Using a slied rule is like riding a bike

[W1GUH]

Once you learn to do it, you never forget how.

Got a slide rule this year at Deerfield. It's a really nice Pickett Power Log Exponential, Log Log dual base. Must have set the original purchaser back a bit. What's happened for casual calculations is that I'll reach for that before I call up he Windows calculator -- it's much more convenient. That's not simply an OF nostalgia thing -- it really is faster.

Image here. (http://sliderulemuseum.com/Pickett.htm)

Scroll down to, or search "Pickett N3-T - Power Log Exponential"

Cost me $20.

[K7SGJ]

Yeah, but can you draw straight lines with it?

[W1GUH]

The lines you got from a French Curve were far sexier.:giggity:

[ab1ga]

My favorite is still a Smith Chart slide rule distributed by Microlabs/FXR in the sixties. A bit rough on aging eyes, though.

[W1GUH]

Got a pic? I'd love to see one. Sounds cool.

[ab1ga]

Not of the one I own, most of my belongings are in random file storage boxes. It came is a dual-pocket folder with instructions on how to use it printed inside, and a list of distributors on the rear cover.

I'll take a look for it, but if you don't hear from me soon, it means I've stumbled across the Lost Ark of the Covenant and I'm being distracted...

On edit:

Found a picture of it on the Web, let's see if the link works.
http://www.rfcafe.com/references/electrical/slide-rules.htm

It's about half way down the page.

The other weird slide rule I have isn't nearly as nice, made of paper with the bit rivets at the corner, used to calculate blackbody radiation parameters.

What I'm really looking for is books on nomography, which, like the slide rule, trades off precision for speed and simplicity, and lets you get on with the job. I'm afraid that's going to become a truly lost art very soon, which would be a shame.

73,

[W1GUH]

Let us know what gopod says.

[n2ize]

Yeah, every now and then I play around with my old slide rule. It's fun but inferior to electronic calculators or, simply using approximations on paper, both of which are faster and more accurate. Nonetheless, the slide rule had its place and time and is worthy on the merit of how it functions.

[W1GUH]

Yea, but how many times do you really need n-digit precision? 2 or 3 digits is usually more than sufficient for what you're doing. For example, I did the 275x4 for the Matchbox on the slide rule. Because I was publishing the number, I checked it with a calculator. But, say, power input. Never need more than 2 or 3 digits.

Faster? Gotta find it or bring it up on the screen, then go through tedious keystrokes.

[NQ6U]

The slide rule was good enough for the designers of the SR-71 but I don't think I'd remember how to use mine any more.

[W1GUH]

Ya never forget.

[K7SGJ]

It's true slide rules are like riding a bike. Just harder to put cards in the spokes.

[HUGH]

No batteries required.

[NQ6U]

No batteries required.

Once, when I was working as a production artist/Mac support guy at a pre-press service bureau, I bought a little box with a glass front and hung it on the wall of my cube. I put an old abacus inside and a label on the front that said "IN CASE OF POWER FAILURE, BREAK GLASS". It got a lot of laughs.

[n2ize]

Yea, but how many times do you really need n-digit precision? 2 or 3 digits is usually more than sufficient for what you're doing. For example, I did the 275x4 for the Matchbox on the slide rule. Because I was publishing the number, I checked it with a calculator. But, say, power input. Never need more than 2 or 3 digits.

Faster? Gotta find it or bring it up on the screen, then go through tedious keystrokes.

Usually I just crunch the numbers in my head, or on a piece of scratch paper and get good accuracy to a decimal place or so using an approximation technique.. For me its faster than a slide rule or a calculator. Keystrokes ? Not necessary For large expressions I do the calculations using either bc or lisp in interactive mode. Just type in the expression exactly as you want it calculated, hit "enter" and I've got the result in a fraction of the time it would take if I were to use a keypad style calculator or a slide rule.

In the old days some of the guys were pretty fast with a slide rule.Unfortunately I never reached that level of proficiency so, for me these other methods are a lot faster and at least as accurate or more accurate.

[n2ize]

Once, when I was working as a production artist/Mac support guy at a pre-press service bureau, I bought a little box with a glass front and hung it on the wall of my cube. I put an old abacus inside and a label on the front that said "IN CASE OF POWER FAILURE, BREAK GLASS". It got a lot of laughs.

The abacus can be a very powerful tool. There are some who are blazing fast with an abacus. It can also be very useful in demonstrating how computers do calculations.

[K7SGJ]

Usually I just crunch the numbers in my head, or on a piece of scratch paper and get good accuracy to a decimal place or so using an approximation technique.. For me its faster than a slide rule or a calculator. Keystrokes ? Not necessary For large expressions I do the calculations using either bc or lisp in interactive mode. Just type in the expression exactly as you want it calculated, hit "enter" and I've got the result in a fraction of the time it would take if I were to use a keypad style calculator or a slide rule.

