Tandberg 3012 Integrated Amplifier Review

Tandberg's Model 3012 integrated amplifier is rated to deliver 100 watts per channel into 8-ohm loads from 20 to 20,000 Hz with no more than 0.02 per cent distortion. It is packaged compactly, measuring 17-1/8 inches wide, 13-3/4 inches deep, and 3-1/4 inches high and weighing about 21-1/4 pounds. Its styling and size match those of current Tandberg tuners, and it can be stacked with one of them to form a “receiver” of exceptional quality. The amplifier is available in either matte-black or pewter finish.

The front panel of the 3012 presents a simple, uncluttered appearance, yet it provides all the operating flexibility you expect from a top-grade amplifier. The large volume knob is concentric with a center-detented balance control. There is a front-panel headphone jack. Small pushbuttons control power, tone-control defeat, bass and treble turnover frequencies, and loudness compensation. A pair of peak clipping lights show when either channel is overdriven. Small detented rotary switches operate the speaker selection (a, b, a + b, off), the tone controls, and the record and program source selection. These last two controls enable the user to select different sources for simultaneous listening and tape recording. The program options include phono, tuner, and two tape-recorder inputs as well as a digital disc input. The record sources are the same except that the tape mode connects the two decks for copying from either one to the other.

It is noteworthy that the digital disc input is not identical to the other high-level inputs, tuner and tape. The latter two are buffered by low-distortion amplifier stages before they are selected by the front-panel switches, so that crosstalk between them is completely eliminated. The digital disc input, however, is switched directly to the volume control, minimizing the number of extraneous elements in the signal path.

The 3012 embodies a number of design concepts that its creators felt would contribute significantly to its audible qualities, if not to its measured performance. For example, Tandberg engineers concluded that dielectric absorption in certain types of electrolytic or ceramic capacitors used in most amplifiers for interstage coupling tends to degrade sound quality in subtle ways. To eliminate any possibility of such degradation in the 3012, no electrolytic or ceramic capacitors are used in its signal path, from the phono inputs to the speaker outputs. Low-loss plastic (polyester) foil-dielectric capacitors are used instead in all signal-carrying circuits. The designers also felt that a high slew rate was a desirable quality in an amplifier, and the 3012’s 1,000-volt-per-mi-crosecond slew rate is the result. Within the amplifier circuitry, every opportunity was taken to eliminate known or suspected causes of signal degradation.

The phono preamplifier of the 3012 accepts either a moving-magnet (MM) or moving-coil (MC) cartridge input; there are separate MM and MC input jacks, but no switching is required. The phono equalization follows the modified RIAA characteristic adopted by the IEC several years ago, which calls for the phono-input response to be rolled off below 20 Hz. The 3012 incorporates a fixed infrasonic filter that cuts off below 15 Hz at an 18-dB-per-octave rate, effectively eliminating any possibility of amplifier or speaker overload from record or turntable rumble.

The tone controls depart from conventional practice, being designed to give useful correction only at frequencies where it is likely to be needed, with a minimum of audible effect elsewhere. The bass and treble controls (eleven-position step switches) use 1 per cent precision resistors for accurate 2-dB control steps and have selectable turnover frequencies (100 or 200 Hz for the bass, 3,000 or 6,000 Hz for the treble). Although loudness compensation—anathema to some purists—is included in the 3012, it is designed to give a maximum low-frequency boost of only 6 dB and does not affect the high-frequency response.

The output stages of the Tandberg 3012 use two power MOSFET’s per channel; these require no current or voltage limiting for their protection and also operate at a very high speed (as demonstrated by the amplifier’s slew rate and by a slew factor said to be more than 1,200). The power supply is based on a compact but powerful toroidal transformer, with a pair of 15,000-microfarad filter capacitors providing the necessary energy storage for handling high-power program peaks. Price: $995.

Laboratory Measurements

The top of the 3012 never became too hot to touch during the one-hour preconditioning period or subsequent high-power testing. The internal heat sinks are effectively cooled by air entering through the bottom and exiting through the top of the amplifier. Our power measurements were made through the digital disc input. The 1,000-Hz clipping-power output with 8-ohm loads was 109 watts per channel (for a clipping headroom of 0.37 dB). The maximum output was 127 watts per channel into 4-ohm loads and 98 watts per channel into 2-ohm loads. Using the tone-burst signal of the dynamic-head-room test, we measured short-term clipping outputs of 156, 213, and 156 watts into loads of 8, 4, and 2 ohms, respectively. The 8-ohm dynamic headroom was, therefore, 1.93 dB.

