Electronic Design

Bob's Mailbox

Dear Bob:
I don't usually react too much to your column other than to enjoy it, but your "Speaker Cable Stuff" for May 13 really hit a nerve. For many years I have read the papers on these lunatic fringe hi-fi audio theories and put as much credence in them as in UFOs. Come to think of it, there are UFOs, they're just unidentified.

In the 60s, I read an article that claimed that if your stereo preamp wasn't flat from dc to 50 MHz (with an M), then you would hear unacceptable "transient intermodulation distortion."

But back to the speaker cables! If they are 20 feet long and the wrong impedance, they are also 20 ns long and might ring at 25 MHz. Big deal! But, of course, if you have some of George Kaye's "solder oxide" junction diodes, they would likely rectify the 25 MHz and create havoc. I can't see how the cable reactance could possibly affect anything. But the resistance is another story if the speakers are 1-2-4 ohms as Kaye implies, and those fancy cables are humongous. Trying to put 800 W into your $3 lamp cord may hurt your dynamic range; you may only get 700-W peaks and, perhaps, George Kaye can hear that difference.

Now, if you invest in two lamp cords (we're up to $6 now) and Kelvin-connect the speakers to the Amp Out and the Feedback, want to make any guesses as to the result? The amplifier is now tracking the speaker cone instead of the zip cords. Time to melt down those silver wires. Oxygen-free hard copper may have lower resistivity than copperus commonus, but I prefer the Kelvin solution.

Does it ever occur to anyone that the weak links in the system are the electromechanical transducers at each end. I have never seen distortion figures for microphones or speakers that come close to the levels achieved by electronics. For the average Joe who doesn't own a pair of exponential horns, the tiny shoebox speakers are common, but they have to be nonLinear because of pressure buildup in the sealed box. To overcome the cutoff frequency of the speaker cone and get some bass out of your 4-in. sub contra bass woofers, the designers pack them into a small box and kill the midrange response until it matches the terrible bass response to achieve flatness. This is why we need these megawatt amplifiers.

I used to have a Goodmans AXIOM 80 speaker with a 17.5-kilogauss magnet and a 20-Hz fundamental resonant frequency in a 2.5-cubic-ft. reflex box with an acoustic resistance load. It took only 4 good watts to level my living-room walls.

As for the valve amplifiers, I know they put out a warmer sound because I used to burn my hands on them. Because so much depends on what happens to the power supply and amplifier bias under overload conditions, any sound difference between tubes and transistors can hardly be ascribed to tubes.

Mr. Kaye doesn't like a 10-kHz waveform with only 4 samples, but that is well above the top note on an organ with a 2-ft. stop, let alone the normal 8-ft. stop. The first overtone is at 20 kHz, and very few people can hear that even by itself, let alone with all the other music. With 4 samples, even the harmonic can be reproduced faithfully.

Remember how all hi-fi preamps used to have a 47-kW input resistor for the magnetic cartridge? Designers used negative voltage feedback to make their transistor-amp input impedance extremely high like a tube, and then used a 47-kW resistor across the input. I never could see the sense in burning up the transducer power in a resistor and then trying to amplify the dregs at low noise. Seems to me you would make the transistor's base look like 47 kW by design, and then stuff all the signal power into the transistor instead of wasting it. We had a simple design for such a beast at Tektronix in the early 60s, and we never had any audible preamp noise. We even used the cartridge inductance to make one of the RIAA corners.

By the way, my interest in audio and music is entirely in the way it sounds, not in what a spec sheet says, and my expectation standards are extreme. But unfortunately for most of us, the available software (recordings) are rather pitiful.

My LPs all have two defects: they are never flat, and some dolt always bores the hole off center. I expect that CDs have overcome these problems, but finding well-done, digitally-mastered recordings of quality performances of interesting music is a tough chore. There still is not a digital recording of "The Ring of the Nibelung" that can hold a candle to the Decca/Solti analog recording from the early 60s.

The real contribution made by the audio engineering fraternity is to provide some very good sound for the average person in his studio apartment at a reasonable price. The trouble with performances on high-end audio gear is that they are usually in public environments and are always marred by coughing, talking, chocolate wrappers, and fire sirens, just like live performances. Wanna split those $900 wires-excuse me-cables with me Bob?
Hewlett-Packard Co.
Optical Components Div.
San Jose, Calif.

I'm sure that there are many audible differences in speakers; everybody to his own tastes. I'm sure that the new, improved speakers that critics raved about, 10 or 20 years ago, now are considered dreadful. My old speakers, on the other hand, sound just fine.

The concept of Kelvin connections causing the power amplifier to force the true response at the speaker sounds very appealing, but it also sounds scary because adding the inductance and delay inside the loop might make the loop very hard to stabilize. However, if you used a low-Z 8-W cable to drive an 8-W speaker, there is zero effective inductance and not much delay. Therefore the loop closure may not be so difficult. Still, isn't it much easier to bring the power amplifiers over by the speakers? Or just use low-Z wires? After all, can anybody hear the difference?-RAP

Dear Bob:
....I bought an "Easy" development kit consisting of 436 files in 416 Mbytes-and it had hardly any documentation. After four weeks I still don't know how to start to use it. But I learned how to change the color of the frames, the font of the letters, and how to change the icon from a garbage can to a geranium. A skunk would be more appropriate.

I was looking for a PLL-clock-generator chip to generate the number to be loaded into three programmable counters. They supply you with 12 programs in 1.1 Mbytes.

The best text editor I ever used was 48 kbytes. It did the same job better-and much more-than the one which is 1000 times larger today.

In spite of the man-centuries of development, I still would like to see a software package without a bug. But at least they now come in a variety of colors and shapes. Well, this is progress, or is it?
PC Peripherals Inc.
Broomfield, Colo.

Nick-my Word Processor uses just 75 kbytes, and it's great (PC Write Lite). I, too, am horrified by the gross excesses of "modern computing," and all their mind-boggling multimedia, and multicolored bugs.... As Yogi Berra said-"Progress may have been alright once, but it just went on too long...." Myself, I am usually NOT required to junk my OLD computers and software.-RAP

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

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.