The Acoustic Research integrated amplifier

The first time we saw an AR amplifier (at a Hi-Fi Show), we were struck by its bland, almost antiseptic appearance. Amidst all those other audio products that looked as though they had been high-styled for Madame's boudoir, the unadorned simplicity of the AR amplifier made it stand out like an Eames chair at Williamsburg.

Well, now that we've looked at that simple, symmetrical control panel for a while, we have come to feel rather differently about it. We still feel it is bland, and maybe even antiseptic-looking, but like simply-designed furniture, it is easy to live with, it doesn't draw attention to itself, and it blends with any decor. As a matter of fact, we began to notice how garish most other control panels are.

To the hi-fi hobbyist, the first reaction to the AR amplifier is likely to be "Where are all the controls?" There are in fact fewer controls on it than on most competing units, but whereas many others have some rarely needed controls, or oft-needed ones that don't function very well, the AR has about the most useful controls of anything currently on the market.

There is a total of nine front-panel controls: input selector, vol ume, tape monitor switch, coaxial (slip-clutch) bass and treble controls, and a mode switch to select Mono (A+B), Stereo or Null.

A Useful Control
The Null position is a test mode, and is used for establishing exact electrical balance between the two stereo channels. Unlike the Mono mode, which adds the input signals (A+B) by combining them in phase, the Null mode combines them out of phase to make the signals subtract from one another (A–B). Thus, when the signals to the two stereo channels are the same, and of exactly the same volume, the result is almost-perfect cancellation, or a "null."

The channel blending takes place after the channel balance control, so if slightly unequal volumes fail to "null" completely, they can be balanced by adjusting the balance control for maximum nulling. And since both amplifier channels have essentially identical gain following the balance control, the signal balance required for maximum nulling will deliver equal outputs to the speakers when the Mode switch is set to reproduce program material.

The null adjustment must be carried out with monophonic program material, because if the input signals to both channels are not identical, they won't cancel each other completely no matter how perfectly they are matched in volume. Consequently, although the Nulling feature allows one to correct for inherent imbalances between the outputs of tuners, pickups, and so on, it can't be used to set the balance of a stereo signal. On the other hand, as far as we're concerned, if you can't hear an imbalance between stereo channels, why worry about it? And if you can, who needs a special test mode to help you correct it?

In addition, it must also be noted that, even with mono material, electrical channel balance does not ensure acoustical balance. Differences in loudspeaker efficiencies, differences in tweeter balance settings, and asymmetrical room acoustics can still make the system sound unbalanced, even when the amplifier output signals are perfectly matched. So although the Null is useful for establishing a starting point, the balance control setting that you use must ultimately be determined by the stereo balance that you hear.

The Phono Input
Most modern preamps are equipped with two sets of magnetic phono inputs: for high- or low-output pick ups. The AR has what we consider to be a far superior arrangement: a single pair of moving-magnet phono inputs and level-set controls. Thus, not only can a wide range of pickups be accommodated, they can also be matched in level to the amplifier's high-level inputs, so the Input selector can be switched from source to source without a drastic change in volume. And with a separate level-set for each phono channel, the pick-up outputs can be made perfectly equal.

Tone Controls
Perhaps the most conspicuous departure from "traditional" design, though, is in the AR amp's lack of any "loudness control." We have yet to find a preamplifier whose loudness control really worked properly. So-called loudness-compensated volume controls, in which the compensation can be turned on or off with a switch, almost invariably provide either too much or too little compensation. Most perfectionists just leave the switch turned Off all the time, but this does nothing to ameliorate the rather poor volume-control tracking that is caused by the volume control taps that operate the automatic loudness compensation circuits.

Acoustic Research decided, wisely in our opinion, to abandon the automatic type of loudness compensation and simply design the unit's bass and treble controls to provide what ever degree of loudness correction that the user wishes to apply. The fact that the resulting tone control action is eminently useful for other forms of tonal corrction may or may not be coincidental.