The 1,000-Hz harmonic distortion, driving 8 ohms, rose from 0.003 per cent at 1 watt to 0.014 per cent at 100 watts. With 4-ohm loads, it was 0.0056 per cent at 1 watt and 0.034 per cent at 100 watts (the thermal-protection circuit cut off the amplifier at this point). The 3012 is not rated for 2- or 4-ohm operation, but with 2-ohm loads the distortion was 0.014 per cent at 1 watt, increasing to 0.55 per cent at 70 watts. The output devices had become very hot at that point and apparently were no longer able to deliver the nearly 100-watt output we had measured earlier.

The distortion at rated power into 8 ohms was between 0.012 and 0.018 per cent from 20 to 20,000 Hz except for a rise to 0.047 per cent between 30 and 50 Hz. This rise may have been the result of an interaction with power-supply ripple, although this was not apparent on the spectrum-analyzer display. At half and one-tenth power, the distortion curve’s shape was similar, with slightly lower readings (0.005 to 0.025 per cent) over the full frequency range at 10 watts output.

The amplifier’s slew factor exceeded our measurement limit of 25 (Tandberg’s own measurements show a slew factor of 1,200, corresponding to a 24-MHz input before waveform distortion becomes visible!). We measured the IHF IM (intermodulation) distortion with 18- and 19-kHz inputs at a 100-watt output, which produced a second-order (1,000-Hz) component of about —86 dB and a third-order (17-kHz) component of —78 dB. The amplifier displayed strong ultrasonic ringing when we drove a simulated reactive loudspeaker load with a 10,000-Hz square wave. This load, whose highly reactive impedance dips to 2 ohms at some frequencies, is designed to be more severe than that of any loudspeaker one is likely to encounter.

The 3012’s sensitivity for a 1-watt reference output was 16 millivolts (mV), with an A-weighted noise level of —77 dB referred to 1 watt. The MM phono sensitivity was 0.17 mV, with a — 80-dB noise level, and for the MC input it was 0.016 mV with a — 75-dB noise level. The phono preamplifier (MM) overloaded at a very high input level, ranging from 240 mV at 20,000 Hz to 323 mV at 20 Hz. The phono equalization was within 0.5 dB of the RIAA characteristic from 35 to 20,000 Hz, falling to —1.5 dB at 20 Hz because of the infrasonic filter. It was not significantly affected by phono-cartridge inductance.

The characteristics of the 3012’s tone controls were as close to ideal as we have seen from a simple bass-and-treble configuration. Even at their extreme settings, where the response at the frequency limits was varied by about 10 dB, the midrange was totally unaffected. Using the 100- and 6,000-Hz turnover frequencies and the middle control settings, we were able to modify the response by 6 dB at 20 and 20,000 Hz with essentially no effect between 60 and 3,000 Hz. The loudness compensation was equally subtle in its action, beginning to boost the response below about 200 Hz at volume settings of —20 dB or less and giving a maximum boost of 6 to 7 dB at 20 Hz. The overall frequency response with the tone controls bypassed was literally ruler-flat (less than 0.1 dB total variation) from 20 to 20,000 Hz.

Comment

Tandberg claims that, good as the measured performance is, the 3012’s special distinction lies in its sound quality. Although we have never found such distinctions among high-quality amplifiers to be audibly significant, we can say that the sound of the 3012 is easily as good as that of any other fine amplifier we’ve heard. Possibly those golden ears who claim to hear differences between the types of capacitors used in an amplifier will be able to appreciate the special sonic properties of the Tandberg 3012 more than we can.

The 3012 has absolutely no vices that we could detect. It is not in the least fragile, since it shuts itself off if abused electrically or thermally, returning to service automatically when conditions have gone back to normal. There are no extraneous noises connected with its operation: no switching transients, no start-up or turn-off thumps, no audible hum or hiss at any control setting, and so forth. We did much of our listening using a digital Compact Disc player as the program source, and the results were as satisfying as one would expect from the combination of a state-of-the-art program source and a state-of-the-art amplifier (the expression “state-of-the-art,” although much abused, definitely applies to both products in this case).

The solid construction and quality of the components used in the Tandberg 3012 set it apart from most of its competition and appear to justify its considerable (but not unreasonable) price. This amplifier is built so well and performs so satisfactorily that it is rather difficult to criticize. It is that good.

Read all

Tandberg reviews

Read all