Charles Babbage is known for having conceived of the first mechanical computers. The original plans specified rather hefty components, roughly what you could expect from early to mid- 19th century mechanisms.
However, as time went on an industry with a similar need for gears, shafts, springs, etc.- the clock industry- continually reduced the size of the components needed. This culminated in the exceedingly fine mechanisms of quality Swiss watches, more or less at the limit of what could be assembled by hand with the right tools and a jeweler’s loupe.
I’m not sure how one would estimate this, but if you could build a Babbage Analytical Engine out of parts sized for the twentieth century watch industry, how much smaller would it be?
I’ve seen the difference engine in the London Science Museum, and IIRC the numbered wheels were about 7 cm in diameter. Using watch technology there should be no problem in reducing the linear dimensions by a factor of ten, and so the volume by 1000 times.
The interesting question to me is how much would that speed things up? Putting it very roughly, Babbage’s tech would have taken a few seconds to multiply two numbers. Reducing the size of the wheels by 10 times will reduce the angular inertia by 100 times. If we scale the forces and torques down by 10 times (because the wheels are smaller) we might hope for a 10 times speed increase, with multiplications done in hundreds of milliseconds.
I’m not sure you’d need “twentieth century watch industry” technology to get the kind of reduction Bert Nobbins writes about. Nineteenth century watch technology was very advanced as far as that goes. I’ve been to the Charles River Museum of Industry and Innovation, which is in the old Waltham Watch factory works, and their array of extremely tiny mechanisms and tools are truly impressive.
Appropriately, considering this question, they have the Watch City Steampunk Festival there in May every year.
In the Wikipedia article on Difference Engines, by the way, it lists several similar devices beuile up through the twentieth centrury. Presumably these did take advantage of smaller technology to reduce the size (and possibly decrease calculation time):
You could certainly reduce the size of the wheels, but I think the bigger concern is the load they have to drive. Watches have fairly small gear trains, but a difference engine (or analytical engine) requires individual wheels to drive huge columns of gears and levers and doodads. With small watch-sized wheels, the load could easily sheer the teeth right off.
I see no reason why the loads to be driven should not scale down too.
In the difference engine, special means were used to make sure that a lot of carries did not all happen at the same time, (which would create a large load) but instead were staggered over a machine cycle.
By the 1940s, there were in fact high-precision (12 digits, IIRC?) mechanical calculators small enough to hold in one hand, and to operate by turning a hand crank. That’s the equivalent of the Difference Engine, not of the much more complicated Analytical Engine, but one could probably scale down an Analytical Engine by about the same factor.
Of course, that was the peak of technology for mechanical calculation, because after that, everyone switched all of their efforts over to electronic calculation.
You’re referring to the Curta Calculator? I understand they’re being 3D-printed now. (checks: printed versions are x3 larger than original, but still…)
Make sure to stick to the original dimensions if you do. Those Curta calculators are designed to fit into one hand, while you crank with the other. Also you will probably need to use metal and correct tolerances to recreate the right “feel”.
Note, pocket electronic calculators in 1970 were not super cheap either.
Yeah, I know about the early electronic calculators. My dad bought one of the first ones, and justified it in terms of how much money he was saving, by not buying one of the mechanical ones.
Of course, he could have saved hundreds of dollars by just waiting a few more years to buy it, and just continuing to use his slide rule. Or nothing at all; it’s not like he was doing all that many calculations that needed it. He just wanted the latest and greatest new toy. Mom (a teacher) probably got more use out of it than he did: She brought it into her class, and let her students use it to check their arithmetic, and it still had enough cool factor to keep the kids interested.
I’ll mention that the inspiration for this thread is several SF novels by Rob Garitta centered on the fictional planet Zaonia.
To briefly recap, Zaonia was one of many colony worlds that was isolated by a major interruption of interstellar civilization, trade only resuming after a couple of centuries. The Zaonians had managed to stabilize their planetary civilization at the pre-electronics level, roughly equivalent to early 20th century Earth.
When interstellar trade resumed, Zaonia initially had a trade balance problem. Their tech was “obsolete” by galactic standards; and although Zaonia mandated remaining as self-sufficient as possible they still needed some interstellar trade, which had to be paid for somehow. To make matters worse, the local trade monopoly (think Trade Federation) cut off Zaonia’s access to the computation resources needed to launch or land traffic from the planet until they agreed to do things the monopoly’s way.
What Zaonia did was to employ technologies no one in the rest of space other than historians knew about. Things like logarithms, slide rules, adding machines, nomograms, and underpaid mathematicians working as freelance “computers” to crank out the calculations necessary. You can read excerpts here
In addition to all this, I was thinking that a scaled down to watchworks sized Analytical Engine might be another thing the plucky Zaonians could have used.
The problem is that, at some point, it becomes easier to switch over to electronic computers. Yes, even if you have to start from scratch: After all, we had to start from scratch on our world, too. As long as Zaonia has sand, they can make microchips.
I’m pretty sure it’s at least a leetle harder than that. If nothing else, maybe Zaonia simply doesn’t have the economy of scale necessary to support a semiconductor industry. This would tend to be supported by the presumption that Zaonia never forgot how to build high-tech, they simply couldn’t manage to pull together the means.
I have a Curta calculator. My dad bought it in the late 50’s. It actually works.
The problem with a scaled down Babbage, I’m guessing, that the problem is not just mass inertia of the gears, shafts, and wheels. That would scale down radically, due to the square-cube law. Another issue might be the resistance due to friction. IIRC toward the end of the mechanical watch era, this was solved by basing the shafts of each wheel in a tiny crystal (the “10 jewel watch”). I understand this provided the ends of each shaft with a base that was unlikely to wear or deform. But then, watch works moved on the order of seconds and minutes - less tress on the components.
OTOH the original IBM card equipment, before digital computers, used electric pulses to trigger things like mechanical counters, I think. The counters could count like an (old) odometer several dozen items a second, but that sort of mechanism would be very simple and unidirectional.