Electronic Design

What’s All This Limerick Stuff, Anyhow?

As I recently THREATENED, here’s a fun story about limericks (see “What’s All This Neutralization Stuff, Anyhow?” October 7). I’ve tried writing poetry and lyrics for songs, but I’m really no good at it. Part of the problem is my memory for old music from 40, 50, and 60 years ago is so good that I can’t do any new music because I remember old stuff too well.

However, I seem to be able to cobble up fun rhymes in the limerick form. I was driving from Denver to Boston, about 40 years ago, and you know how boring I-80 can be. I just happened to see the name of a town that suggested a limerick. So I made up one—and another—and another.

Since towns and exits pop up every five or 10 miles on the interstate, that was just enough spacing for me to concoct a limerick for every town. It was awful silly, but it did help pass the time. ALL THE WAY TO BOSTON! My wife wrote down some of the best ones and some of the worst ones. Here’s the best one that I still remember:

There was a young man from Chicago
Who went to see Dr. Zhivago.
He was such a Prude,
He considered it Lewd,
And decided to call an Embargo.

And if that’s one of the better ones, you don’t want to see the worse ones.

Back 15 years ago, I was having some debates with the supporters of Fuzzy Logic. Various readers had written some poems about Fuzzy. Finally one night I started to write a whole string of about 15 limericks. I typed them up in all caps and faxed them to Fuzzy Logic supporter Camerone Welch. Some of them were pretty good. Here’s one:

“On an elevator you may perch;
Only Fuzzy can prevent a lurch”?
—Such egregious claim
May yet Pease inflame,
And for refuge, you’ll hide in a Church.

Settling in to Limerick
I was riding in the passenger seat of our rented car as we drove through Limerick, Ireland, last year. As we came up at about 15 mph to a railroad crossing, I looked back on the right to see if anything interesting (such as a train) was on the tracks. Nope.

I immediately snapped my head around to the left to see if anything was coming up (or at least worth looking at). There wasn’t much there either. But I was surprised that I had turned my head almost 180° in about a fifth of a second, and it settled to a stop pretty well, so I could easily see what the tracks looked like. Wow. I didn’t know my neck could do that!

I’m not arguing that my head SETTLED perfectly, but it slewed well enough that my eyes could take up the slack and focus on the rails with full resolution. My neck must have settled to well within 2° or 4° so my eyes could take up the rest of the small motion, the fine focus.

Continue on next page

Measuring settling time to high precision has been a challenge for many years. I wrote a good research story for The New Lightning Empiricist about 40 years ago. Of course, it’s a little outdated, but at the time, it was a good start.

I wrote about the general principles of how to avoid stupid errors from overdrive, like how nice little Schottky diodes don’t necessarily just turn off when brought to 0 V. They may have tails of current that bleed out at low levels.

Old Tom Edison knew that to make a movie camera and a movie projector, he needed to get the film to stop and settle very quickly and stably. Of course, he cheated. He didn’t just servo the film to a stop, he CLAMPED it to a stop in a few milliseconds. Pretty ingenious.

Back to Work
Settling, and the instrumentation of settling, is a challenging business. How do we prove that it’s done right? You need a calibrator or “lie detector” to put in a signal that must be admitted to stop and settle fast. And that’s not dead easy. Even mercury-wetted relays aren’t perfect. But those are some of the techniques we have to use. What did Jim Williams just write? Two app notes at www.linear.com, AN120 and AN128, describe settling measurements to 0.1% in 2 ns.

Back in 1972, I designed a monolithic bit switch per U.S. Patent 3,995,304. It worked because the bits and current sources were all common-centroid. They rejected all kinds of gradients. And, the PNPs that set up the current sources never turned off, but the currents were just steered to ground, or to a summing point. They didn’t have a lot of thermal tails.

I was trying to brag about what a brilliant invention this was, and I realized I had never measured this settling. Was it down below the 14- or 16-bit level? Sure, but how about 18? MAN! I tried to guess how good it was—somewhere south of 16 bits?

If I could find one of these old bit switches, how would I measure its settling and its thermal tails? I soon got discouraged and quit planning that test. But if I had to, I could measure the settling a lot better than my neck did in Limerick.

Comments invited! [email protected] —or:
R.A. Pease, 682 Miramar Avenue
San Francisco, CA 94112-1232

Bob Pease obtained a BSEE from MIT in 1961 and was a Staff Scientist at National Semiconductor Corp., Santa Clara, Calif.

TAGS: Components
Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish