I have owned a Teac A-360S for months and have had no
difficulty with playback or recording except with some recordings made
on LH Super 90 BASF cassettes. Each of these cassettes now has a high-pitched "scratch" superimposed
on the music. The scratch, however, is present only when more than half of
the tape is played and is independent of head condition--I have noted the noise
is present even after cleaning and de-magnetizing the heads. A friend, who
also has a Teac, has experienced the same difficulty. The scratch is not present
on any other recordings made with Maxell, Scotch, and TDK tapes.
It seems that the difficulty lies in the BASF cassettes, not my tape deck. Have you ever heard of this problem before? What can I do to salvage some of my favorite recordings?
Sidney D. Machefsky, Nashville, Tenn.
We have not been able to duplicate your problem with LH Super (which, incidentally, is no longer in the BASF line), but it sounds like a variety of so-called scrape flutter, which can be influenced by changing tape tension on the supply side as the size of the "pancake" decreases. Scrape flutter can pro duce extraneous noises during recording, in which case the only remedy is filtering in playback. If the noise is produced only in playback, however, it may disappear if you play the cassette on another deck. If it does, you can then copy the music onto a cassette that you know to work well in the Teac.
In your June 1976 issue a reader complained of excessive cone vibration in his speakers. You advised either the use of a subsonic filter or backing off on the bass. (He was a "bass freak.") I have the same problem. My twelve-inch Lafayettes vibrate so much as to produce distortion. Cutting back on the bass does not help much. And, besides, I like bass. When I play FM on my receiver (Technics SA-5200A), I get perfect reproduction. But when I put on my turntable (Pioneer PL-12D ii), I get this fluttering distortion and the cones go crazy. If it's the turntable, would a subsonic filter help? Does it (audibly) affect the sound (dampen it)? Where could I get one? Would I be able to push my bass up without fear of distortion if using one of these? Perhaps I should send these speakers back while they're in warranty.
-Gary Grebstein, Binghamton, N.Y.
Perhaps we did not make this point clear in June: There is nothing wrong with the speakers. They are simply trying to reproduce rumble that originates elsewhere-probably, in your case, in the turntable or as feedback between the Lafayettes and the pickup. The Pioneer PL-12D II has a rated signal-to-noise ratio on the order of 48 dB, a figure that is acceptable in a low-priced turntable but far from outstanding. The newer (just under $200) PL-510, in contrast, is rated at 60 dB. (CBS labs measured 63 dB with ARLL weighting.) You can get good results from a unit like the PL-12D II as long as your speakers do not have extended bass response (which your Lafayette speakers seem to have). The first thing to try is greater vibration isolation for the turntable. There are a number of products marketed for this purpose, though foam or fiber pads such as are used under office machines often are effective and cheaper. It that doesn't help-or doesn't help enough the next solution is a new and quieter turn table. If that doesn't work completely, a low cut filter (available, as far as we can tell, only as part of a receiver or preamp) is the next thing to try. Using the filter may involve a small sacrifice in bass response, but this is minimal if the filter has a steep enough roll-off (say 12 or 18 dB per octave) and a low enough turnover point (around 30 Hz).
In "What You Should Know Before Buying a Speaker" [October], you stated that a 6-dB-per-octave crossover introduces no phase distortion and that 12- and 18-dB crossover net works will produce phase distortion. After reading Passive Audio Network Design by Howard M. Tremain, I find that to be incorrect. A 6-dB network is 90 degrees out of phase at the crossover frequency. Although you can cure this, it often is impractical to do so because it requires physically separating the transducers by one-quarter the wavelength (of the crossover frequency) in depth. The 18-dB crossover (M-derived type) is approximately 221 degrees out of phase. This can be cured also, but it requires more time and patience to position the drivers correctly. The 12-dB crossover is 180 degrees out of phase. This is corrected by switching the leads to the one driver.
I think Bang & Olufsen is trying to be deceptive, because changing the polarity to the speaker couldn't affect its rolloff characteristics. Even if it did, an "audible suckout" couldn't exist, because the speaker transition would occur 6 dB down. (The crossover frequency would change.) At that point the combined acoustic output would be 3 dB louder, and the total drop-off would be only 3 dB down, which is truly a big difference from the audible suckout that they picture.
-Greg Aiken; DelMar, Calif.
Bang & Olufsen may have confused you. but the statements made by this company concerning the phase response of crossover net works can in no way be called deceptive. If we assume that we are working with drivers whose acoustic output is a perfect replica of their electrical input, we can proceed by looking at the outputs of the sections of the net work. At the crossover frequency of a first-order (6 dB per octave) network, the high- and low-pass sections are down by 3 dB; that is, the voltage output of each is 0.707 of that de livered in the passband. Also at this frequency, the output of the low-pass section is shifted in phase by -45 degrees, that of the high-pass by +45 degrees-adding up to a 90-degree phase difference between the two. Note, however, that the sum of the voltage outputs (0.707 + 0.707) is greater than 1. The response peak that this would imply does not materialize because the magnitudes of waves that are 90 degrees out of phase must be added via the Pythagorean theorem. Therefore we have (0.707)^2 + (0.707)^2, which adds to 1. Since the square root of 1 is 1, the output of the two networks is correct in amplitude. And since we have dealt with the phase shift, is also correct in phase. Looking at this from another view point, there is just enough "excess" output from the two sections of this network to make up for the phase shift.
We have assumed tacitly that the two drivers occupy the same position in space. It is the impossibility of meeting this requirement that makes the positioning of B&0 speakers for any listening location fairly critical.
With a second-order (12 dB per octave) network, the phase shift increases to 180 degrees between the two sections, which means that the outputs exactly cancel and leave a total of zero-a suckout, if you will.
While this can be "corrected" by reversing the leads for one driver and causing the out puts to add, the net result is a 3-dB peak that in and of itself implies a phase shift in the total output.
For higher-order networks (18 dB or more per octave) the situation is more complex and the analysis more tedious. But it can be shown that the total output will not be correct in phase however the drivers are connected and, further, that the associated phase shifts may exceed 360 degrees. Positioning drivers in space so as to compensate for a crossover's phase deficiencies is obviously not viable, as phase (time) response is then incorrect for all frequencies away from the crossover point.
I have sent my solid-state preamplifier for re pairs, and while waiting I called on my old tubed version of the same preamplifier. I found that in general there is little difference in sound, but the tubed version reproduces more realistic string sound than the solid state version does. I see no advantage to the solid-state instruments. Besides, the tubed preamp can be repaired locally; not so the solid-state. How about a 200-watt triode amplifier? I will buy one at once if somebody makes one.
-David Fonseca, East Ridge, Tenn.
An output of 200 watts per channel translates to 23 dBW. If you can settle for just a hair less, the Audio Research D-150 at 22 dBW (150 watts) per channel comes to mind.
Another possibility would be a pair of Dynaco Mark VI mono power amps, which will produce 21 dBW (120 watts) each. Our files don't reveal whether the output devices in either of these are in fact triodes, but we know of no other tubed amps designed for such high output power. We, by the way, do find advantages to solid-state circuitry.
-------------
(High Fidelity, Jan. 1977)
Also see: