The News in Power Amplifiers: Mostly Very Good (Vol. 2, No. 3, Spring - Fall 1980)

The News in Power Amplifiers: Mostly Very Good: Amber Series 70 (follow-up) , Bedini Models 25/25 and 45/45 (follow-up) , Denon POA-3000 , JVC M-7050 (follow-up) , The Leach Amp and The Leach Superamp , Sonotron PA-2000

----

By the Staff of The Audio Critic

There are some exciting new developments in high-priced power amplifiers. In the under-$500 category, however, we have no new recommendations to make.

Since our last issue, nothing has come up to change our mind about the severe limitations of ' 'black box' bench tests (THD, IM, phase shift, slew rate, input/output nulling, square waves into complex loads, etc., etc.) when it comes to determining the ultimate quality of a power amplifier. A recent invitation to a Bruel & Kjaer seminar on advanced measurement techniques referred to ''the concept of measurement used as an indication of good or bad performance rather than as an absolute measure."' Now they tell us, after we spent every cent of the kids' college money on our lab bench! What manufacturer in his right mind would send us an amplifier with downright ' 'bad performance,' anyway? For the unconscionable price of B & K test instruments, we want absolute measures indeed, not just the ability to tell a Mark Levinson from a kitchen radio. . .

Seriously, though, it still makes very good sense to put power amplifiers through our routine series of measurements, just to back up and elucidate our listening perceptions. We don't believe that the purely subjective underground reviewers (the ones who intone with mystically closed eyes that the highs are whitish and the upper midrange insufficiently liquid, but don't know an ampere from a volt) would have been able to identify, for example, the gain-control peculiarities of the Denon and JVC amplifiers reviewed below. Signal generators and oscilloscopes do have their use.

Otala's last stand.

Dr. Matti Otala, quite possibly the world's foremost researcher on subjects having to do with power amplifier design, is back in his native Finland and turning his attention to technologies other than audio, at least for the moment. His last act on behalf of audiophiles before he left the United States was to finish the design of the Citation XX, a 260/260-watt class AB stereo power amplifier that represents, we're told, the sum total of his expertise. Harman/ Kardon will begin to manufacture the amplifier in Japan as soon as certain production and marketing decisions about it are final. We have seen but not heard the prototype; it certainly creates the impression of a supreme engineering effort, inside and out. We only hope that Harman/Kardon will be able to maintain a no-compromise attitude in all phases of Citation XX production, even though the company's new Japanese owners come from an unabashedly commercial school of manufacturing.

Until we can actually listen to a Citation XX, the most exciting thing about it from our point of view is the large number of new ideas on amplifier performance that went into its design concept. Among these are the recognition of distortion-producing mechanisms such as the modulation of the output impedance (in class AB amplifiers) by the momentary signal amplitude; feedback-induced phase modulation of the higher frequencies by the amplitude of the audio signal (see also Vol. 2, No. 2, p. 37); and the modulation of transistor junction temperatures by low-frequency audio signals at high amplitude. A full discussion of these phenomena is beyond the scope of this article, but the power amplifier chapter in The Audio Critic Handbook (see announcement on page 2 of this issue) will take them into consideration, and eventually we hope to modify our power amplifier measurement routine accordingly. Dr. Otala already claims to have a workable phase modulation test using signals of 200 Hz and 20 kHz, but has not published it so far.

Look for new horizons, then, in the not-too-distant future.

Amber Series 70 (follow-up)

Amber Electronics, Inc., 500 Henry Avenue, Charlottesville, VA 22901. Series 70 Power Amplifier, $499. Three-year warranty. Tested (and retested) #700100, on loan from manufacturer.

In the course of our follow-up tests it became quite apparent that the Amber doesn't just barely nose out other power amplifiers in its class (PS Model One, Hafler DH-200, Audionics CC-2, etc.), as we had reported in the last issue, but handily excels them in overall listening quality. This is now definitely our favorite ''cheap'' power amplifier. It has a nice, solid bottom; a midrange that lacks the ultimate transparency obtainable at much higher prices but is open and musical nonetheless; and a clearly etched top end that doesn't harden or smear even when the program material has a wide dynamic range and is rich in high-frequency energy (e.g., 30-IPS master tapes of band music). This last virtue is probably the result of judicious filtering at the input. All in all, the Amber sounds more like a high-end power amplifier than any other under-$500 unit we're aware of.

