Electronic Design

Bob's Mailbox

Dear Bob:
I worked in the music recording industry for a while. The recording studios in the U.K. in the '70s wired their 1-kW amps to the speakers with 30-A, single-strand cooker wire. Sounded okay to me, but a bit loud.
Allan Hurst
via e-mail

Hello, Alan. It's easy to compute that. If you don't mind poor (high) inductance, you can use these big, fat wires. But they're awfully hard to manipulate—bulky, lumpy, stiff, awkward. Plus, low-inductance wires will give less high-frequency phase shift and less skin effect—not that anybody can hear that. If you're going to drive 1000 W into 4-Ω speakers, it would probably make sense to use heavy wires to avoid excessive heating. Hey, 15 A is a lot. I'd guess that 12- or 14-gauge ordinary lamp cord would be less awkward with less attenuation for long wires, compared to 16 or 18 gauge. But just think about this: pick any speaker wires you want, and nobody can hear any difference anyway! Thanks for writing.—RAP

Allan Hurst replies:
Bob, you're right. The wire was fat and cumbersome. But it was installed just like for a cooker—in the plaster of the walls. After all, studios weren't portable installations! I agree with you. I bet nobody could tell the difference between expensive, fancy Litz wire and cooker cable, given that both were fat enough to carry the current without undue loss.

P.S. There was a great fuss at the time about the phase shift introduced by transformers. The firm I worked for, Neve, decided to see if the problem was audible. You could certainly measure significant phase shift through a good audio transformer. So we rigged up a test jig in a posh London studio with soft-switchable amounts of shift at 3, 5, 10 kHz, etc., and ran blind tests with the local "golden ears." The results were entirely random, as we expected. I don't think that people can hear phase shift as such at high frequencies, especially because a full 360° shift at 10 kHz corresponds to moving your head about 1 inch! Most stereo imaging is down to your brain being pretty good at using the difference in time-of-arrival of transients at the two ears, to a lesser extent, amplitude, and much less again, phase (if at all), except at low frequencies when the wavelengths are big enough for coherent cancellations at both ears.

I gave up on the industry after a while, as it was impossible to correlate what the "golden ears" who bought the stuff said they wanted with any sensible engineering—though they generally had engineering managers who were a bit more sensible. I hear that the HiFi trade is now about 100 times worse.

I've been debating the evils of "objectivism" versus "subjectivism" with some guys. Some of the fantasies of the golden-ear guys are quite amusing, except that they start taking themselves seriously.—RAP

Re: Tee networks. When I sent you the e-mail that you published in your column (Electronic Design, Aug. 20, p. 90), I thought you would realize it was a joke and get a laugh. Maybe I could create an urban myth for electrical engineers: Bob Pease had discovered a way to tell the difference between Delta and Tee networks, other than the parasitics. I thought the excited tone of the first sentence of the second paragraph—and its exclamation point—would tip you off. But I guess I didn't have my tongue far enough in my cheek. Perhaps you should mention that it was a joke in a subsequent column.
Rodger Rosenbaum
via e-mail

No, I did not realize that it was a joke! Every week I get a letter, or e-mail, from some very serious guy arguing that I'm obviously wrong about something I said in a column, which is a bit annoying when I know that I'm right. I get responses from guys who think that I'm wrong about Taguchi Methods. I'm contacted by readers who think I'm wrong about Fuzzy Logic. I receive letters from guys who think they have discovered perpetual motion with output power bigger than the input. I even get mail from guys who think that I'm wrong about Tee networks! So I'm sorry to be so slow, but I did miss the joke!—RAP

Dear Robert:
I believe that in one of the past 12 issues of Electronic Design, there was a reference to a "socket/adapter" for the SOT23 family of ICs. I have searched the site with zero success. Can you help? Thanks!
Ron Raspet
via e-mail

Hello, Ron. According to NSC's Emmy Smaragda Denton, "You can find them at Fry's and in the Digi-Key catalogue. We used to send out samples of the LM45/50, etc., on a little conversion board, but we no longer do that. You can also order an LM62 demo board. The LM62 comes in a three-lead SOT23 package. URL: www.national.com/store/view_item/index.html?nsid=LM62EVAL. It costs $4.00 from us. But if you buy the parts from Digi-Key and solder them up yourself, it would probably be cheaper."

Emmy is right. I grabbed a May-August catalog for Digi-Key, and you can buy these as a "Surfboard" at about $1.50 each, for the five-pin SOT23-5, in quantities of 10. The six-pin version is a few pennies more, but it's more versatile. These are sold by Capital Advanced Technologies. I bet you can even search and find this on the Digi-Key Web site, www.digikey.com, now that we have proven that they exist. Look for Digi-Key part number 33205CA-ND. Best regards.—RAP

All for now. / Comments invited!
RAP / Robert A. Pease / Engineer
[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.