New Generation of Power Amps (Vol.2, No.2: 1979)

The New Generation of Power Amplifiers: Amber Series 70; Audionics BA-150; Audire 'Crescendo'; Audire DM700; Bedini Model 25/25 ; Bedini Model 45/45; Hafler DH-200 (follow-up); JVC M-7050; PS Model One; Rappaport AMP-1 (epitaph); Sonotron PA-2000.

It doesn't look as if anybody really knew what makes a power amplifier perfect, but through sheer tenacity a few designers are getting within striking distance.

Now hear this. Forget about THD measurements. For get about SMPTE IM. Forget about CCIF IM. Forget about slew rate. Forget about bandwidth. Forget about square wave response, into any kind of load, unless it's truly horrendous.

Forget about 'black box' tests (comparing the output against the input) in general. They just won't tell you what you want to know about power amplifiers. Sure, they'll identify specific malfunctions and crude design errors; they'll separate the dogs from the half decent ones. But if you're after the best sound that money can buy-or the next best sound that a lot less money can buy-forget it.

As you know, we've had our reservations about black box tests for several issues now; with this latest batch of power amps we've reached a firm conclusion. There just isn't any sonic correlation. We still put all power amps through our routine bench tests, mainly to get some idea of the design philosophy behind them and to eliminate the possibility of attributing audible effects to nonexistent technical causes, in the manner of certain underground reviewers. But we certainly don't claim at this point that we can predict the sound of a power amplifier from its measurements, even approximately. If nothing else, the Bedini Model 25/25 (see review below) has cured us of that illusion. All our evaluations are therefore based on how the amplifier sounds when inserted into our 'Reference A' and 'Reference B' systems.

Feedback falls into further disrepute.

Shortly before press time we received a devastating document that appears to confirm our worst doubts about standard methods of amplifier measurement, not to mention amplifier design in general. It's the preprint of a technical paper to be delivered by Dr. Matti Otala on February 25, 1980, at the 65th Convention of the Audio Engineering Society in London.

The paper presents rigorous mathematical proof, for the most generalized, all-inclusive case, that feedback can not make amplifier distortions go away; all it can do is to change one kind of distortion into another. By the application of feedback, the amplitude nonlinearities of the open loop are converted into phase nonlinearities of the closed loop. That's all. The garbage cannot, by definition, be made to disappear; it's simply swept into another corner. In the typical feedback amplifier, the amplitude of the audio signal phase-modulates the high-frequency components of the signal. Furthermore, any amplitude intermodulation distortion in the open loop is converted into phase intermodulation distortion in the closed loop. What about TIM, alias SID? It turns out that it (he?) isa limit case of this feedback-generated phase modulation effect, with all shades of gray possible before the actual black eruption occurs. None of this shows up on standard tests.

Scary, isn't it? Dr. Otala promises further papers on the audibility of the effect and on a method to measure it, adding that it seems to be particularly annoying in amplifiers that use high values of feedback to suppress crossover (notch) distortion.

This is no small matter; once the analysis and the conclusions are familiar to the audio engineering community, amplifier design will never be the same again. We wonder how prompt the Audio Engineering Society will be to publish the paper in the Journal, since there are many AES members with a vested interest in high feedback and triple-oh amplitude distortion figures. We trust that scientific objectivity will prevail. Meanwhile we feel justified in many of our previously expressed but not nearly as rigorously derived opinions and relieved that the best minds in the business are also having trouble finding measurement methods that correlate with listening.

Amber Series 70

Amber Electronics, Inc., 917-B Preston Avenue, Charlottesville, VA 22901. Series 70 Power Amplifier, $459.95. Three-year warranty. Tested #700100 (also two earlier, not yet optimized samples), on loan from manufacturer.

There isn't much we can say about this solidly built, sensibly designed, conservatively operated 70/70-watt power amplifier, except that we discern in it no obvious errors of concept or execution and that it sounds better to our ears than any other amplifier in its class, including the Hafler DH-200 and the PS Model One, though not by a wide margin.