As we go to press, some minor circuit modifications are about to go into production. We plan to follow up on these as soon as we can get our hands on a sample; meanwhile we have no reason to believe that the revised version of the Amber Series 70 won't sound even better than the original. These people obviously know how to listen.

Editor's Note: A follow-up review of the Audire DM 700 "monster amplifier'' was supposed to appear in this space, to report on the promised modification mentioned at the end of the original review in the last issue. The modification turned out to be something of a disaster; meanwhile, however, the DM 700 has been discontinued, so that the matter is now largely academic.

Bedini Models 25/25 and 45/45 (follow-up)

Bedini Electronics, 13000 San Fernando Road, #9, Sylmar, CA 91342. Model 25/25 Class A Power Amplifier, $845. Model 45/45 Class A Power Amplifier, $1300. Tested samples (final production versions) on loan from manufacturer.

Since our original review in the last issue, the marvelous little Model 25/25 had undergone a $195 price increase and some major changes in its internal physical layout, eliminating virtually all wiring other than the circuit boards themselves. The sound is, if anything, even better; the silkiness of the highs and the transparency of the midrange are unsurpassed in our experience, except possibly by some-not all-versions of the Futterman tube amplifier and one or two experimental solid-state prototypes. The bottom end of the Bedini is very impressive for a 25/25-watt stereo amplifier with a single power supply, but of course there are many large amplifiers with all-out dual power supplies that will give you firmer and subjectively deeper bass. (John Bedini has been experimenting with a heavily beefed-up power supply for a variant version of the 25/25; we have looked at a slightly miscalculated sample and found it quite promising, but for the moment this is not a commercially available product and would have to sell at a much higher price if it were marketed.) We want to state again that the Bedini Model 25/25 appears to be just about the ideal amplifier for driving the Quad electrostatic loudspeaker; the combination provides the utmost clarity, very satisfactory SPL capability and complete safety, since the largest voltage swing the 25/25 can deliver into a load of any impedance is considerably short of what ever could damage the Quad, even on a long-term basis. And that $845 price tag seems a little more tolerable in a low powered amplifier when you can tell yourself that you're buying a Quad accessory.

We also wish to withdraw our recommendation of the Cotter NFB-2 noise filter/buffer for use at the input of the 25/25. That was our initial lab bench determination; several months of listening experience have convinced us that real world input signals, even the fastest, are unable to get the Bedini into any kind of audible trouble.

As for the Model 45/45, we tested an improved version that was considerably better than the original, very nice in fact, but still without the extraordinary qualities of the 25/25.

At $455 additional cost, the 25 dB increase in power and somewhat fuller bass are insufficient to offset the slightly less pellucid midrange, shallower spatial perspective and not quite edgeless top end. The 25/25 is a classic, whereas the 45/45 remains just another good (and very expensive) amplifier.

We're still waiting for the Model 200/200; meanwhile we must register at least a temporary vote in favor of John Bedini's argument (and Andy Rappaport's, for that matter) that pure class A and no negative feedback loop are the way to go.

Denon POA-3000

Denon America, Inc., 27 Law Drive, Fairfield, NJ 07006. Model POA-3000 Stereo Power Amplifier, $2300. Tested #1110021, on loan from manufacturer.

This is a "sliding" class A (i.e., dynamic-bias, not-quite-pure class A) Tokyo super-spectacular, with huge peak-reading meters, lacquered rosewood sides and 180 watts rated output per channel. Very appetizing indeed. That $2300 price tag invites a hard-nosed critical attitude, however; so it should be stated right up front that we much prefer a pair of Leach Superamps ($1598 total)-and that's basically the whole story. The Denon is a very nice amplifier, but the big Leach is better. Once again, the best is the enemy of the good.

As a matter of fact, had the POA-3000 come in for testing at the same time as the JVC M-7050 we reviewed so favorably in the last issue, we would have rated the two as approximately equal in quality, with the nod going possibly to the Denon for perhaps a slightly better trade-off between brute power and sweet transparency, especially with the aid of the Cotter NFB-2 filter at the input. The arrival of the less costly, much more powerful and sonically much more neutral Leach Superamp, however, makes such a judgmental balancing act unnecessary.

The sound of the POA-3000 is almost, but not quite, free from the hardness or edginess we object to in so many amplifiers, and our investigation of this phenomenon (both with and without the Cotter filter at the input) revealed a rather startling error in design. The input level control of the amplifier acts as a variable low-pass filter, instead of merely controlling the input sensitivity for biamping and other applications that require level setting. With the control fully clockwise (''max''), the measured rise time through the amplifier is approximately 1 microsecond, which agrees quite neatly with the 350 kHz bandwidth spec. With the control turned back counterclockwise to a still perfectly reasonable level, the rise time slows down to as much as 15 microseconds, which is definitely slower than the human ear.

What happens is that the amplifier has a perceptibly different sound in each position of the level control, edgier in some positions than in others and benefiting from the additional slowdown effect of the Cotter filter only in the edgier positions. Weird, isn't it? The location of the level control in the circuit is obviously incorrect; perhaps it was an after-thought.

Otherwise the amplifier measures beautifully; you'll have the best Japanese specs on your block if you own one.

And you'll also be satisfied with the sound-big, plush, clean, detailed, and only subtly flawed by the peculiarities discussed above-unless somebody else on your block al ready owns a pair of Leach Superamps.

JVC M-7050 (follow-up) 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.

Now that we've had time to put the M-7050 on the lab bench, we can report that its behavior is quite exemplary on all standard tests, in the expected manner of a Japanese high-end product. It amazed us, however, to discover exactly the same input level control anomalies as we found in the Denon POA-3000 review above. In the JVC, the slowing down of leading edges isn't quite as drastic (about 7.5 micro seconds was the slowest rise time we measured), probably accounting for our initial insensitivity to the problem. There are distinct differences in sound, nonetheless, as you turn down the level control, although the need for an external filter such as the Cotter isn't as obvious even in the relatively edgy '"'norm'' (fully clockwise) position as in the case of the Denon. With their essentially similar sliding class A circuit concepts, huge illuminated front-panel meters and almost identical level control peculiarities, the two Tokyo heavy weights appear to be virtual twins; nor are they far apart sonically, as we have already indicated. Luckily the new Leach Superamp (see review below) resolves the potential dilemma of ranking either of the two above the other.

The Leach Amp and The Leach Superamp

LSR&D, Inc., 481Buckingham Circle, Marietta, GA 30066. Sales: 100 Hiawatha Drive, Mount Pleasant, MI 48858. The Leach Amp, Model 101 (stereo), $799. The Leach Superamp, Model 102 (mono), $799 each, $1598 the pair. Three-year warranty. Tested Amp #001105 and Superamps #001020/#001021, on loan from manufacturer.

The Leach Amp is a 160/160-watt class AB stereo power amplifier with a single power supply. The Leach Superamp is a rather similarly conceived class AB mono power amplifier capable of more than 300 watts output. The two look physically identical (low-profile matte black boxes with 19-inch rack panels, small rack handles, huge rearward heat sinks and removable walnut side panels) and their construction reeks of quality without high-end cultist excess.

Everything about them is buttoned down, functional and well made, inside and out.

The designer of the amplifiers is one of the very few genuine authorities on the subject to turn manufacturer, namely Prof. W. Marshall Leach, Jr., who teaches electrical engineering at the Georgia Institute of Technology in Atlanta.

Prof. Leach is well known to readers of learned journals, as well as of more popular technical magazines such as Audio, for his numerous papers and articles on amplifier circuitry, TIM, phase distortion and assorted other subjects having to do with high-quality sound reproduction. As a matter of fact, Prof. Leach has saved us the trouble of explaining his latest design philosophy with his own two-part article 'Build a Double-Barreled Amplifier'' in the April and May 1980 issues of Audio. The amplifier presented in great detail in this do-it-yourself article is the immediate ancestor of the Superamp, although the latter incorporates certain refinements and is considerably more sophisticated in packaging.

Since just about every subscriber to The Audio Critic also reads Audio (or so it appears from our correspondence), we shall confine ourselves to a few general remarks about design features and then go on record with our evaluation.

