Conrad-Johnson MF2500 power amplifier
Like many Conrad-Johnson designs, the MF2500's higher performance goals have been attained through experience and the refinement of established circuits, not by some magic new topology that's claimed to solve all previously known audio-fidelity problems. Thus the input stage is a standard differential circuit executed with FETs. The second stage comprises a buffered voltage-gain amplifier, and uses bipolar transistors. It drives a regular class-A/B push-pull bipolar output stage, which is direct-coupled to the speaker. The output stage is run at quite a low idle power, which partly explains why the amplifier takes a little time to get up to speed from a cold start (footnote 1). But it does mean that the amplifier does not run too hot to touch.
While some of the MF2500's technology is derived from the older MF2300A (reviewed in October 1996), the big change has been from FET to bipolar output devices. In this circuit, at least, bipolars confer greater gain stability with output load and current variations, imparting a more consistent and more dynamic sound quality. The MF2500 has high-current, resin-encapsulated Sanken power transistors: two 2SC3281 and two SA1302 devices are paralleled for each channel's output stage. These transistors have good thermal properties, high-quality electrical connections to their junctions, and are endowed with good safe-operating areas. This means the amplifier can be protected with simple quick-blow fuses.
C-J favors a predominantly low-order distortion spectrum in which some second harmonic is designed to dominate the distortion residual, this in the MF2500 courtesy of the input FETs. Minimal loop negative feedback is used, in this case just 12.5dB, or approximately 4x. (Some amplifiers may run with x1000 to x10,000 of feedback.) As Lew Johnson explained, "the more the performance of the basic circuits was improved from the new MF series, the more important was the upgrading of individual component quality. Then negative feedback could be reduced yet again."
The MF2500's regulation topology is of Premier quality. For example, the input stage has its own transformer tap stabilized at a voltage higher than the output stage requirement. Discrete open-loop regulators have audiophile-grade polystyrene and polypropylene capacitors. Only the main current reservoirs are electrolytic, but here there is no choice. Voltage-reference and feedback-control resistors are costly metal-foil Vishays. Six 10,000µF Nichicon electrolytics form the central reservoirs for both channels, giving a total of 30,000µF per DC supply phase and ensuring good current delivery at low frequencies.
The massive EI-lamination transformer has split secondary sections to feed the various regulators. I also counted no fewer than 22 of C-J's exclusive lead-foil, large-value polystyrene capacitors, plus eight higher-value, audiophile-grade polypropylene caps.
The MF2500's 3/8" alloy front panel is anodized in C-J's traditional gold, while the rest is finished in satin-black textured enamel. The baseplate can be removed to access the underside of the printed circuit board. The only control is a front-panel power switch, and there's no soft start. The heavy transformer is affixed to the right side of the chassis, while the single, external, full-size finned heatsink is to the left. The inputs are single-ended phono sockets, the outputs 5-way binding posts, all gold-plated. Thermal protection is via a self-resetting thermal trip circuit. The power cord is a captive three-core type.
I found I preferred to leave the MF2500 on most of the time, its low idle power proving a blessing. The amplifier also benefited from extended break-in.
In some respects, the MF2500 gave an understated view of musical events—with casual listening it's possible to dismiss it on grounds of a lack of forthrightness for its rather measured pace, and an almost rich, distant presentation. Yet, over the long term, and along with its even-tempered if slightly sweet tonal balance, it was the very absence of a forced or excitable character, false hardness, or excessive attack that endeared the MF2500 to me. There was a sense of ease; aural fatigue was significantly lower than usual for this class of amplifier. Just when I thought I'd got about as much out of the MF2500 as I was going to get, it got better by stages.
Elements of this sound—almost a tube experience—had been dimly evident in earlier solid-state amplifiers from C-J. But some of the necessary tonal consistency, fine textures, vitality, and clarity were not expressed well enough to make those models truly viable alternatives. Their FET output stages suffered from some variations in both clarity and tonal color with reproduced sound level, this to some degree also dependent on the loudspeaker load and the program complexity.
In the case of the MF2500, two key aspects have become firmly established. First, this amplifier offered the ability to play all the way to full output with consistent quality. Second, while the MF2500 won't put C-J's Premier Eight A monoblocks out of business, or even the smaller Premier Eleven A, its sound conveyed much of the general nature of a fine tube amplifier—a classic type of moderate negative feedback, wide bandwidth, and linear, push-pull design—coupled with greater dynamic range and sheer load-driving grunt.
Tonal colors were natural, never strident, thin, or hard. While the bass didn't have the kind of deeply extended impact of the biggest, most costly reference amplifiers, it was satisfying. Not only was the bass powerful—more than sufficient to do justice to cathedral organ at full stretch, and able to carry the low-frequency presence of a large symphony orchestra—it was also articulate, in terms of both detail and tune-playing. The bass also had a kind of bouncy, springy recovery that lent drive and interest to rhythm bass lines on jazz and classic rock.
Through the midrange, the MF2500 had a thermionic tonal poise that appeared to have a payoff in terms of natural image depth and perspective. While the amplifier's transparency was rated fairly well, the natural tonality helped create coherent, focused well-layered images. Stage width also rated above average, if a little more like that of a fine moving-coil cartridge than a CD source.
In the upper frequencies, the amplifier remained even tempered, well balanced, and even quite delicately resolved. String tone was nicely presented, and no false edge or breathiness was added to vocal sibilants. In fact, vocals in general were handled with near-tubelike quality and good clarity, and the inner harmonies of massed choir were well read, with well-maintained perspectives.
Within that slightly "relaxed" temperament, the MF2500 nevertheless managed to pack quite a punch—this is a bighearted amplifier, and sounds it. As it continued to break-in and I grew more familiar with its balance, musical values asserted themselves. Pace, rhythm, and timing, plus dynamic resolution, were all rated significantly above average.
On my personal scoring scale, this amplifier achieved a commendable 22 points, a significantly good result for type and price.
Conrad-Johnson's MF2500 is a successful product on many grounds. It closely approaches the classic tonal character of a fine tubed amplifier—a C-J hallmark—and holds on to that quality over a great range of output power and possible speaker loading. Not easy to accomplish.
It also avoids the cheap, quick thrill of a bright, fast, forward sound. Instead, it builds on quality and depth of performance to achieve its commendable rating. Its sound is musically entertaining without being hyperactive, while its broad, bighearted approach is easy on the ears. All kinds of music were handled with equanimity.
In the lab, the MF2500's high output power and good load tolerance were confirmed. The amplifier could dump up to 900Wpc into 2 ohm loads for short-term peaks, and cruise happily at around 300Wpc driving 8 ohm loads. This matches the attainment of significantly more costly amplifiers. Except for the modest channel separation at high frequencies, no test result gave rise to concern.
Well-built and well-finished, with effective if old-fashioned circuit protection, the MF2500 idled at quite low power and could be safely left on for extended listening periods. This musically well-balanced powerhouse represents fine dollar value, and wins a warm recommendation.
Footnote 1: In discussions with amplifier designers, I have suggested that a microprocessor could be engaged to manage a pre-heat routine that would accelerate warm-up from a cold start. This could even extend to the power-supply reservoirs, which several authorities maintain do not reach optimum sound and electrical characteristics until 20 or 30°C above ambient temperature.