You may think that's a pretty heavy statement, but it isn't really, not in a product category where things are changing very fast these days and mostly for the better, and especially not since the Bedini Model 25/25, for only $190 more, offers considerably superior sound, though not as much power. Furthermore, our ranking of the Amber applies only to the very latest production version, incorporating certain circuit modifications without which the amplifier doesn't sound nearly as good, and then only when its speaker fuses are changed to 5 amps or heavier. (See also the preamble to the speaker reviews in this issue.) Thus modified and fused, the Amber sound very smooth, open, neutral and clearly detailed, with a good, solid bottom. By comparison, even the best sample of the Hafler is a little zingier on top and the PS perhaps a wee bit coarser in grain-but just. On balance, the Amber is our new power amplifier choice for ''Reference One feature that makes the Amber even more desirable and certainly more versatile is the built-in capability to be strapped for mono operation. At the flick of a switch, you have a 200-watt (into 8 ohms) mono amplifier that sounds just as good as the stereo channels, even though the strapping is accomplished in a theoretically less pure way than the classic balanced bridge connection, which requires accurate phase inversion.

How do we account for the good showing of this first product of a new and little-known company? It seems that they have their act pretty well together. Even in physical appearance, the Amber looks much more finished and professional than what we expected. Really nice. The power supply is also considerably more generous than what is generally seen in this price and power category; the audio signal path appears to be simple, straightforward and trouble free. Perhaps basic good sense, no penny-pinching, and the avoidance of vulgar design errors will get you further in product development than half-baked originality.

Audionics BA-150

Audionics, Inc., Suite 160, 10950 SW 5th Avenue, Beaverton, OR 97005. BA-150 Analog-Digital Stereo Power Amplifier, $2950. Three-year warranty. Tested #10163, on loan from manufacturer.

This one is a whole bucketful of paradoxes. First of all, in immediate contradiction of our generalizations above on amplifier measurements, black-box testing does offer some clue to the sound quality of the BA-150. Secondly, its enormous price tag, although promising more than what the amplifier delivers in audible benefits, does have some justification in the beautiful and thoroughly professional construction. Thirdly, the design looks nostalgically backwards, to the good old days of vacuum tubes and output transformers, and at the same time aggressively forward. to bias and balance regulation with CMOS digital logic. Finally, with three different choices of feedback and three output impedance connections, the BA-150 has nine different kinds of sound-but we're getting ahead of ourselves.

Designed by David Berning of the National Bureau of Standards (as a private project, of course, not for the NBS), the BA-150 is a hybrid design, with a pair of 6LF6 tubes in a new Class B configuration driving each speaker through an output transformer, and with both tubes and solid-state de vices used in other parts of the circuit. The aforementioned CMOS digital logic circuitry continuously monitors the operating characteristics of the output tubes and automatically adjusts biasing for optimum performance. Either 150 watts per channel or 40 watts per channel (for speaker protection) can be chosen by means of a front-panel switch; seven different colored lights provide information at a glance as to what the various parts of the amplifier are doing. It's a neat and reassuring package.

The output transformers are tapped for 4, 8 and 16-ohm loads; in our opinion, a combination of series and parallel strappings, utilizing all windings of the secondary under all loading conditions, would have been a more elegant and better optimized solution. (Remember the old Partridge out put transformer? You're no spring chicken if you do.) An other aspect of the design that we aren't convinced about is the feedback selector switch, which allows the user to decide whether to apply 0, 8 or 14 dB of feedback around the open loop. After three minutes of listening, it becomes quite apparent that 8 dB is the right figure, with just about any speaker load, neither 0 dB nor 14 dB giving anything even approaching the same clarity and freedom from colorations and/or stress. Obviously, there are audible amplitude distortions with no feedback and audible time-domain problems with too much feedback, leaving 8 dB as the right way to split the difference. It should have been the built-in design choice, without giving anyone the opportunity to mess it up. With the three positions of the feedback control and the three output taps, there are nine possible combinations, each of which sounds totally different from the other eight. The cleanest, most transparent-sounding combination is probably 16 ohms out with 8 dB of feedback-but what if you're driving speakers that drop to 3 or even 2 ohms somewhere in their range, as so many do? At its best, the sound of the BA-150 is amazingly sweet, smooth and edgeless, often in dramatic contrast to that of the transistor amplifier used in a particular A-B comparison. The initial listener reaction is always very positive; only after a few minutes of exposure does it become apparent that there's always a certain amount of sonic mud, thickness and veiling of detail. Just this once, we believe we can offer some laboratory data to corroborate our ears. For one thing, the amplifier produces a tremendous amount of IM distortion.

