Explaining the Friden Mechanical Calculator (while fixing it)
HTML-код
- Опубликовано: 15 окт 2024
- We fix a subtle carry problem in our Friden STW-10 mechanical calculator, which gets us into explaining how it works.
Friden Calculator Playlist: • Friden STW 10 Mechanic...
More about the Friden STW10 on my website:
www.curiousmar...
Our sponsor for PCBs: www.pcbway.com
Support the team on Patreon: / curiousmarc
Buy shirts on Teespring: teespring.com/...
Learn more on the companion site: www.curiousmar...
Contact info: www.youtube.co...
17:18
Zaphod Beeblebrox: "It's the weird color-scheme that freaks me. Every time you try to operate one of these weird black controls, which are labeled in black on a black background, a small black light lights up black to let you know you've done it."
- Hitch Hikers Guide to the Galaxy.
I still marvel that this amazing, complex, clever apparatus was designed, engineered, debugged, and built entirely by hand without any modern design aids like CAD/CAM, collaboration tools, etc. Engineers of the 1950s were a force to be reckoned with. Of course they did bring us space flight as well. I feel like our overall intelligence might have slipped a bit since those days now that we have so many tools to do our jobs for us. Same goes for old vs. modern programmers. It's comparatively easy to write some bloated program that has gigabytes of memory and GHz processors to play with. Not so easy to do the same task in assembly on the 'bare metal'. We've lost a lot of elegance since those days too. Forgot to mention that yours is one of my favorite channels and you do a great job of making forgotten technology accessible to a wide audience.
The knowledge and skills you mention with programmers has not disappeared. It still exists and many of us know it.
There just isn't a valid reason to do it now days. It almost always is going to produce less efficient code and will cost significantly more time to produce.
Use the right tool for the job. Regardless of if it is a right soldering iron tip, programming language or library. It is just good engineering.
Drafting despite it being manual labor is quite good and technical. Look at the results like you said. Good drafters back in the day were the CAD basically.
Economic reasons is probably the reason why we no such ambitious project not intelligence even then there still are such as race to a COVID vaccine, electric cars, different forms of energy etc etc....
One certainly has to hand it to the designers.
This is mostly a matter of perspective. This machine seems more complex because you can see all its internals moving, but modern machines are even more complex, you just can't see it because they use electronic gates instead of levers and gears. Have you ever seen the insides of a modern processor? It has literally billions of transistors. CAD programs can make the engineers' jobs easier, but they also allow them to increase the complexity of their projects instead. The algorithms have also improved a lot: it's more efficient to do math with binary numbers, loops and branches than to do math with decimals and dedicated hardware for each operation, but binary algorithms weren't yet well developed at the time.
Today, you are able to watch a video made by someone that is probably on the other side of the world from you, on demand, on HD, accessible all over the world, probably over wi-fi. You ever think about how complex this whole system is? This calculator is like a toy in comparison. Yes, a lot of modern applications are not elegant, and use "brute force" rather than good engineering, but they are only possible because someone else is so good at their job that they were able to develop some very powerful pieces of hardware that are also very cheap and small.
Also, it's not like the programmers of today don't program on bare metal; it's literally my job. I use C instead of assembly, because compilers can generate better code than I could and are also faster, but that's not because I can't program like the programmers of old, it's just that today's compilers are also better than what nearly all of them could do. C is still close to the actual machine code though, so to write drivers I read and write to registers directly, debug looking at the actual memory and registers contents, etc. Not all programming is done on PCs and smart phones, a lot of projects are still done on relatively weak processors that cost less than a dollar and have 64 to 256kB flash and 8 to 32kB RAM.
@@Destroier534 Humans don't design modern CPUs, software does. While that software was written by a human, it made the CPU design very abstract to us, compared to old school CPUs or some cogs and levers.
I worked on all Friden products from 1964 until 1975. In school (Rochester N.Y.) we had to strip the mechanical monster to it's frames them reassemble to working order. My last training was the mainframe computer which supported the first electronic cash registers mostly used by Sears. This required machine language programming in hexadecimal.
Both Redstone Arsenal (missile Command) and Marshall Space Flight Center (NASA) had about 500 of these each.
Or nation's space programs were founded using slide rules and mechanical calculators. Some would actually perform square roots.
Back before I realized how bad I was at hardware, a friend of mine and I used to do all sorts of cast-off-machine hackery. We would Dumpster dive office buildings looking for machines that they didn't bother to upgrade, they just replaced entirely. Local colleges and universities were treasure troves, as well. We tried cobbling together a Beowulf cluster, but no joy.
I am so glad that RUclips recommended your channel to me. Every single video is a joy.
With your talent for taking 'dead' hardware and not only breathing new life into it, but making it sing and dance, you, sir, are, to use a phrase we coined in the 90s, obsol33t.
Hey Marc, thanks so much for putting these videos on RUclips, and in with all the detail. I was lucky enough to happen upon a Friden at my local thrift store (what a find!!) and picked it up for a song. After combing through the STW manual you put online, combing through your videos & those of a couple other RUclipsrs, I've finally got it running!
I've got a nagging issue with the motor outrunning the rest of the machine and spinning endlessly, with the clutch making that "buzzing" ratchet noise. Sort of remedied it by massively reducing the engine RPM as described in the manual, but I'm working on tightening the spring tension on the lever that closes the clutch & electrical contacts.
These are a tremendously fun puzzle because they evolved so much over the years - between your models & the STW service manual, and mine, produced just a few years earlier, there are a remarkable amount of changes that I had to translate to my model.
I just checked mine to see if I have an issue with carry (mine has 17 accumulator dials, 9 in the register/counter, and 8 number key columns). Luckily, looks like the bell and the transfers work correctly, but looks like mine caps out at 12 transfers.
Mine's a Supermatic with Multiplication, the earlier 1930s type model with chrome trim and round green buttons.
Love your enthusiasm, thanks for inspiring a young guy to take on such a big task and follow through!
Awesome, and congratulations!
Curious Marc's RUclips videos are probably one of the most educational videos on RUclips. Watching as he bull doggedly delves into a complex electronic or electromechanical problem is awe inspiring. I recommend these RUclips videos to students studying electronic engineering. They might learn something!
Don't apologise for rambling, those that watch your channel call it "EDUCATION". Ramble on, dive deep, bounce off the bottom, ramble on.
Well worth repeating Ramble on!!
Also no need to apologize for your videography!
My dad used to sell these when I was little. I remember him having several different models of Fridens, Monroes and Merchants in our den as he was leaning to use them so he could demo them for customers.
Those carryover roller/gears are beautiful. Such craftsmanship of design.
"Use your imagination" ... that's something like watching SpaceX barge landings!
You're like the 8-Bit Guy, but before 8-bit computers!
The fact that the calculator actually does not need carry for all 20 digits, but only for a limited subset, to successfully multiply 10-digit numbers is fascinating. I need to think some more about the theory of it.
Oh I was just hoping for a nice wind-down thing to watch… perfect timing. Thank you Marc, I will enjoy this.
Another great Friden video Marc! I've been restoring my STW and almost everything works with the big exception of the lower dials which don't move at all. It's been driving me nuts but your videos help demonstrate how a properly operating machine functions way better than a fuzzy 1950's manual.
I suggest you get yourself a broken one, really, a parts machine and take it apart. I did the same to understand how my machine (another entirely) worked, and by the end, i managed to fix both. Start with the notion that the second one is just a parts machine and that regardless of how you take it apart, you do it just for parts. Obviously take pictures, and maybe even film things. To familiarize yourself, there's no other way.
@@aserta good suggestion. I do take lots of pictures along the way so I know how to put things back the way I found them. Marc's patience has been an inspiration so I will keep plugging away at it.
It's extraordinary that within my lifetime we have gone from these mechanical marvels to a quite different marvel, the modern computer that can do the same thing and more for just a few dollars.
My Monroe IQ-213 sat for over ten years before I found it in the back of a friend's curio shop. It actually ran flawlessly despite sitting for so long. The guy who ran the curio shop was a master machinist, retired engineer and technical educator who actually used it in his machine shop.
I still have a very early Rheinmetall at my bench (without electrical motor and without automatic division. so must be early 1920s ) for maintenance. :D
Thank you for bringing up the mechanical calculators to the younger audience, Marc!
I have the Friden DW10 which is the older model without the separate multiplication unit. I honestly find these machines to be amazing pieces of mechanical engineering. I like to imagine how these machines must have been received at the time - it must have been revolutionary just like the computer was 30-40 years later. I take mine into local schools to demonstrate what we did before computers - always find everyone to be amazed.
i enjoy reading comments such as yours and the appreciation of such mechanical (near) perfection. My grandfather (Carl Friden) would have a smile on his face. His daughter, Barbro Friden was my mother. When I'm in the San Francisco bay area I make it a point to visit the cemetery where my mother is buried and he is interred. Yes, my grandfather was the one who built this machine and his hugely successful company back in the day. I have some original blueprints of his machines. Fascinating.
Rick Alexander wow thanks for your reply. Your grandfather was a brilliant engineer for sure. Those blueprints should be in a museum somewhere - the impact these machines had on society should not be underplayed. At least as revolutionary as the computer on the ability to allow companies to scale up. Thanks for your reply.
Beautiful. Its hard to think about what it took to even design the thing. Brilliant.
please show more of this mechanichal calculator. it is so cool seeing it do math the same way you would with logic gates. I built a basic 8 bit adder/subtraction circuit in vcv rack recently and seeing a mechanical alternative to shifting after a digit moves past 9 (or 1 in bianary) is pretty cool
Thanks! I have 6 videos to date, here is the playlist: ruclips.net/p/PL-_93BVApb5_GbzkFwdKpuyPo8Ty83h8b
Best steampunk calculator ever! I have watch all your uploads Marc your passion for vintage electromechanical machines is just wonderful thank you for saving these machines for posterity!! Cheers from Salt Lake City, Utah
The amazing and differing methods of carrying numbers on mechanical calculators - great fix on the extended carry section of your Friden.
I finally got a Lagomarsino (non electric) mechanical calculator apart last night - it's been sitting on the bench for a couple of months totally seized, to the extent that I couldn't get it apart - finally penetrating oil and heat got the shaft moving that allowed me to get the carriage off and now I can start working on it properly.
Before you pull it out of storage, warm the machine up by a few degrees, more, than what was in the storage when you got it. Don't absurdly heat it, just enough to get everything going. On my machines, i pull the covers off, and blast it with my hair drier (old one) on medium in the mechanisms to heat things up a bit, and i find that it makes first operation, which is always a basic number movement of all the operations it's capable of, in 00 digit skips, a lot, lot, lot easier than doing it cold.
It makes sense, in my head, that a machine, any machine of this integrated level, needs some kind of warm up, and like i said, a humble hair drier does it.
Get all the old sticky grease a bit more fluid, and expands the metal. Nice idea
The ENIAC (the first electronic general-purpose digital computer) was essentially an electronic version of this mechanical technique. I can imagine John Mauchly and J. Presper Eckert thinking to themselves: what if we made an electronic version of the mechanical calculator. We make 20 electronic "accumulators" of 10 digits each, provide some switches and a common bus to connect the outputs to inputs and we are in business. The mechanical gears are replaced with "ring counters" using vacuum tubes. I believe this is the only computer designed this way. I read somewhere that late in the construction of ENIAC, the designers realized that using binary adders is a better idea.
I think some of the earliest electronic desk calculators used decimal-based computation, like the ANITA and the Friden EC-130. The Friden's memory was based on magnetostrictive acoustic delay lines and the numbers were stored as decimal digits represented as trains of pulses, almost like a rotary dial phone system but not quite. Each digit was a string of five bits, and from 0 to 9 they were 00000, 10000, 11000, 11100, 11110, 11111, 01111, 00111, 00011, 00001. Very weird.
Thank you Marc, a great explanation and fix as usual, looking forward to the next video as always!
