Electronic Design

Bob's Mailbox

Dear Bob: Just got around to seeing your June 9 column. The apparent shift in perceived pitch you asked about is due to the lack of compensation in the inner ear for changes in environment. While not an expert on ears, here's the way I understand it: The inner ear has resonant fibers in the cochlea that stimulate attached nerve endings at a level corresponding to the fiber response amplitude. An array of nerves scans the outputs from these fibers and develops what amounts to a Fourier transform of the received sound. If a single frequency is received, only one fiber has a max response. The array locates it and reports that to the brain. If biting changes the pressure in the inner ear, the resonance of the fibers shifts, a new fiber becomes the max responder, and the analysis array, not being compensated for pressure changes, sees only that a different fiber is now responding and reports a change in received pitch.

Another effect that is fun to observe is the AGC function of the ears. If you are hearing a medium-level broadband noise and then drive a nail with a hammer, the AGC will respond to the peak sound of the hammer blow and suppress the broadband noise for a fraction of a second. Recovery follows the usual exponential curve and restores the original sensitivity in a few tenths of a second. You have to do this outside so the echo of the hammer does not mask observation of the recovery process, which is faster than the echo time constant of most rooms.

Richard D. Lee (via e-mail)
Pease: That's one of the most thoughtful replies on the subject I've seen. About a third of my readers say they hear this effect. We don't know if younger or older people hear it most.

Dear Bob: Without further evidence, I can't accept Dave Weigand's assertion (Electronic Design, April 28, p. 20) that today's embedded microprocessors are going to last only three to seven years, limited by electromigration. I would guess that electromigration goes as voltage squared. But smaller silicon geometries are running on lower voltages. My rough guess is that operating voltage is roughly proportional to feature size. Therefore, since a roughly square function beats a function roughly to the power of one, it seems likely that electromigration should be "less" of a problem as chips are shrunk to ever smaller sizes—not more of one. (Not at all! Electromigration (E/M) is proportional to (current density in aluminum paths)N and to (temperature above 100°C)M. It may be proportional to the product of those two terms, where N and M are at least 2, or 3, or 4—or maybe more. NSC's parts are designed very conservatively for E/M. Our QA guy told me we get less than one complaint per year about E/M. /rap)

Newer, faster CPUs seem more sensitive to ESD events than older ones. So, perhaps ESD is the culprit for any finding of shortened lifetimes. (Nope. Next time you wander into a computer store, ask the clerk how fast the 2300-MHz processor is expected to wear out due to electromigration. If you plan on living in, and computing in, an igloo, you might be quite pleased with the answer. If not, probably not. /rap) By the way, I am a member of the mountain-bike-to-work crowd.

Alastair Roxburgh (via e-mail)
Pease: Very good, but at 40 miles out, I bicycle to work very rarely. Sometimes I ride my bike down to the train and ride that to work. That is easy—five miles down, and 330 feet down the hill to Caltrain—about 21 minutes. But riding back up that hill is rather slower, like 45 minutes. Naw, as long as my 1970 Beetle is running, it is cost-effective.

Dear Bob: Regarding your reply to Lawrence King in "Bob's Mailbox" (Electronic Design, June 23, p. 20): You had relays? We only dreamed of relays. All we had were these stones marked with 0 on one side and 1 on the other. We flipped them over manually to perform calculations in binary arithmetic. Yep, we were real familiar with the original Latin meaning of calculus.

Bill Frank (via e-mail)
Pease: Thanks, Bill. I love it!

Comments invited! [email protected] —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

Hide 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.