With 0 dB feedback, the SMPTE figures are ridiculous: 5.4% at 30 watts into 8 ohms; over 10% at 100 watts into 8 ohms.

With 8 dB of feedback, these figures drop to 2.35% and 4.2%, respectively. Into 16 ohms, we measured a halfway respectable 1% at 30 watts (again with 8 dB of feedback):

the question is, where does the relative forgivingness of the human ear to amplitude distortions reach its limit? These numbers are too large for comfort. Furthermore, the amplifier is very, very slow. Rise time with an 8-ohm resistive load is between 5.5 and 7 microseconds, depending on the setting of the feedback selector; with capacitive loads, considerably worse low-pass effects can be observed, greatly varying with the combinations of transformer taps and feedback settings.

In a worst-case situation, still using real-world capacitive loads as encountered in electrostatic speakers, square waves change into virtual sine waves. This is the first power amplifier with which we hesitate to recommend the use of the Cotter NFB-2 noise filter/buffer, since the rms sum of the rise times of all the signal-path components in series might then exceed the threshold of hearing, so that the slowing down of the signal would be definitely audible. Unfortunately, this filtering effect of the BA-150 takes place at the output, not the input, and thus the benefits of the Cotter filter aren't necessarily duplicated. What all this boils down to is that lots of IM distortion combined with borderline speed on transients may very well be responsible for the sonic personality of the amplifier.

Even so, we're convinced that there exist installations in which the power, rugged construction, virtually certain reliability and velvety sound of the BA-150 will make it the right choice for a particular user, high price notwithstanding.

We continue to prefer, however, a few solid-state power amplifiers that offer comparable smoothness in combination with greater clarity and better resolution of inner detail.

Audire 'Crescendo'

Audire, Inc., 18474-E Amistad Street, Fountain Valley, CA 92708. 'Crescendo' Power Amplifier, $350. Three-year warranty. Tested #5070 and #5129, on loan from manufacturer.

For $350, this surprisingly well-built 60/60-watt power amp gives you a midrange that's close to SOTA in openness and transparency. In fact, when we first heard it, we were reminded of the midrange quality of the Rappaport AMP-1. At the same time, the 'Crescendo' sounds quite hard and edgy on top, though a little less so in the second sample we auditioned, which had been a bit more carefully adjusted for correct bias as a result of our complaints about the first. Even this less irritating sample, however, was clearly not as smooth and neutral in the upper frequencies as the Hafler DH-200, which itself is a small step or two from the ultimate in this respect.

On the laboratory bench we observed some square wave anomalies even with a purely resistive load, not to mention quite a bit of overshoot and ringing when capacitance was added, but we no longer set much store by these tests. Most probably the cutting edge we hear has to do with the operating characteristics chosen for the various devices in the circuit, hence the variability with small bias adjustments.

One nice feature at this price is the very graphic power output indicator; it flashes twelve LED's per channel, in green, yellow and red. We could watch it for hours, but we'd rather listen to the Amber, PS or Hafler, all of which admittedly cost a little more. We call this one a near miss.

Audire DM-700

Audire, Inc., 18474-E Amistad Street, Fountain Valley, CA 92708. DM-700 Power Amplifier, $1400. Three-year warranty. Tested #7078, on loan from manufacturer.

We expected a lot from this one, the flagship of the Audire line, rated at 350 watts per channel into 8 ohms (that's right, three-five-oh, or 500 watts into 4 ohms) and presumably designed with the same understanding of what a transparent midrange is all about as the 'Cresendo' reviewed above. As we learned, it does share the latter's midrange qualities (excellent!) but also its hard, over-bright character further up in the audio range. We found our sample of the DM700 very fatiguing to listen to for any extended length of time; the zinginess was much too aggressive.

After a while we began to suspect incorrect bias set tings, as in our first sample of the Crescendo, so we made an attempt to readjust the bias. What we didn't know was that the bias pots in the DM700 don't just operate within the allowable bias range but can actually be rotated all the way to zero resistance. (This saves the price of a protective resistor in series with each pot.) As a result we zapped one of the boards with too much current, and it went up in smoke. End of our tests, as El Cheapo strikes again.

