W Scales -- Faber Castell 2/83 Style
Вставка
- Опубліковано 2 сер 2024
- I describe the usage of the double length W scales which give a 5" slide rule the precision of a 10" and a 10" slide rule the precision of a 20". My apologies for the wacky brightness -- after UA-cam killed the video editor, it's a bit harder to work with the raw video.
- Наука та технологія
I was the last user for slide rules at my university on 1980 and I decided towards 2/83 mainly for the W scale I named it as "double precision" - of course it is not double precision as we found in FORTRAN variables, but I liked the joke.
Thank you for post that video. You made me rescue my dear 2/83 slide rule that was in a drawer for ages. I have also another slide rule, a "pocket" model, Aristo 868.
As a just retired Engineer I think it is an opportunity to "re-learn" the features of that jewel.
Thanks for your video on the W-scales! I have to practice a couple of times to get used to the W-scales and when to read from which scale.
After a couple of weeks not using the W-scales, I forget how to use them. Practice, practice, practice....
Thanks for the video on the 2/83N and 62/83N. I enjoy these rules very much and think they are very powerful rules for computation. These rules came out at the end of the slide rule era and perhaps more manufacturers would have used the W scales if electronic calculators had not come out when they did. It does take some practice using these scales but in my opinion the effort is worth it.
These are the 2/83 and 6/83 NOT the 2/83N or 62/83N
looks great! the W scale might be a better substitution of A/B scale!
it's longer so more decimals!
I recently tried out the 2/83 and I think the solution of the W-scales does a good job of filling the 20 in scale onto the 10 in rule for compact relatively quick operations (rather than a long unwieldy 20 in rule). The real advantage seems to guarantee 3 sig figures esp. when doing log10 operations! I don't much like the vague guess you get sometime on the 10in for the 3rd sig fig. The downside seems to be its a bit slower and is harder to keep track of the decimal point and to get top/bottom misreadings. In a hypothetical "what is the best basic scale set?" I thought taking the standard scales and replacing the A/B to get something like K W2 [ W2' CI C W1' ] W1 D with [S L T] on the back would be optimal. I don't know if anyone put such an idea into practice. It might be fun to do a speed vs accuracy competition comparing, say, the 5in (e.g. N600-ES) vs 10in (e.g Post 1447) and the 20in 2/83 W-scales based on these (quite involved!) basic operation tests: mikeyancey.com/uil-slide-rule-resources/
That scale set is an interesting proposition. I think the K scale seems out of place on rules like this. Just like on the 2/83 (and Versalog, etc.), when you square a number you land on D, giving you the normal precision of the rule, but when you cube a number you land on K giving much lower precision. Now, squaring occurs much more often than cubing in most applications, but a handful of rules add the three cube root scales to address this asymmetry.
As for L, I never found the 3rd decimal place to be a problem on 10" rules that have graduations of 1/500 -- you just need to decide if the cursor is closer to the middle or to one of the marks. On the 2/83 though, I'd probably guess a fourth decimal place sometimes. Still, I don't find myself using L that much when there are log log scales, so this is probably more pertinent for your theoretical rule than for the 2/83. Similarly, I don't understand why a rule like the 1450 Versatrig doesn't use the 2/83 style L scale. I would have made that change (and probably swapped L and K's positioning).
So the problem I was thinking of is a situation of calculating y^x. the steps typically are find log(y) = .abc so that x * log(y) = d.ef and so you get y^x = 10^(.ef) * 10^d however as you can see 10^(.ef) is missing the last figure which sometime seems to cause problems. E.g. 4.62^5 on the 10'' gives log(y) = .664 and then x log y = 3.32 so ans = 10^(.32) * 10^3 = 2.09 * 10^3. If I do it on the 2/83 I get 10^3.32(3) = 2.105 * 10^3. The last digit in brackets is a bit of a guess but basically determines the final accuracy as you can finally do 10^(.efg) not just 10^(.ef) which is the advantage of the W-scales in the multiplication step use to calculate x log(y). If you use the log-log version you get very quickly 2.1 * 10^3. In other words the issue is determining 10^(.32?) = 2.?? which from the above versions is looking like 2.1? and of course for bigger numbers the log-log scale become unusable.
My comment about is about taking a 10'' and have the option to calculate with the C/D scales and the W-scales (by removing the A/B). This feature of the 2/83 duplex seems to make it stand out as unique. Maybe the preference not to go for the W-scales is on grounds of speed? Or is the fact that if you are in need of going beyond 2 sig figures (the 10'' gives normally a good guess for the 3rd sig figure) you really should get a log table and a curta peppercorn!?
Even the square roots and squares can be sometimes done more precisely on LL1/LL01 scales. For example as 0.8^2=0.64. But AB scales allow to get area of circle or volume of a cylinder almost instantly. The majority of the European slide rules have cursors for diameter to area conversion using S=pi*r*r=d*d*(pi/4) and there's 4/pi ratio on AB scales. ABCD scales are even extended on both sides and hairlines are on both sides, so it's possible to do area-diameter conversion on whole range without overflow.