In the old days some of the guys were pretty fast with a slide rule.Unfortunately I never reached that level of proficiency so, for me these other methods are a lot faster and at least as accurate or more accurate.

I went to school with a girl from bc that had a lisp. Couldn't do math for shit, but was a real looker, eh?

[W1GUH]

I should be a littler clearer about this....I'm only talking about basic multiplying and dividing. And, yea, head crunches work good, too. Of course, heavy-duty software is the convenient and fast way to go, IF your computer is on and ready to go! But that goes without saying!

As for abacus', I always preferred the Japanese Soroban, refinement of the abacus. You don't really need two beads on top or 5 on the bottom. 1 and 4 make the calculation easier.

http://www.ee.ryerson.ca/~elf/abacus/images/tomoecloseup.jpg

[KC2UGV]

Yea, but how many times do you really need n-digit precision? 2 or 3 digits is usually more than sufficient for what you're doing. For example, I did the 275x4 for the Matchbox on the slide rule. Because I was publishing the number, I checked it with a calculator. But, say, power input. Never need more than 2 or 3 digits.

Faster? Gotta find it or bring it up on the screen, then go through tedious keystrokes.

Whenever you multiply or divide more than once. The rounding errors start getting noticeable on the second operation.

[W1GUH]

Tough crowd! Force me to specify the problem precisely! ;)

I'm talking about things like plate current x plate voltage; MPG;

[ab1ga]

Tough crowd! Force me to specify the problem precisely! ;)

I'm talking about things like plate current x plate voltage; MPG;

If you formulate the problem more precisely, you'll wind up in the land of false precision.

I agree with you about significant digits for "real world" values. Most measured quantities are know to no better than 5 or 10 percent accuracy, so even multiple computational steps won't add significant numerical error. Measuring a quantity, or even knowing a physical constant to better than one percent is actually hard work, especially in the face of environmental variation.

The trap of false precision can even trap those who should know better. A professor of mine once took a number with one significant digit, multiplied it by a number with two significant digits, divided it by a number with three significant digits and voila, wrote down an answer with four significant digits! I could have cried.

73,

[n2ize]

I went to school with a girl from bc that had a lisp. Couldn't do math for shit, but was a real looker, eh?

That sounds nice. I knew a girl in school who had a lisp. She was cute as a button. Ability to "do math" is relative. I still can't always add, subtract, multiply, and divide, correctly but, I can prove complicated theorems. One of my math professors put it like this. There are 2 types of mathematicians. Those that can add, subtract, multiply, and divide and those who cannot. He admitted he was in the latter class.

[NQ6U]

One of my math professors put it like this. There are 2 types of mathematicians. Those that can add, subtract, multiply, and divide and those who cannot. He admitted he was in the latter class.

You know, John, that makes me feel a lot better about myself. I have never been able to do simple arithmetic worth shit but I had no problem understanding algebra. I usually got the wrong answer because I screwed up the multiplication or something but I got the concept just fine. Not that I can remember any of it, of course.

[WØTKX]

http://sokol-blog.com/wp-content/uploads/2006/11/dilbert200611295196.jpg

[W1GUH]

I went to school with a girl from bc that had a lisp. Couldn't do math for shit, but was a real looker, eh?


"lisp" is the cruelest word in the English language. If you have one, you can't say it.

BTW, is lisp actually used anymore? I know it was the Next Great Thing a couple of decades ago, but haven't heard hide nor hair of it since. Like AI or Neural Networks.

Matlab makes a great calculator!

[WØTKX]

A strange version used to have me counting parenthesis late into the evening...

http://en.wikipedia.org/wiki/AutoLISP

[n2ize]

"lisp" is the cruelest word in the English language. If you have one, you can't say it.

BTW, is lisp actually used anymore?

Sure its still used. Just as FORTRAN is still used. Not very much these days but there are still a lot of people using it for a wide variety of different things including AI.These days there are multitudes of Lisp variants and dialects, many derived for specialized purposes. Emacs is programmed in lisp as are most of its extensions and it is still widely used in the *nix/L*nix world. I like it for writing mathematical expressions and certain functions can be expressed very easily in Lisp as opposed to languages such as C, Java, C++, etc.

It's like anything else, it is one of many tools available to do a job. The one you chose depends on the job you are doing. There are some projects where I would chose Lisp. Other projects where I would chose C, or maybe C++, or Java. Or maybe Python or Perl. If I were doing web programming I would chose php or some of the popular web oriented langauges. Sometimes for the fun of it I might even toss a little assembly into the mix.