Square wave tests, just before this fiasco, revealed some of the same strangenesses we had observed in the Crescendo. No big deal. It should be noted that the huge output of the DM700 is obtained by means of a bridge circuit, which in some ways is also a money-saving gimmick (four times the power into the same load by just doubling the number of output devices, at the expense of various other important design considerations); nevertheless, we could certainly use and would welcome this kind of power any way we can get it, as long as the sound remains sweet and smooth.

One thing the DM 700 does right is the use of entirely separate power supplies for each channel; it's an amplifier with considerable appeal and potential at this price, one of those we would like to like better. A modified sample has been promised to us and is supposed to be on its way as we go to press; look for a follow-up review in the next issue.

Bedini Model 25/25

Bedini Electronics, 13000 San Fernando Road, #9, Sylmar, CA 91342. Model 25/25 Class A Power Amplifier, 3650. Tested #250029, on loan from manufacturer.

Here's the sleeper of the year: a 25/25-watt (14 volts out) little Class A unit that performs most unimpressively on the lab bench but sonically wipes out all other amplifiers known to us-in transparency, depth of perspective, definition of inner detail and smoothness of highs! Its only rivals are the now defunct Rappaport AMP-1 and a last-minute arrival, the JVC M-7050, both of which are in a totally different power and price category. We said rivals, mind you, not equals. The whole thing is rather hard to believe.

John Bedini claims that his circuit is *'pure Class A" rather than ''sliding Class A'' (a la Threshold, etc.); we aren't sure whether his definition of *'pure'' would be accepted by the designer of the Mark Levinson ML-2 Class A amplifier, for example, but that's beside the point. Technical hairsplitting isn't the long suit of the Model 25/25. In fact, hairsplitting laboratory tests will reveal that it suffers from rather severe slew rate limiting; we strongly recommend the use of the Cotter NFB-2 noise filter/buffer with this amplifier to prevent it from getting zapped with superfast rise times devoid of auditory information. It will function a lot more happily that way. But that's not all. We could hardly believe our eyes when we passed square waves through the Model 25/25 into an 8-ohm resistor, with an additional 2-microfarad capacitor connected directly across the output terminals. The amplifier went into such severe oscillation that it almost immediately blew both 8-amp fuses in its power supply.

Mamma mia! Since this test load is a fairly accurate model of what is presented to the amplifier by certain electrostatic speakers, the relatively small series inductance of a length of speaker wire acquires crucial importance as protective isolation in an actual music system utilizing the Bedini. It has no equivalent built-in protection. (See also the speaker wire article in this issue. Incidentally, nothing makes the highly capacitive Quad electrostatic sound quite as perfect as the Model 25/25-with a bit of wire in between, of course.) How come this test-bench turkey sounds so beautiful in the real world? Ah, if we knew that, we could devise real world oriented bench tests in which the Bedini would obviously not come out a turkey. Our best guess is that all the solid-state devices used in the amplifier are deliriously happy at their particular operating voltages, currents and temperatures; those big output transistors are just coasting; nothing is stressed beyond the most comfortable linearity; and whatever little glitches exist are amplitude modulations rather than time modulations. At least with the Cotter filter at the input and a nice length of mildly inductive speaker wire at the output, right? In addition, the amplifier can put out a fair amount of current, even though it's limited in voltage output, so it sounds more powerful driving complex speaker loads than a typical 25-watter. And that's the best we can do on that one, folks.

Anyway, while trying to figure out what makes the Bedini Model 25/25 tick, we're shamelessly enjoying it in our ''Reference A' system, replaced by the JVC M-7050 only when we want to play something very, very loud.

Bedini Model 45/45

Bedini Electronics, 13000 San Fernando Road, #9, Sylmar, CA 91342.Model 45/45 Class A Power Amplifier, $1300. Tested #450047 (AC coupled) and #450053 ( DC coupled), on loan from manufacturer.

This is supposed to be a scaled-up version of the Model 25/25, with everything essentially the same except the bigger power supply. Well, there's one other thing that isn't the same in our opinion, and that's the sound. The Model 45/45 isn't even unequivocally superior to a good sample of the Hafler DH-200, at one third the price, let alone the smaller Bedini or the JVC. Where the Model 25/25 is utterly smooth and edgeless, the 45/45 exhibits that characteristic little transistory zing and hardening, and its midrange transparency and delineation of high-frequency detail are merely good, not great.

Obviously, there's something going on in the circuit that even John Bedini hasn't quite got a handle on at this point.

We do hope he gets there before his coming 200/200-watt model is finalized; with that kind of power and the sound of the little amplifier, we could all just wash up and go home. A 100/100-watt version has been heard by a number of people whose ears we trust, and they report that it isn't quite in the same class sonically with the 25/25.

Very early samples of the Model 45/45 had huge output capacitors (a la Futterman); we see nothing wrong with such AC coupling, but the current production model has been changed to the DC-coupled configuration of the entire line.

We could hear no significant difference between the two versions. The test-bench behavior of the 45/45 is very similar to that of the 25/25, and you already know how relevant that turned out to be. Wouldn't it be nice if we all knew as much about measuring amplifiers as we did only a few years ago?

Hafler DH-200 (follow-up)

The David Hafler Company, 5910 Crescent Boulevard, Pennsauken, NJ 08109. Model DH-200 Stereo Power Amplifier, $429.95 wired.

(In kit form, $329.95.) One-year warranty. Tested successive production samples and modifications, all on loan from manufacturer.

Our enthusiastic recommendation of this most attractively priced 100/100-watt MOS FET power amplifier in the last issue has created something of a credibility problem for us.

Early production units turned out to be not nearly as excellent as the preproduction sample we had tested in good faith, having been solemnly assured by the manufacturer that the final production version would be electrically and sonically indistinguishable. Lets straighten out the record by tracking through the entire sequence of events following our review.

The first thing that happened was that the designer of DH-200, Erno Borbely, left The David Hafler Company to become head of the European training program of National Semiconductor Corp. Ed Gately, Hafler's president and an excellent engineer in his own right, was left with the problem of following through on the DH-200 project, without being on intimate terms with every minor component in the circuit.

Somehow or other, perhaps as an untested afterthought by the departing designer or on someone else's instructions, a high-beta transistor was substituted in some production runs for the low-beta device used in certain low-level stages of the circuit. These DH-200's, all of them having serial numbers below 3939000, sounded hard, zingy and dimensionless.

The low-beta transistor was then reinstated in all applicable positions. Later it was discovered that certain transistors had to be more carefully matched for best results; still later it was decided that certain capacitors had to be replaced in order to eliminate some remaining vestiges of zinginess. By the time we tested production samples in the 3946000 series, some were already better than our original review sample, though not all. Finally we got a modified sample that looked forward a couple of months; it was supposed to incorporate all the minor capacitor changes and other little tweaks that would be in production early in 1980. This one was definitely cleaner and sweeter than any previous version we had tested, including the preproduction one. Meanwhile, however, we still kept hearing about the variable quality of the DH-200's being received by dealers.

If the Hafler were still out top choice in the “Reference B'' category, we would probably explore the situation still further. Our latest listening comparisons show, however, that both the Amber Series 70 and the PS Model One are marginally superior even to this ' 'final' version of the DH-200.

The latter still has a tiny bit of that treble shimmer or zing we pointed out in our original review; the Amber and the PS are both a shade better in this respect and also have a more solid bottom end, as well as a slightly more transparent midrange.

Since they all cost about the same (the DH-200 has gone up in price by $30), there's not much point in agonizing over the possible variability that may or may not exist in the latest production runs of the Hafler. If you happen to own a DH-200 that isn't quite up to snuff, you probably aren't just imagining things and should holler until your dealer or Hafler fixes it free of charge.

None of this means that the DH-200 isn't still an outstandingly fine audio component; in fact, we're convinced that, other things being equal, a MOS-FET power amplifier is inherently superior to one that uses bipolar output transistors. But other things are seldom equal.

JVC M-7050

US JVC Corp., 58-75 Queens Midtown Expressway, Maspeth, NY 11378. M-7050 Stereo Power Amplifier, $1500. Two-year warranty. Tested #13400021, on loan from manufacturer.

This one came in so close to press time that we didn't even have time to put it on the lab bench. All we could do was to have it in our reference system for a week or so, listen to it as much as we could, write this brief last-minute review, and insert a word or two in the other reviews where a comparison with the M-7050 seemed essential. A follow-up in the next issue should take care of the loose ends, including measurements (if at all relevant.) Even such brief exposure reveals outstanding qualities in the M-7050. Here's one Japanese amplifier that doesn't sound Japanese. Its basic sonic character is in the mold of the most sophisticated American high-end electronics. At this point, we rank it immediately below the Bedini Model 25/25 in transparency, lack of ''personality,"" smoothness, and delineation of detail. Just about all other amplifiers sound coarse, veiled or edgy next to the JVC, at least a little bit.

And here comes the punch line: the M-7050 is rated at 150 watts per channel into 8 ohms! It's the only 'monster amplifier' we've tested so far that can hack it in the purist category.

The output stage of the M-7050 is biased for what JVC calls Super-A operation; it isn't pure Class A but it appears to be more linear, judging by the sound, than Class AB amplifiers we're familiar with, and the power efficiency is still reasonable. The amplifier weighs about 50 pounds. Two neat features are the huge power level meters and the switching facilities for A-B-ing two pairs of speakers (or connecting both pairs simultaneously). For further discussion and amore finely tuned evaluation, see the next issue.

PS Model One

PS Audio, 3130 Skyway Drive, #301, Santa Maria, CA 93454. Model One Power Amplifier, $449. Five-year warranty. Tested #0397, on loan from manufacturer.

Another very late arrival that we just had to mention before going to press, the 80/80-watt PS also proved itself amazingly quickly against all amplifiers in its power and price category-and then some. After somewhat limited exposure to it, we rate it just a hair below the Amber Series 70 and an equally small distance ahead of our best sample of the Hafler DH-200.

The PS III phono preamp already demonstrated to us that these people know what totally neutral, uncolored audio amplification is all about, and their first power amp now clinches the proof. It has no personality. The bass is deep and solid, the midrange is impressively transparent, the highs are clean and nonirritating, spatial perspective is excellent.

The Amber has perhaps a slightly more refined and suavely detailed sound; the Hafler is a wee bit zippier on top and maybe not quite as firmly controlled and focused; essentially all three are in the same class, and a very classy class it is.

PS claims that a special, patent-pending circuit they call Storage Linearized Amplifier is responsible for the good sound. The amplifier came in much too late for us to have an opinion on that, but any awareness of storage as a cause of audible time-domain perturbations earns an automatic 'right on!" from us. This is clearly an audio product that has been engineered and listened to.

Rappaport AMP-1 (epitaph)

In case you haven't heard, Andy Rappaport is out of business. A. S. Rappaport Co., Inc., has declared bankrupt cy; Andy has a salaried job in Boston with a high-technology company that has nothing to do with audio; his plan is to remain an audio designer and consultant in his spare time but never again a manufacturer. All a case of too much too soon with too little capital, we'd say.

This creates a problem for those who own Rappaport equipment, some of which is rather temperamental and re quires expert servicing. The AMP-1, for example, has an almost 100% failure rate, if you count minor but annoying malfunctions (hum, excessive DC offset, asymmetrical clip ping, overheating, etc.) as failures-which they are. Andy informs us just as we go to press that an 'official' servicing agency is about to be set up in a New York suburb, which will have his full cooperation and the benefit of his engineering advice. Repairs, however, will be charged for; the warranty died with the bankruptcy. Quite independently, the largest Rappaport dealer in the New York area has declared to us his intention to honor the warranty and provide free service to his own Rappaport customers (but not to any other dealer's), probably by going through this same agency. So all is not lost, even if you didn't heed our warning about reliability and gambled on the Rappaport sound.

That sound, incidentally, hasn't been clearly bested yet; the Bedini Model 25/25 is the only power amplifier we know of that may quite possibly be superior to the AMP-1, but we'll never know for sure, having bailed out of the two we owned when the news of the dissolution reached us. The JVC M-7050 is probably comparable but not better.

