On What’s Happening Now…

By Bryan Geyer

It’s clearly apparent (well, maybe not to those that cling to vacuum tube technology) that major change is ongoing in the design and configuration of home audio systems. Obviously, one of the hot trends involves the arrival of class D high power, high efficiency amplifiers—mostly as replacements for the classic A/B bias amps that have been dominant for ~ 70 years. The class D drift began a few years ago with the release of various new ICEpower and Hypex advanced class D high power amplifiers that were sold, in “open chassis” form, directly to the established OEMs (original equipment manufacturers) that make power amplifiers. The intent was that these traditional OEMs would add some proprietary peripheral content, mainly in the form of their own enclosures and preferred input/output terminations, and then market those class D amplifiers as their own finished product.

This course represents major change. All OEM participants that buy the same class D power amplifier module (complete with captive switch-mode power supply) would then be reselling the same basic product, as supplied from the same common source. Although their finished retail products would likely exhibit distinctive external shape/size/weight, and (maybe) sport unique in/out terminations and/or special peripheral features, they’d all use the same class D engine, hence offer the same native performance and same sonic signature.

The compelling benefit that drives this new marketing concept is high efficiency. A modern class D stereo power amplifier can deliver full power output to the load with efficiency on the order of some 86%. As a result, Class D bias can dramatically reduce internal waste (heating), and thereby eliminate much of the massive heat sinking that’s essential when operating in conventional class A/B (and class A) mode. The attendant saving in cost can be appreciable, and the related size reduction can be dramatic.

There’s nothing conceptually new about class D; it’s been around since the late 1950s. What is new is that a very talented and highly select coterie of creative audio engineers—largely concentrated in Denmark and The Netherlands—have now created new class D high power amplifiers that can deliver this high efficiency advantage without any of the attendant performance penalties that formerly plagued class D application. Finally, we have elegant, effective, and efficient class D stages that can output 300-400-600 Watts/channel without the noise and distortion artifacts that always limited this technology to a benign “yes, it’s interesting” reality. The key Danish and Dutch engineers responsible for this breakthrough currently represent three independent competitors: ICEpower and Purifi Audio, both located in Denmark, and Hypex Electronics in The Netherlands. These outfits are the principal innovators. Some other nearby companies are immersed in class D product development; e.g., Lyngdorf Audio, in Denmark, and Nord Acoustics, in the U.K. The latter offers a wide assortment of modular high power amplifiers (for OEM resale) that are exclusively based on the latest advanced class D products from Purifi Audio and Hypex Electronics. (Nord appears to be an authorized distributor.)

There’s also now a smaller new class D innovator here in the U.S.: Orchard Audio, of Succasunna, NJ, some 40 miles west of Manhattan. Orchard specializes in select class D power technology using gallium nitride (GaN) materials. GaN devices exhibit a higher operating temperature capability than conventional silicon semiconductors, so I presume that Orchard Audio’s unique offerings will be primarily optimized for non-consumer applications.

The engineers representing these key companies have been actively seeking a solution to the class D high power puzzle for more than a decade, but it has only been within the last few years that success seems assured. The class D challenge is demanding. Every step of the exacting process requires special expertise. Complex interactive design and assembly includes the necessity to identify, specify, obtain, and test unique semiconductors with select characteristics, as well as the use of special passive components like the novel LC filter inductors evident on every class D power amplifier PC board. At present, no manufacturer of high quality audio power amplifiers can replicate the excellence of the modular class D high power amplifiers available from these dominant Danish and Dutch providers. Indeed, nobody other than Orchard even comes close.

At present, various traditional OEM producers of power amplifiers buy these new modular class D amplifiers directly from the Danish and Dutch makers, or from their authorized distributors. These OEMs then add the requisite housing and input/output interfacing, plus maybe some peripheral features, and market the composite package under their own trade label. That course has now proved feasible for several well known U.S. participants that traditionally sold class A/B bias power amps of their own design. Marketing proprietary product that’s been created and provided by an outside sole-source patent-holder is a new endeavor for most of these established OEMs. This sort of selling is probably common in the overall scope of worldwide consumer electronics, but it’s new and novel in the more restricted confines of the hi-end audio power amplifier business.

