Jeff Rowland Design Group Model 2 power amplifier Page 2

The output section features four individually selected bipolar transistors per phase per channel, totaling 16 transistors overall. A sophisticated temperature-sensing circuit optimizes bias. The Model 2's output stage operates in class-AB rather than class-A. Jeff Rowland believes that a well-designed class-AB design can perform as well as a class-A, without the penalty of high power consumption and heat generation. In common with many high-end amplifiers, the Model 2 has no overall loop feedback.

All amplifiers have transformers in the power supply, and tube amplifiers have transformers on the output side (except, of course, if they're the output-transformer-less variety); the Rowland Model 2 is unusual in having transformers in the signal path on the input side. The input transformer features proprietary winding techniques and high-percentage nickel core materials. The claimed advantages of using an input signal transformer include total isolation of ground currents, effective RFI filtering, pure resistive loading, response bandwidth to 200kHz, a common-mode rejection ratio (CMRR) greater than 60dB at 20kHz, extremely low distortion (±0.01% at 20Hz), and insensitivity to a wide range of source impedances.

The power supply itself features toroidal transformers with lower flux density than conventional designs (1.5 Tesla at 50Hz, if you must know); 260,000µF capacitance; and separate regulated positive/negative supplies for input, output, and "housekeeping" functions. Rowland points out that the advantages of balanced operation include a 50% reduction in power-supply voltage, which permits more reliable high-speed operation.

One of the problems in reviewing audio equipment is that it's sometimes difficult to be sure of the extent to which the "sound" of a particular component represents an interaction between the component and the rest of the system, rather than being the sound of the component per se. To guard against errors of this sort, it's important to use matching components that are as neutral and accurate as possible so that deviations from the absolute sound can be attributed to the Device Under Test.

Still, if an amplifier being tested is disappointing, one wonders if it would have performed better with a different set of associated equipment (footnote 3), or if one found the right tweak that would allow it to sing. Of course, if an unfavorable interaction is suspected, the reviewer will try swapping components, changing cables, and tweaking the setup, but there's a practical limit to this, and one can question whether the typical consumer should be expected to follow this path. (Actually, my feeling is that, if a component needs to have everything tweaked to a fare-thee-well, there's something wrong with the design.)

From the reviewer's perspective, the best situation is one in which the component sounds so good right off the bat that there's little or no need for tweaking or substitution of components. This was the case with the Rowland Model 2. Listening to track 3 of the first Chesky Jazz Sampler with the Model 2 in the system for the first time (after a few hours' warm-up), my immediate impression was of a level of detail and transparency that went beyond anything I had heard previously in my system. It was an impression that was to persist throughout the extended evaluation period.

Now, it may be that I happened to hit on a particularly synergistic system interaction (Rowland and Dunlavy are both located in Colorado Springs, so it could have something to do with that mountain air), but I wasn't about to search for components that would make the Model 2 sound worse!

In any case, my impression of the Rowland Model 2 was the same with any of the three preamps, two D/A processors, and the two sets of loudspeaker cables that were part of the system at one time or another. I also know that it wasn't a matter of the Model 2 somehow compensating for loudspeaker colorations: the Dunlavy SC-IV is as neutral and accurate as just about any speaker out there, with a linearity of impedance, amplitude, and phase response that makes it ideal for identifying the sound of other components—especially amplifiers.

Rowland electronics have had a reputation for sounding "dark," with a top end on the soft side of neutral. A couple of years ago I had a sample of the Rowland Consonance preamp to compare with some preamps I was reviewing; although I liked the Consonance quite a lot, I thought it did sound a bit soft and forgiving.

Footnote 3: Given a negative review, Standard Manufacturer's Comment No.2 is: "It's obvious that our product allowed the reviewer to hear, for the first time, the colorations of his system."