I have one - an stw10 - I thought I was good at mechanical things . Every time I get something to work something else is stuck . Now the multiplier wont move and I have no idea how it works. I am afraid to take it apart. I will never get it back together again. How could any one design one of these things ?
Now I need one of those calculators. Thank you Marc.
I still want you to restore a Linotype machine!
That is the pinnacle of complicated mechanical contraptions! You probably have to devote your life to it!
That would be very cool ... just please not like the one from " The Twilight Zone"
Also would love to see the ibm machine you got a few years ago ... a 360 I think, i remember the tape machines and the high speed printer coming on the back of a truck.
wasn't teletyp-o-operated linotypes a thing?
My mom operated a Linotype in the 70s. There were four or so in a vast museum in Cape Town a while ago. Highlight of all the mechanical exhibits. Many great videos on YT. Thank you Marc for all the fascinating videos.
What I have read of mechanical calculator maintenance back in the day was that the whole machine would be lowered in to a vat of kerosene and light oil. The kerosene would dissolve the gummy old oil, and the oil would lubricate it until the next time. I didn’t have the confidence to do that to my MADAS. I used q tips with various solvents then a drip of light oil. Now it is working, I try and exercise it every month.
Sounds scarily close to my hacky as hell method of dealing with sticky complex light machinery, hose it with WD40 (naptha/kero and oil), excercise until functional, blow off drips, once the solvent evaporates there's a film of oil on everything. It makes sense though, not like there's anything going on in there requiring extreme pressure lubricants or whatever, just getting everything sorta greasy again is adequate.
Sounds a lot like how in the 70's the watchmaker's tool supplier Elma tried to convince watchmakers that you would get good results by leaving a movement assembled and running it through their fancy machine that had a solvent/oil bath at the end to lubricate the movement.
That is exactly what we did.
These videos are so good!! I doubt I’ll ever work on anything mechanical the way you do but I love watching!
I love electromechanical calculators. Someday I will have on of my own
Start with a mechanical one. Cheaper, and if you get one that's broken, you won't cry yourself to sleep. :D
@@aserta any recommendations on a type or style to work with?
Hi Marc! Thanks to this video I was able to figure out why the carry between the 12th and 13th digit wasn’t working on my STW-10. Unfortunately it seems that the two levers you dremeled in this video have shorn off. Do you have a source for parts you can suggest?
Not really. I don’t think there is a Mr. Friden equivalent to our Mr. RTTY. That said these parts are mostly flat, should be possible to have new ones laser or water jet cut. Or weld a new tab on top. Not that easy to get them out of there though, this is really buried.
CuriousMarc I figured as much. Any chance you could give me the dimensions of the part in question? Do you have a scan?
@@timothycolegrove4365 Go on my web page, www.curiousmarc.com/mechanical/friden-stw-10-mechnical-calculator, I have all the manuals there. It's drawn somewhere in one of the manuals (I think the service manual), the page I show in the video where I explain the ca - but can't remember exactly where. Usually the drawings are to scale. It will give you a good idea. Before you make a copy you'll have to extract the real part. The ratchet must be attached to it, so it must have a pivot and several spring attachment points.
What a wonderful mechanical contrivance. I am saddened that it would very difficult and costly to reproduce it. Still, they should : ) The flat parts could be scanned on a flatbed scanner for reverse engineering.
I think the proper way to service that thing is equivalent to old mechanical clock. It needs to be fully disassembled, cleaned and reassembled (with fresh oil and adjustments). But of course it is very challenging thing to do.
That would be a proper restoration indeed, except the complexity of clocks and this is not comparable. It's not even close. I considered it, but it's extremely risky in this poorly documented machine: we are talking something like many thousands of parts. And it's the most non-modular design I have ever seen: everything is connected to everything. Then, assuming you did not lose or break a part and put them all back together were they belong, recovering the correct timing without the proper factory instruction would be a nightmare - if even possible. We are talking several month of work, with a good chance that you'd never be able to put it back together. So I decided against it, and only took apart about a third of the machine, including the entire backplate assembly, which was quite risky. I had no choice, it was entirely stuck.