The ICEpower 700AS2 amplifier module is a good example of how this cooperative concept can succeed, at least for ICEpower and for the consumers that ultimately acquire their product. The traditional OEM houses that currently buy, repackage, and resell these advanced class D power amplifiers might feel conflicted about this new means of serving their primary market, but they really have no other choice. There’s now a hot demand for top quality class D high power technology, and the only way that these OEMs can serve that need is to ship the existing Danish and Dutch product—it’s their only way to remain relevant.*

Within the past 4 years, several prominent U.S. producers of power amplifiers have released new class D amps that feature the same ICEpower 700AS2 engine. As a result, those products all exhibit identical performance and share the same sonic character. For example, Legacy Audio announced their Powerbloc2 “dual mono” power amplifier in early 2016 (it’s still in current production). It has a rated output capability of “325 Watts per channel x 2 @ 8 ohms, 650 Watts per channel x 2 @ 4 ohms”. Then, in late 2017, Parasound Products announced their ZoneMaster 2350 stereo power amplifier, with outputs rated at “350 watts x 2 @ 8 Ω” and “600 watts x 2 @ 4 Ω or 2 Ω”, as measured per “RMS, both channels driven, 20Hz to 20kHz”. The cited deviation between these amplifier’s power ratings is insignificant: 350 Watts is +0.3dB > 325 Watts, and 600 Watts is -0.3dB < 650 Watts. There’s simply no real difference. There are obvious variances in some peripheral features—stuff unrelated to power output—that might make one amp more or less optimal than the other for a given installation, but both amplifiers will deliver identical performance when driving the same speakers in the same listening room, and both will sound precisely the same when driven by an identical signal.** Given this congruence, it’s hard to explain the persistent (since 2017) list price gap of ~ $500 between these two closely related components, but I’ll do that now; see footnote.†

It’s reported, in some discussion forums, that both Wyred-4-Sound and PS Audio also utilize this same ICEpower 700AS2 amplifier module (or its single channel direct derivative). I’ve not tried to confirm this as fact because there are surely other OEM customers as well. ICEpower has apparently created a fine product, and the audio buffs that use it seem enthused and satisfied. Early feedback indicates reliable service life. Class D now looks literally as cool as promised, and it presents compelling possibilities—like a practical way to double your power output and consume half as much space—while spending no more than you’d normally allocate for half as much power and twice the size.

Be wary when shopping. Research class D high power amps with care, and question your judgement if your ears sense a brand-related advantage that can only be traced to random differences in exterior hardware. Amplifiers that use the same internal engine will always deliver the same sonic performance. Ears are not always so reliable as some think, and listening trials are never comparable when demonstrations involve different rooms and different programs. Focus on the published specifications, and on the detail that’s presented therein. Assure that the stated measurements are fully comparable, and that the results represent equivalent conditions. Beware when vital data is omitted. I feel that sloppy specs reflect a corporate focus that’s inconsistent with a healthy engineering culture.

This ongoing class D upheaval is big, but it’s not the only new audio trend afoot. Following are some of the other recent innovations that will eventually change the way we listen.

LOUDSPEAKERS: There’s a persistent trend away from big all-in-one-box loudspeaker systems and toward the use of smaller mains that are coupled with paired subwoofers.†† This evolution is the result of improved subwoofer design in combination with a new awareness of the ways in which we can tame the acoustic limitations implicit in typical home-size listening rooms. Multiple subwoofers can be utilized to effectively achieve partial cancellation of the reflected modal bass over large portions of the listening area†*. Two subwoofers will work well; more subs can be more effective—space permitting. The subs’ low bass output will naturally be ~ 180˚ out-of-phase with the back wall modal bass, so significant cancellation will result when those wavefronts converge. This can potentially improve near-field accuracy and generally enhance the perception of realism.

