
So again, the issue of wearing down mechanically is an issue of lubrication and compromise between how much maintenance you want to do, and how much product you want to use, vs its expected end of life for replacement with newer and better and its overall value.
@@neglectfulsausage7689 Maybe I'm reading you wrong, but feels like you think you're contradicting me, when nothing you said contradicts anything I said?
can it run pong?
lol, Maybe, if you custom built one for that task.
Brilliant thanks for posting this.
They're genius. I was impressed.
Prior MK111 UBFCS Tech (an ancient ASW fire control). So, only the really decrepit truly enjoyed this. Which I did with great relish! Every minute..
Pay attention boys and girls, this is how the guns of Battlestations Pacific find their targets without your help from Battlestations Midway. Also, see the gunnery problem taken up to eleven from this point in this anime Last Exile - 2 - Lut Vanship (SUB) 9:06
That multiplier was epic!!
I bet those gears are precision ones and installed with virtually no backlash.
Great video. Is there a textbook that covers all of this as well as this video?
Now I'm interested to know how they did these educational videos with overlays and animations without computers!
Simply beautiful.
Absolutely incredible. America was amazing
These mechanisms are so clever
Great film.
Great film but I wonder why they didn't put the torque amplifier in it? Maybe because it doesn't do any computation per se but allows many integrators, etc. to be connected without the whole thing grinding to a halt.
they may have but as the guy said this was just a short intro to the system.
cubesinanutshell
could be. Maybe this is a guide for the users of the system to understand how it operates. But someone that repairs it must also know things like the electric motors that drive it, the torque amplifies, and likely a hundred other things.
MrGoatflakes i think this may have been fire control computer 101
Well, the target audience was the US Navy so I'd imagine the level of effort was put in because peoples' lives would depend on their understanding it.
I agree that this video is great but mechanical computers are limited by "backlash" inherent in all mating surfaces. The resulting error adds, multiplies or squares depending upon the mechanism. As a result, a complex chain of mechanisms will hamstring the supposed infinite resolution of the analog systems. In practice, even simple relay driven digital computers have better resolution in fire control systems.
Wow, these 'primitive' analog computers were still REALLY COMPLICATED!!!!
It would have to be re-digitized from the original film, which is presumably somewhere in the government archives. I assume that this exists because it was converted at some point by the government itself, which would have used standard NTSC 640x480 (which it is), the resolution of most American TVs well into the 90s.
War truely is the great driver off innovation
That is a Mark 1A Fire Control Computer. It's an analog computer and believe it or not a analog computer is faster than a digital computer. Your solution is present at the output as soon as you finish entering the problem. No clock cycles, just answers.
let us all thank minecraft red stone engineers for laying the ground work for computing.
Nah
Pretty cool
Bird's ffffffgnnmmm ,
Blew my mind....
Well one would be shocked to know how accurate this was....and not vulnerable to EMP's or the like...
A large X-class solar flare is more than the equivalent of an emp attack, and could take out every computer on earth, as well as every power grid, and it would be at least a decade before it was restored, if it ever was. It would not take out mechanical computers.
@@jsat5609 Exactly. Btw...a German engineer (Konrad Zuse) made the first mechanical computer. A remarkable task as he was literally building this thing in his parent's living room. He even wrote his own programming language called "Plankalkül"
for the Z3 computer. Imagine how far we could have got back then if not for a psychopathic Austrian with a ridiculous beard.
Neural networking would possibly already be a reality?
Fascinating!
Great video
This is effffinnnn awesome!!!!
My mind was blown.
Awesome!
I wonder who was the intended audience of this film back in the 50s. It seams a lot of work was put in it's realization (or were the images generated with the help of a mechanical computer as well? :-))
Does anyone know a good resource to learn more about this kind of stuff? A video series or book perhaps? Really interesting to see how everything fits together!
Tbh these were probably some of the more advanced mechanical computers before everything gradually turned into being electronic. I know analog computers might be something that will return as they are apparently really efficient at simulating AI, however these are also running on electricity.
How does gear backlash not just completely overwhelm the computation? Don't even perfectly machined spur gears need some backlash to function?
I am going to build one of these one day.
Having a bit of programming background, I wonder if it's at all possible to have endless precision. But, it all comes to naught when you can only fit a couple digits in output, no matter how precise the calculation was
Awesome video.
This is beautiful
Nice info, thanks :)
How did they make that barrel cam with such a complex surface?
does anyone know a book where i can find this sort of mechanisms explained?
Look on Amazon for Norman Friedman's "Naval Firepower: Guns and Gunnery in the Battleship Age." Despite the title, it concentrates almost entirely on describing the development of fire control systems in most of the major navies of the world from about 1890 to 1950.
I love this stuff
Omg this is mechanic art
A question that comes to mind, say I'm a country with a less developed industry than the US, would it be economical not to go for the continuous calculation that the US went for but to go for step calculations dismissed by the narrator as not how the world works. My thinking is that although range doesn't change in steps, but loading your guns does, so if your average reload time is say 90 seconds, you would only need to be able to have a fire solution for each 90 second interval. Would building a less able computers and modifying your doctrine to suit be workable?
"This ain't no party, this ain't no disco, this ain't no foolin' around."
~~David Byrne, The Talking Heads
can it run music ?