Dietzgen 1741 (sort of a clone of the 2/83N made in Germany) also has W1/W2 scales
Uh. Interesting mental exercise. I even know paper slide rule, actually two cards where one card has 10 slots and and scale from 1 to 10 wraps ten times. Putting cards somehow on top of each other can be used as unwrapped cylindrical slide rule where scale is a spiral. Actually I'm not sure if these cards had horizontal scales or they had some slopes. I guess I may try to buy transparent foil or paper into printer and make it.
Heh thank you for the explanation brother. I don't see it on here anymore for some reason but I see it in my notifications so, whatever... maybe a glitch.
Anyway, I just wanted to mention the scale difference from the one you've got. Speaking of, do you know the model you are using on this tutorial? Also, since you are the ultimate slide rule junky (lol) I have one more question. The reason I'm trying to learn about slide rules is that my aunt, a math teacher and school administrator, is nearing retirement and I have decided to donate a collection/display of slide rules and other analogue/mechanical calculators to the Bossier Parish School System in her honor. As far as items go, I have a Reliable Co. Addometer, an Arithma Addiator, a (homemade) set of Napier's Bones lattice multiplication rods hand engraved in mammoth tusk ivory in an ebony box, and a huge list of more things. For slide rules, I have a mid-sixties Post Versalog I (black and red with green cosine) and this Faber-Castell 2/83N. What, in your opinion, are other the other absolute must-have slide rules I need to include in the display to make it feel "complete" insofar as including the most popular and representative models from the slide rule era. I'm kind of going for the opposite of obscure rare expensive ones, rather on the lines of popular, loved, "famous" and highly functional rules that define the overall slide rule category.
Not sure about the comment. The /N Faber Castell models came later and have expanded scale sets. I'd definitely add a K+E 4081 to that list. Maybe also a simpler model or a popular Pickett. I'll think about it a little more tomorrow.
After a little more thought, I'd definitely add a 4081/4181 and a metal Pickett (hard to pick which one but maybe the n600 or 800/n800). Top it off with a Mannheim/student model like the Post 1447 or K+E 4053.
If you're going "international" I'd add either a Nestler or Faber wooden Reitz and/or Darmstadt rule (e.g. A Nestler 23 like Von Braun's or the FC 1/87. For Darmstadt, the Nestler 21 or 0210, or FC 1/54). The Aristo Studio is also popular in Europe and produced for a long period.
@@ProfessorHerning awesome! Thank you so much!
@@ProfessorHerning Profe, one more question. If you had to toss out all the slide rules in the world apart from one which you could keep and use, which one would you choose and why?
That's a tough question! I'd probably be happy with a K+E 4081: classic mahogany construction and a balanced scale set. I usually keep one of these out on my desk and very rarely do I wish I had another rule handy. I also keep the pocket version in my briefcase.
That said, there are a number of other rules I'd be happy with like the Versalog II, Decilon, or Hemmi 260 (or similar).
My 63/83N has a C1 scale above L and a C scale below L ,Also a D and LL0 scale above LL1. It must be a different year . It is a 62/83N Novo Duplex.
The "N" models (2/83N and 62/83N) are introduced later and add some extra scales to the 2/83 design. Both 83N models are fairly sought-after on the collector market. I believe the late-model 63/83 has essentially the same smaller scale set as the rule I have here, but on an all-plastic frame.
Great video! After seeing this I managed to get my hands on a 2/83N and a 63/83. You're correct about the 63/83. Mine has the sames scale set as your 62/83 but with a build quality similar to an Aristo Studio 868 as seen in your video about trig with Euro-style rules. Same length and width, but the 63/83 is a hair thicker.
Hmm that's odd, profe. I have my hands wrapped around a nice, thick, long, lubed up Faber-Castell 2/83N right now. I've been using it while I follow along with your videos and at least the one I've got does, in fact, have an A scale (on the front, at the top, between the K and the DF scales) as well as a B scale (on the front, on the slide, between the CF and the CIF scales). What's up with that, papi chulo?
🍌✊💦😋
This one in the video is the pocket model, so I guess they decided some sacrifices were necessary for the sake of compactness.
Professor H - can I request you do a series on best ways of solving some of the higher numbered questions from the slide rule contest sheets? Example test 107 from mikeyancey.com/uil-slide-rule-resources/
Most of the last half of the questions combined square and square root functions or cube with cube root in single equation. Must be a better methods to solve than writing down intermediate results.
Note that the W1W2 sales help for first case but on a versalog or cube/cube root case you are back to pencil.
I'm thinking that in these timed tests there was no time for intermediate calculation and pencil work. Maybe they had better methods that I can't find in my readings.
Thanks
Also comparing the advantages or disadvantages of various rule types on solving the same equation efficiently would be very interesting.
For example versalog without AB but with R1R2 compared to European style Aristo studio or 2/83 or simple Nestler 23