The action of these controls strikes us as being as close to ideal as any we've ever used. The bass control is of the so-called variable-inflection type (figs.1 & 2), which affects only the lowest frequencies at moderate control settings, and moves the inflection point—the frequency at which boosting or cutting starts to take place—upwards as the control is turned farther from its Flat position. Thus, it can be used to fill out the extreme bottom end of a speaker system that rolls off gradually below 60Hz (as do most moderate-sized systems), or can be advanced farther to fill out the entire low end of a thin-sounding program source. You can also, incidentally, do a better job of loudness compensation with it than you can with practically any so-called loudness control.


Fig.1 Variable-slope bass control (arrow shows point where control action begins).


Fig.2 Variable-inflection bass control (arrows show points where control action begins).

At the high end, the tone control action is essentially variable-slope, which is to say, the inflection point changes relatively little but the steepness of the curves increases as you rotate the control from its mid point. This is rather less effective for dealing with loudspeaker problems, but is ideal for coping with most of the high-end deviations encountered in typical program material.

Not Enough Inputs
About the only thing for which an audio hobbyist might fault the AR amplifier is its relatively small number of inputs. Besides the phono inputs, there are only two high-level inputs. This arrangement will handle a turntable, a tuner, and a tape machine, but if you hope to include also a TV sound input or a second turntable or a videotape recorder in your system, you will need external switching facilities for them. On the other hand, adding the necessary switching should be no problem for anyone who's handy with tools. Just remember to shield the switch connections, keep high-level signal grounds isolated from phono grounds and left-channel grounds isolated from right-channel grounds, and use the shortest possible lengths of shielded cable to and from the switches.

As far as protective features are concerned, AR seems to be much more confident about the durability of their amplifier than most other solid-state amp manufacturers. There are no instant-acting shutdown circuits in the amp at all; the line fuse and the thermal cutoff relays on the output stages are both slow- acting devices intended to protect only against prolonged overloads or catastrophic power-supply failures. There are however a couple of fast- acting fuses in series with the loudspeaker lines, and while these are ostensibly there to prevent loudspeaker damage (footnote 1), they also provide a degree of overload protection for the output transistors themselves. Rather than satisfy our curiosity about this at the outset of our tests, though, we left the blowout-protection checks until last.

We tested the AR amplifier with three varieties of loudspeaker: an all-dynamic system (Dynaco A-25), a dynamic/electrostatic system (Janszen Z-600), and a full-range electrostatic (the KLH Nine). On all three, the sound was very clean and lucid, but not without some degree of coloration.

Transparency was very good despite a barely detectable hint of graininess, crescendos were handled very cleanly and solidly, overload recovery was extremely fast (unless the protective thermal breakers let go, in which case it took several seconds for the dead channel to come on again), and when mild clipping occurred, the resulting clicks were judged relatively innocuous.

In comparison with other top-of-the-pile amp/preamp combinations we have heard, the AR amplifier sounded somewhat heavy and soft, as though most of the frequency response curve was tilted downwards towards the right (fig.3). Frequency response measure ments revealed a very slight rolloff at the high end, but nothing to suggest why the amplifier sounded the way it did. To the contrary, the AR amplifier sounded as if it were very slightly tipped up at the extreme top—above about 10kHz—when driving either of the electrostatic-tweeter systems, so we were forced to conclude that our measurements were meaningless. As has often been the case in the past, the best thing we can do this time is shrug our shoulders at the measurements and say "Well, that's the way we read it, and this is the way it sounds." If you can visualise balance from a response curve, fig.3 will give you an idea of the extent of the coloration. It is by no means pronounced, but it is audible.


Fig.3 The AR Amplifier, subjective—that is, as opposed to measured—frequency response with all tone controls set to Flat (1dB/vertical div.).