ACTIVE EXTERNAL CROSSOVERS: Better bass is isn’t the only benefit that paired subwoofers can bestow. After the paired subs are installed, it’s then possible to achieve a cleaner, clearer mid-range response by inserting an active crossover controller at a point in the signal path that's immediately before the power amplifier stage. This will allow you to dictate what feeds to the self-powered subs and what feeds to the main power amplifier + main speakers. It's always those big fat bottom bass notes that perpetually smear midrange clarity. When you divert the low bass and send it directly to the subwoofers (where it belongs), the main speaker’s mid-woofer drivers are then free to handle the upper bass and midrange frequencies independently, without disruptive low bass modulation. You'll get a clearer, more articulate midrange. The improvement is especially apparent when listening at high sound pressure levels. This midrange benefit is one of the most significant assets that a properly managed subwoofer setup can bestow, and many experienced listeners feel that it’s more important than enhanced bass.

BG (August 24, 2020)

*A well known and highly respected U.S. microelectronics company, Analog Devices, conducted intensive research and product development work on class D amplifiers over the course of the last two decades, and they published numerous technical papers for open peer review. Analog Devices also created and sold numerous class D products of their own design. However, their focus was the market for military and space research applications, and primarily involved high efficiency at modest power output, e.g. 5-10-20 Watts. Consumer electronics and power outputs ≥ 300 Watts at audio frequencies was never their objective.

**Legacy Audio’s Powerbloc2 and Parasound Products’ ZoneMaster 2350 stereo power amplifier exhibit near identical input sensitivities. The former specifies 2.19Vrms input for 325 Watts out across 8Ω (calculated voltage gain = +27.4dB), while the latter specifies 2.0Vrms input for 350 Watts out across 8Ω (calculated voltage gain ≈ +28.4dB). The related gain ∆ of just 1dB is simply inconsequential; it likely traces to calculated round-offs in combination with the manner in which the respective gains were originally derived, or measured, or expressed. In any case, it’s apparent that both Legacy Audio and Parasound Products are using the same active input, namely the amplifier’s basic “on board” input, as provided by ICEpower. There’s no evidence of any auxiliary active buffering.

†Legacy Audio, like Sanders Sound Systems, is primarily recognized as a producer of ultra hi-end loudspeaker systems. In both cases, these independent companies utilize transducers that tend to be lower in efficiency than what’s generally characteristic with conventional electro-magnetic drivers. (Legacy uses AMT type midrange drivers and Sanders builds planar electrostatic speakers.) As a result, both companies need on-site power amplifiers with very high output capability to properly demonstrate (and stay well clear of clipping) their proprietary loudspeakers to full advantage. The best way to assure this result is to utilize their own (house brand) high power amplifiers—amps that they know to be absolutely optimum for use with their speaker systems. The Sanders solution is a custom-designed proprietary amplifier with a high-bias class A/B final stage and multiple paralleled output devices in lieu of the usual (power-limiting) protection circuitry. Conversely, Legacy Audio’s Powerbloc2 is a high output class D amplifier that features the modern ICEpower 700AS2 engine. (As a consequence, the Powerbloc2 can be far smaller in size/weight than the Sanders amp, which requires conventional heatsinks.) It’s quite natural that customers who purchase a high quality speaker system from one of these sources would also want to buy the same power amplifier that was used when their speakers were demonstrated. That’s just good sales synergy, and a valid tactic that’s often linked to a nice discount on the complementary product. So maybe (?) Legacy’s Powerbloc2 amplifier is priced with that expectation; ask ’em.

††There are very few full-range floor-standing loudspeaker systems with woofers that are able to match the performance of a modern “mid-fi” self-powered subwoofer at frequencies from 20Hz to 50Hz. Nearly all of the very best floor-standers sag from 50Hz down, which is not surprising, given the fact that they’re also expected to reach 600Hz to 1kHz or more. Conversely, good subs are intended purely for 20Hz to 100Hz bass, a demanding but restricted niche that they’re specifically designed to serve, and where they consistently excel.

†*Refer pp. 234-262 of Floyd Toole’s “Sound Reproduction”, 3rd edition (2018), Routledge, ISBN 978-1-138-92136-8.