@@CuriousMarc It is surprise to me that such thing doesn't have fully detailed service manual. How did they managed to fix such things back in the times? I agree that you need to dive deep in specialized literature (if it ever exists) before doing anything with such mechanical wizardy. I remember my first tries to fix old soviet mechanical alarm clock when I was just a kid... Its spring almost broke one of my fingers☻ Well, after years I managed to read some books and get proper tools. So now I have modest collection of old clocks and watches with everything in perfect shape. And yes, even simple alarm clock mechanism can easy take a whole day for restoration. With that calculator I can't even imagine amounts of time even with detailed documentation. At least you can lubricate everything you can reach with oil. Not really everything of course because obviously some mechanics needs strong friction to work properly.
We actually washed them in a machine with a turntable and spray nozzles using solvent. They were then rinsed, dried and relubricated
They were serviced by techs who attended a 13 week school.
People say things today are a marvel of technology.
This thing does relatively complex math with gears, levers, and springs!
Beautiful piece of hardware.
Awesome, as always.
Your opening sounds like Dr Seuss book. :D Just something sing song about it.
impresionante, son unos grandiosos ingenieros, algún día seré como ustedes.
Wow what an amazing machine
Never heard of Nye oil... I wonder if it's more similar to marvel mystery oil, liquid wrench or seafoam.
The SDS says it's half mineral oil, the other half is trade-secret... Likely various detergents and lubricity enhancers.
It was squeezed from Bill Nye, back in his misspent youth..
It's called a "lovejoy coupling". I didn't name it) Love the videos
Thank you for your work
Charles Babbage, surely, was eventually going to invent this machine.
Amazing mechanizm.
Disponible en Caracas Venezuela
Every time I see the innerds of these machines I’m more impressed than before. The design and fabrication and assembly of that many machine parts, in mass production boggles my mind. Hardened steel gears, cams, rollers, springs, levers, and sheer compactness and density. I assume just about 100% assembled by hand.
I hate to think how many working Fridens were just tossed in the bin due to obsolescence and still perfectly good.
My imagination tells me there are tiny unicorns missing making this thing tick... mm sorry, ping...
Cool trip thanks.
Only CM can make something longer by filling it. 😁
I hardly ever witnessed the sound of mechanical calculators, which obviously were commonplace until the late 1970s and before integrated circuits took over. However, I very well know this kind of sound from films and TV shows which aired even ten years later, for example Knight Rider. To me it seemed ridiculous to put a "modern" computer on a show which is that noisy. But I suspect that most viewers and directors were still conditioned to believe that calculation power had to be in line with the noise a machine makes.
Great stuff :-)
Piece of engineering
"In the morning I bent a lever. In the afternoon I bent it back."
Have tried to do a divide by zero? Long ago when I had a student summer job in an office they had a calculator that looked a lot like this one. Once at lunch when no one was around I tried to divide by zero. It ran and ran and ran and wouldn't stop! I had to pull out the plug!
Yes, he did. Just in the last video.
ruclips.net/video/-MLQ0yI1BrQ/видео.html
Division by zero is done in the previous video. And this machine has a "division halt" lever for this exact reason.
@mirskym: You bet: ruclips.net/video/7Kd3R_RlXgc/видео.html. For some odd reason, one of my most popular videos.
I think today the more apt joke would be that you are Jeff Bezos and you are counting your trillions of dollars.
👍👌
"too much wear" Hmmmm, isn't that some of the least used digits?
But the reset cam does one rotation for every cycle of the machine. Regardless of the actual use of of the digit.
The whole mechanism moves, doesn't matter what you do. Limitation and flaw in one.
These machines, unfortunately, have a finite life. Once the system gets worn down enough, that's it. It's over.
I have a french make machine from the 1800's and it's at that stage where the components are so utterly worn down that it cannot operate anymore. And the wear is exactly like here. The digits are bleached only in the upper digits, meaning those were barely used, but the whole mechanism is worn down because the whole machine resets.
:D "patato vision"
and people think automobiles are complex xD
Just make it programmable and put a gameboy emulator on it.