In case you're wondering, it was not possible to correct for this mild coloration by means of the tone controls, for when these gave adequate correction in the mid-bass and lower treble ranges, they gave much too much correction in the low-bass and extreme treble ranges. If they had been able to do the job, they would have been much less efficacious in treating program-material aberrations. You can't have everything, you know.

Footnote 1: Solid-state amplifiers are much more prone to blow out loudspeakers than were tube-type amps because of their superior low-frequency power-output capability. A subsonic pulse that would simply saturate the core of an output transformer may pass through a solid-state amplifier at close to full power output, and few speakers can handle such an impulse without tearing themselves apart.
Acoustic Research, Inc.
3502 Woodview Trace
Indianapolis, IN 46268
(844) 353-1307

mrounds's picture

I really wanted to get one of those back in the day, but couldn't afford it as a starving student. Got a Nikko instead, of which the less said the better though it was greatly improved after being the subject of a reconstruction project by an EE roommate.

dalethorn's picture

The biggest dog I ever got was the Advent Receiver, containing a preamp by one Tomlinson Holman. I'm sure the preamp was great, but the receiver had a hum that couldn't be fixed. Worse yet was after borrowing a set of LS3/5a's that I discovered had a defective woofer, the owner of those speakers was sure that my Advent receiver blew the LS3/5a.

soundhound's picture

I bought the AR amp when it first came out. I use it up to this day, but now its in my workshop doing 3rd system duties. Like all examples of this amp, the bias circuitry went wacky due to the cheap bias-setting potentiometers, but that was easily fixed. I never noticed the sound quality issues mentioned in this review, but I regard all reviews in magazines like this more entertainment than something which provides actually useful information.

tonykaz's picture

Back in 1969 was any information "useful"?

Our superb Music Systems today are conduits of Suberb Recordings, aren't they?, capable of sounding rather horrible when playing the rather horrible source material we had back in those times -- the 1960s.

I'm not-much interested in this AR Amp. I'm reading Mr. Holt and fellow panelists, who don't seem to merit identification. Who are "we" in J.Gordon Holts constant references ?

I don't recall knowing about Holt (or Stereophile ) until recently ( 2014 or so ) because Stereophile was the Parent of Innerfidelty and Publisher of our Tyll the Headphone Authority that I happened to meet at RMAF in 2011 and thought to be conspicuously brilliant. In fact, I discovered ( or perhaps uncovered ) that most of the entire group at Stereophile are working at a very high level of Journalistic Integrity, along side the Guardian, NYTimes, Washington Post, my 4 regular daily reads.

For me it's interesting that you own an AR Amp that still sings after nearly 50 years, geez, that's gotta make it an A+ Recommended piece of electronics.

So Mr. Holt should've concluded with :
Buy this Amp and own it for the rest of your Life!

I'm say'n :
It doesn't get much better than that! ( at any price )

Tony in Michigan

dalethorn's picture

The principles of hi-fi, well-presented by Gordon Holt, are essentially the same today. What are different are the gear and the recordings. Good stereo recordings that sound great on today's gear do exist, albeit some remastering usually helps.

I've subscribed to Stereophile since 1971, and can appreciate what Holt heard in speakers, amps, and even headphones. Holt knew the differences between the better amps and the not-so-good amps all the way back to the 1960's, and I came to appreciate his advice after buying the likes of Crown and SAE, to name a couple of brands.

I even soldered out the tone controls from a Dynaco preamp on his advice - time well spent. But as they say of many subjects as well as hi-fi, those who don't know the history very precisely are likely to repeat some of the mistakes.

dmineard HT's picture

I actually remember reading this review when it was published. I had owned the Dynaco Pat-4 pream & Stereo 120 amp for about 3 years when the AR integrated amp was released.

At the time I agreed with the review, consistent with what I heard. I had that Dynaco system for 5 or 6 years and sold it to move up to McIntosh preamp and amp.

Although I did not subscribe to that magazine for a year or two, I really liked the technical reviews of Gordon Holt. Really enjoyed reading this review...brought back fond memories of a simpler time. Thanks.