The Superamp, which is definitely the more impressive of the two Leach units, even on a per-dollar or per-watt basis, has a stupendous power supply (130 joules energy storage-that's enough energy to lift an adult male Doberman pinscher one foot off the ground, or two Dobermans in stereo) and is therefore capable of meeting almost any current demand made by the load. The Superamp is also the first class AB monster amplifier we've run across whose ''AB ishness'' isn't obvious even after prolonged listening; the circuit design minimizes all class AB compromises so cleverly that you can almost pretend you're listening to the world's largest class A amplifier. (Think what a 300-watt class A job would cost.) As far as minimizing all conventional, old fashioned distortions is concerned, not to mention the new and fashionable ones (TIM, SID, DIM and all those gents), not many people in the world know more about the subject than Marshall Leach, so we weren't surprised to find out that the amplifier performs exactly as he specifies. On the other hand, we firmly believe that the real reason for the amplifier's superior sound qualities lies outside these tests, since sonically inferior units are sometimes also capable of passing them with flying colors.

Justhow good does The Leach Superamp sound? Super.

Perhaps we're letting the stupendous dynamic range influence our judgment (for the first time, we're hearing our 30-IPS piano master tapes without any amplifier clipping, even at front-row concert levels), but we also find the bass rock-solid. the midrange totally transparent and the highs superbly detailed. Perhaps there is just a smidgen more hardness to the upper midrange and lower treble than in one or two pure class A amplifiers of much lower power, but then comes a huge orchestral climax with the high-frequency textures truly unsmeared for the first time, and were disarmed.

Except for fussy limited-power applications such as driving the Quad electrostatics, a pair of Leach Superamps is now our reference. And, we might add, at a fraction of the cost of some of the more absurd high-end amplifiers, whose superiority remains to be proved to our ears.

As for the Leach Amp, with its mere 160 watts per channel at half the price. we find it beautifully transparent in the midrange, very well controlled on the bottom end, but a bit over-bright and glassy on top. (Our bench tests won't tell us why.) Not that it doesn't sound extremely impressive on first hearing: the highs aren't so aggressive that the clarity and dynamic range of the amplifier are lost on the listener.

For example, the Sound-Lab R-1 electrostatic panels are extremely happy with The Leach Amp driving them; the Superamps are capable of swinging just a bit too much voltage for our peace of mind in this particular case. At $2.50 per watt per channel, we don't know of a better power amplifier than The Leach Amp, but for our Reference B choice it would have to cost less and for Reference A it would have to sound better. For example, the little Bedini Model 25/25, at $845, sounds unquestionably smoother, less fatiguing and more musical.

As a total engineering, packaging and marketing effort, however, the two Leach power amplifiers are impressive beyond anything we've seen from other sources for the past few years. At its debut, LSR&D looks to us like our idea of a high-end audio company, combining superior technical knowledge with a commitment to practicality and value, apparently striving to help the consumer, not to impress or one-up him.

Sonotron PA-2000 (follow-up)

Sonotron A/S, PO Box 2114, N-7001 Trondheim, Norway. PA-2000 stereo power amplifier, approx. $1500 (estimated U.S. retail price if and when available). Tested #12228 (second sample), on loan from owner.

We are pleased to report that the sample we had reviewed so unfavorably in the last issue must have been defective in some way, or at least quite untypical, since our second sample didn't even resemble it sonically. Not that the newer unit was totally free from the hard, over-bright quality that bothers us in all class AB amplifiers except a very few, but overall the sound was quite respectable by any standard.

Again, the availability of The Leach Superamp in roughly the same price category makes a fine-tuned evaluation unnecessary. The Sonotron isn't even close.

Recommendations

Once more, we must preface these with a disclaimer of omniscience. We haven't tested every power amplifier in the world that might be of interest to our readers. Nor has anyone else.

Best-sounding low-powered amplifier tested so far, regardless of price: Bedini Model 25/25.

Best-sounding high-powered amplifier tested so far, regardless of price: The Leach Superamp.

Best-sounding power amplifier at a much lower price than the above: Amber Series 70.

-----

[adapted from TAC, Vol.2, No.3 ]

---------

Also see:

The Ever-Changing Preamplifier Scene: Audio Research SP-6B (interim report) , Beveridge RM-1a/2a , Cotter System 2 (reappraisal) , Mark Levinson ML-7 (interim report), Precision Fidelity CTA , Robert Grodinsky Research Model Four

Various audio and high-fidelity magazines