I took Herb's comments into consideration as I went searching for a power amplifier to complement my Sugden LA-4 preamplifier. The $10,000 LKV PWR+ ranked top on my list. I put in a request to LKV Research's majordomo, Bill Hutchins, who dispatched the large, 50lb PWR+ to my Greenwich Village flat for evaluation.

I enjoyed the New Hampshire–made, 200Wpc (8 ohms) amplifier; it became my solid state/class-D reference power amplifier (footnote 1). The PWR+ presented an extravagant soundstage, bold dynamics, and clear, pure sound, with effortless, exemplary microdynamics and microdetail.

When I spied the shoebox-size LKV Research PWR-3 power amplifier ($3350) at Capitol Audiofest 2021, I wondered, Could this new class-D amp deliver a solid chunk of the sonic finesse of the PWR+ amp at a third of its price?

Design The PWR-3, like the PWR+, is a "solid state stereo power amplifier with Purifi (footnote 2) Audio's 1ET400A class-D power amplifier modules," Hutchins wrote to me in an email. The PWR-3 is much smaller, though: Its aluminum case stands just 3.5" high, 17" wide, and 13" deep. It weighs just 17lb. Its power output is rated at 175Wpc into 8 ohms, 360Wpc into 4 ohms, and 225Wpc into 2 ohms. Voltage gain is specified as 26dB, and input impedance is 200k ohms single-ended; there are no balanced inputs.

With the PWR-3, Hutchins wanted "to bring to a much lower retail price the essential characteristics that make the PWR+ sound as good as it does."

Hutchins explained that his design approach in the PWR+ and PWR-3 combines "class-A, zero-feedback voltage-gain circuitry implemented in a differential configuration with tightly matched JFETs" with two Bruno Putzeys–designed class-D modules. "Together, these circuits solve a problem that has bedeviled designers of feedback amplifiers for many years: the loss of effectiveness of negative feedback in the upper octaves of the audio spectrum. This loss degrades high-frequency sound."

Hutchins elaborated further on his use of negative feedback. "In an amplifier employing loop negative feedback, the amplifier's contribution (or lack thereof) to the character of its sound is determined largely by the characteristics of the feedback loop," he wrote. "Specifically, that loop controls (a) the amount and type of distortion you hear, (b) the amplifier's output impedance, (c) the amplifier's output gain, and (d) the variations in frequency response created by amp/speaker interactions."

When open-loop gain is increased, I learned, open-loop distortion rises. "At the upper frequencies, where feedback weakens," Hutchins wrote, "the distortion will start to re-emerge, which can result in a pronounced hardening and sterility of the sound. Also, the effects of amp/speaker interactions on frequency response and balance become less predictable and potentially problematic."

"Until very recently," Hutchins wrote, "there was only one solution to this problem: Avoid negative feedback. But doing so negates the distortion cancellation and impedance reduction that feedback provides at lower frequencies. Audio engineer Bruno Putzeys," of Purifi, "created a solution by devising a feedback loop that maintains high, very effective levels of negative feedback up to the top of the audio spectrum. Gone is the sterility of high-frequency sound. I use this solution in my Veros PWR+ and PWR-3 amplifiers.

"For the voltage-gain circuitry," he continued, "I use my own, zero-feedback circuit, [which] delivers lifelike sound [thanks to] careful design and component matching, employing differential (balanced) circuitry, carefully hand-matched, low-noise JFETs, folded-cascode voltage biasing, current-source control, very high headroom, and extensive regulation and filtering of the low-level power rails. For the high-current output stage, I use Putzeys's Purifi 1ET400A module, which implements his breakthrough," described above. "These comprise the essential elements that create the sound of the PWR+.

"In designing the PWR-3, I eliminated as much [hardware] as possible," Hutchins continued, "but kept these two essential elements—Putzeys modules and my zero-feedback circuit—and [thereby] preserved the sound character of the PWR+ at a dramatically lower price.

"Putzeys's groundbreaking work provides an intricate, carefully calculated feedback loop that ... maintains high open-loop gain and high, relatively constant feedback throughout the audio spectrum without causing oscillation. [In the PWR+ and PWR-3], this loop creates an output stage that sends the signal from my LKV-designed JFET circuitry to the speakers without adding or subtracting (virtually) anything."

Hutchins used the same Purifi class-D modules in both the PWR+ and PWR-3. In the PWR-3, they are powered by a Hypex Switch Mode Power Supply (SMPS), replacing three heavy, expensive toroidal transformers and some other circuitry found in the PWR+. Accommodating the switch-mode power supply required "operating at lower rail voltages"; other cost-cutting measures included "greatly reducing filtering and regulation of low-level power ... and [having] all circuitry on one PCB rather than four, [as] in the PWR+," Hutchins told me.

In addition to doing away with those toroidal transformers of the PWR+, Hutchins eliminated more than half the electrolytic capacitors found in the PWR+'s power supply. Also gone are the balanced inputs, two high-quality switches, the illuminated front panel power switch, and 16 costly, low-noise rectifier diodes.

Surely, stripping away all those high-quality parts from the design would degrade the sound.

"I wasn't sure I could use a [switching power supply] and still get the high level of performance I demand," Hutchins wrote. "High-frequency switching noise does nothing good for audio circuits." When he started working on the new design, that is what he saw. "Initially, switching noise was leaking into the voltage-gain circuit, where it showed up as noise and, to some extent, distortion. But trial and error, careful tweaking of internal shielding, cable routing, and component positioning solved the problem."

The PWR+ and PWR-3 both use JFET-based circuits for voltage gain and for the driver circuitry that feeds the Purifi modules. The PWR+ has 72 hand-matched JFETs; the PWR-3, 40. But there are many similarities between Hutchins's offspring, and many shared components: JFETs, PNP and NPN transistors, 1% metal film resistors, high-quality film and electrolytic capacitors, red LEDs (used as voltage sources), various switches, and voltage regulators.

Footnote 1: The Shindo Haut Brion remains my tube reference.

Footnote 2: Purifi is a Danish audio firm building amplifiers, amplifier modules, and transducers. It was founded in 2014 by audio engineer Bruno Putzeys and direct-switching pioneer Lars Risbo.

Log in or register to post comments