All we can say at this point is that we never imagined it would end this way. Even so, it isn't our intention in the future to run a Dun & Bradstreet credit check on every bright young man who sends us an amplifier for review. We've got our hands full with just the performance aspects of consumer protection.

Sonotron PA-2000

Sonotron A/S, PO Box 2114, N-7001 Trondheim, Norway. PA-2000 stereo power amplifier, $650 (approx. export price, not retail). Tested #1189, on loan from manufacturer.

We don't quite know what to make of this 100/100-watt Class AB amplifier from Norway. It's a rather attractive package; the parts appear to be of good quality; even the price seems to be within the bounds of reason for a European high-end product. The reputation of the amplifier is quite high in Northern Europe. And yet--we just can't get decent sound out of the PA-2000.

On every little dynamic peak at the higher frequencies--strings digging in, sopranos hitting a high note, trumpets punctuating a phrase-the amplifier literally protests, creating a hot, piercing spike of sound that goes well beyond the typical hard or edgy quality we often complain about. It's extremely fatiguing and makes the amplifier virtually unlistenable.

In this case we really expected to see some kind of ringing on square waves or other obvious transient anomalies when we put the amplifier on the test bench. We were wrong.

The Sonotron passes all routine black-box tests with a clean bill of health. In fact, it looks very, very good on all conventional spec-sheet type measurements. We now suspect that what we heard has to do with feedback-related time modulations, which are very hard to catch on the wing and display on meters and CRTs.

While we were in the middle of all this, a Scandinavian audio dealer happened to mention to us that the Sonotron sounds much better when thoroughly warmed up. We routinely warm up all amplifiers for a good many hours before we listen to them critically, but this time we took special pains, even attempting to overheat the amplifier by preventing the circulation of air around the heat sinks. It made no audible difference, however, and we reluctantly gave up.

Recommendations

We still haven't been able to get our hands on a Mark Levinson ML-3 (Class AB, 200/200 watts), so we'll just have to take Mark Levinson's word that the less powerful ML-2, which we tested in mid-1978, is a more accurate amplifier and hasn't changed all that much since. We haven't tested the very latest from Audio Research or Threshold, either; otherwise we have the field reasonably well covered, we feel. We'll keep trying to test all the most promising contenders.

Best-sounding power amplifier tested so far, regard less of price, but of limited output capability: Bedini Model 25/25.

Best-sounding power amplifier of large output capability: JVC M-7050.

Best power amplifier per dollar: Amber Series 70 (see review above for qualifications).

Important Announcement from the Publisher

The last few issues of The Audio Critic, including this one, have been in effect double issues for the price of one- and also double issues in the time it took to publish them. We can no longer go on like this.

Charter subscribers will recall that The Audio Critic was originally a publication of more limited contents that came out more often, at only a 6.7% lower subscription price per issue. Instead of adjusting for the most devastating inflation in recent history, we've been writing, producing and pricing issues as if costs were lower than ever and money easier to come by. This must come to a halt, right now.

The next issue will contain fewer reviews and features than this one and won't take nearly as long to be published.

Our mail indicates that most of our subscribers would rather pay the same price per issue for smaller issues more often. In the course of 1980, we plan to convert to an entirely new format, which will eventually consist of quite thin issues published with considerably greater frequency, plus one very big, fat and highly educational reference issue, a kind of semi-permanent handbook to which the thinner issues will keep referring. Full details will be announced in the next issue; there may have to be another price increase, alas;

however, the six-issue subscription cycle will continue, only the page allocations will be different. The total six-issue package, including the reference handbook, should be richer in contents than ever.

It must also be pointed out that the cover date of the present issue refers to the period during which the equipment tests took place; revisions, updates and last-minute additions went well into the first two months of 1980, however, so you can consider all the information presented here to be thoroughly up-to-date, even to the extent that some of the equipment reviewed is barely available yet as this issue is about to be mailed to you.

---------

[adapted from TAC]

---------

Also see:

Further Developments in Step-Up Devices for Moving-Coil Cartridges: Denon AU-340; Denon HA-1000; Marcof PPA-1 (Improved) ; RWR Audio MCT-1 ; Signet MK12T

Why We're So Mean, Vindictive, Arrogant, Negative--and Truthful

Various audio and high-fidelity magazines