From a 10-watt tube amplifier to UcD, NCORE, and Eigentakt, his work has helped change both amplifier circuits and the industry's view of Class D.
Few names are harder to leave out of a history of modern high-performance Class D than Bruno Putzeys.
Many listeners have never owned a component bearing his name on its front panel. They may nevertheless have heard products built around his work. UcD helped extend Class D's role beyond efficiency into a wider high-fidelity market. NCORE raised expectations further and found its way into established high-end brands. With Purifi's Eigentakt, he continued exploring control theory, output feedback, and error suppression. Beyond amplifiers, his work includes converters and active loudspeakers at Grimm Audio, electronics for Mola Mola, and system-level loudspeaker design at Kii Audio.
Read together, these projects suggest a question pursued over three decades: how can the equipment impose less of its own character and preserve more of the music?
At sixteen, an old tube amplifier upset the expected answer
There is something of a family story behind Bruno's interest in audio.
In a Sound & Vision interview, he recalled that his maternal grandfather and great-grandfather had manufactured radios in the 1930s. His grandfather later predicted that a powerful amplifier would one day fit inside a matchbox. Bruno never met him, but the prediction became part of the family inheritance he carried into electronics.
The more immediate turning point came when he was about sixteen. Until then, he had been deeply interested in computers and programming. His father and friends compared a modern transistor amplifier with a small, old tube amplifier using EL84 valves. In a later account, Bruno described the tube amplifier as a roughly 10-watt unit. To their ears, it outperformed the newer amplifier.
This was a remembered listening experience, not a controlled comparison establishing the superiority of tubes. What mattered to Bruno was the surprise. A result he had not expected made him want to understand what was happening.
That response provides an early clue to the approach seen in his later work. A conflict between expectation and experience could become a reason to examine the assumptions more closely: was the problem fully described, had the measurements missed something, or was the supposed theory only a simplified approximation?
The episode appears in his interviews with SoundStage! Ultra and Sound & Vision.
Philips gave him a month—and a career began to take shape
Bruno studied at the National Technical School for Radio and Film in Brussels, graduating with distinction. His thesis concerned Class D amplification and was sponsored by Philips. After he graduated in 1995, Philips offered him a job.
Joining a large electronics company did not immediately mean working on the ideas that most interested him. He spent the initial period refining the conventional amplifier circuits used in the compact black stereo systems popular at the time. In later interviews, he described the cost-optimization work as deeply frustrating. He kept trying to persuade management to give Class D a chance.
A small budget connected with the television business eventually created that opportunity. Another Philips team had been developing a Class D IC without achieving the result needed. Bruno was given a month to see whether a discrete circuit could do better.
He delivered a small, inexpensive, self-oscillating 25-watt prototype within that deadline. IEEE Spectrum's account says three weeks; his later Sound & Vision recollection describes thirty days. The accounts differ on some details, but agree on the short deadline, the 25-watt prototype, and its importance to the work that followed.
One success bought time for the next: a 100-watt version, another project, and gradually more freedom to investigate unfamiliar ideas. By around 2000, he recalled having strong support from the laboratory's management, including tolerance for his unusually direct opinions.
He later chose that little 25-watt amplifier as a personal career highlight. The compelling part of the story is not simply that a young engineer challenged a longer-running project. He turned the challenge into working hardware within a severe deadline, giving the company a reason to let him continue.
The episode is documented in IEEE Spectrum, SoundStage! Ultra, and Sound & Vision. Their retrospective component-cost estimates also differ; no single cost figure is needed to explain the achievement.
UcD mattered for more than its schematic
In 2001, while working at Philips, Bruno developed UcD: Universal class-D.
Its significance went beyond operating the power devices as switches. UcD includes the output filter in its control loop and takes feedback from the output delivered to the loudspeaker. Its self-oscillating architecture helps provide consistent control across the audio band. The engineering objective is accurate output with low distortion and limited sensitivity to changing loudspeaker loads. Hypex's UcD overview
There was also a personal reassessment behind this development. Bruno has described his earlier suspicion of negative feedback. He had built a zero-feedback power DAC that processed DSD directly, a design of which he remained technically proud. Yet he reported preferring the sound of his UcD designs.
That listening judgment belongs to Bruno's own account; it is not an independent test performed by Maroon. Its importance to this story is the conclusion he drew from it. A design principle such as avoiding feedback or keeping the path entirely digital was not, by itself, a useful answer to an engineering problem.
The starting point had to be the problem itself. If practice seemed to contradict the model, the model needed closer examination. If a standard measurement did not reveal a suspected behavior, the measurement method needed further thought.
He described building successive UcD prototypes and measuring them until their behavior agreed with the intended theory before sitting down to listen. The result became a design that could be licensed, manufactured, and used by other companies. His later article, The F Word, examined feedback terminology, distortion, and the importance of loop behavior across frequency.
See The F Word and his SoundStage! Ultra account of UcD's development.
Hypex turned the circuit into an industrial platform
An invention needs more than a successful prototype to influence an industry. It must be manufactured, delivered, adopted by other designers, and supported over time.
Hypex founder Jan-Peter van Amerongen recognized UcD's potential. Hypex's company history records the introduction of Philips-licensed UcD in 2003. Bruno recalled handing over schematics, a sample layout, and notes; Hypex then produced a working amplifier and brought it into manufacture.
The two contributions complemented each other. Bruno developed the circuit architecture. Hypex made it a practical platform that other people could obtain and build around.
Bruno subsequently left Philips to work at Hypex. UcD reached both DIY builders and commercial manufacturers. IEEE Spectrum's profile described dozens of amplifiers across different price levels using the modules and portrayed him as an important figure in Class D's growing high-end presence.
The same profile includes a lighter moment from the 2006 AES convention. Someone standing in front of Bruno began talking enthusiastically about Bruno Putzeys without recognizing him. Recognition arrived only when the visitor noticed his name badge. It is an apt image for a designer whose circuits and name were often better known than his face.
Hypex's history and the IEEE Spectrum profile provide the background to this period. Accounts differ on the precise year of his move to Hypex, so this feature does not assign a single date to that transition.
NCORE raised the expectations again
UcD offered a successful platform on which to continue working. NCORE took the investigation further.
Bruno's explanation begins with a mathematical description of the stable operating state—the limit cycle—of a self-oscillating amplifier. That gave him a tool for increasing audio-band loop gain while addressing stability. Switching frequency and available bandwidth impose constraints; stronger error suppression has to coexist with predictable behavior near clipping and a clean return to normal operation.
This was a control problem, not simply a matter of fitting more components or declaring that more feedback must be better. The loop and the power stage both needed careful engineering.
He described NCORE as an evolution of UcD, while emphasizing how substantial the result could be. Hypex introduced NCORE in 2011. By the time of Sound & Vision's 2017 profile, the publication identified its use by ATI, Marantz, Jeff Rowland, and Bel Canto. The technology had become part of products from companies with very different histories and audiences.
That adoption matters to the larger story. Class D was increasingly being considered on its actual performance in full-range high-end systems rather than being confined in discussion to subwoofers or applications where efficiency dominated every other consideration.
Bruno himself was clear about the difficulty. In the interview, he identified efficiency and the resulting power density as Class D's inherent advantages. High audio performance still had to be earned through demanding engineering. His case for the technology was informed by its difficulties, not by overlooking them.
Read his SoundStage! Ultra discussion of NCORE's control and stability, the Sound & Vision interview, and Hypex's NCORE overview.
Following the music through the whole playback chain
Calling Bruno only a Class D amplifier designer leaves out a substantial part of his work.
At Grimm Audio, his projects have included high-performance converters and the LS1 active loudspeaker system. Mola Mola brought together NCORE amplification, analog circuitry, conversion, and front-end functions in products for home listeners. At Kii Audio, the team's work combines amplification, DSP, drivers, and control of acoustic radiation to address the relationship between a loudspeaker and its room.
These projects can be read as investigations of different parts of the same chain: how an analog signal is converted, how a power stage drives a load, how a loudspeaker radiates, and how a room changes the sound at the listening position.
That is the editorial connection we draw between the projects. It does not require treating every product as one person's work, or imagining that accurate amplifier output alone settles the questions of loudspeaker and room behavior.
Purifi's team biography summarizes his work across UcD, NCORE, Grimm, Mola Mola, and Kii. The Kii Audio team page places his amplifier background alongside the company's loudspeaker development.
Purifi: starting again in an office nicknamed Siberia
By the time he co-founded Purifi, Bruno had an established reputation in Class D. What makes this next chapter interesting is the willingness to revisit familiar questions and return to the mathematical foundations.
Purifi's current team page identifies Bruno Putzeys, Lars Risbo, and Peter Lyngdorf as co-founders. Bruno and Lars are listed as CTOs and Peter as an investor. Claus Neesgaard is listed as co-owner and CEO, while Carsten Tinggaard is co-owner and COO. Those roles distinguish the founders from other important people in the business.
The company's account of its early premises is far removed from the image of an elaborate high-end laboratory. Inherited desks were cut down to fit the available space, and the team worked close together. The building became so cold in winter that they nicknamed it Siberia.
Eigentakt's development returned to a difficult question: how could stronger error correction coexist with predictable behavior in a self-oscillating Class D amplifier? Purifi describes how simply increasing loop gain could move the onset of instability to lower signal levels, while conventional analysis did not adequately capture the combination of PWM sampling effects and small- and large-signal dynamics.
The team put practical circuit work aside while developing a more complete mathematical description, then used numerical optimization to refine the loop.
This is a significant part of Bruno's contribution: helping turn behavior that is difficult to navigate through rules of thumb into something that can be modeled, calculated, measured, and implemented.
Purifi's work combines feedback encompassing the output filter, investigation of magnetic hysteresis, and attention to large-signal stability. It is a team effort. The company's own retrospective specifically credits Bruno with leading the collaborative mathematical work on Class D amplification, while also describing the separate expertise required to turn research into products and a functioning business.
Sources: Purifi's company retrospective, team biographies, and Eigentakt loop research.
Measurements, listening, and a room full of music
It is tempting to describe Bruno simply as a measurement-led engineer. His interviews describe a more interesting relationship between measurement and listening.
He advocates measuring and refining a circuit before evaluating it by ear. An unexpected listening result can then prompt another look at what was measured and how. Listening becomes a check on the completeness of the engineering investigation, rather than a substitute for it.
IEEE Spectrum's portrait of his home and workplace gives that method a human setting. The laboratory contained oscilloscopes, analyzers, transformers, and prototypes. Upstairs, the living space held his audio system and recordings ranging from Johnny Cash and Art of Noise to Jacques Brel and Senegalese singer Baaba Maal. The writer noted that there was no television.
It is an image of technical work organized around a lasting interest in music. The instruments downstairs and the records upstairs belonged to the same life.
There is also a recognizable independence in the published stories: curiosity provoked by an unexpected tube-amplifier comparison, persistence at Philips, a willingness to challenge received views of feedback, and a reassessment of his own zero-feedback work.
The lesson we take from those accounts is that confidence can include the ability to revise an answer when the evidence changes.
IEEE Spectrum's profile and the SoundStage! Ultra interview describe these working habits and surroundings.
His influence changed the questions being asked
A list of UcD, NCORE, and Eigentakt is a useful timeline. On its own, it does not fully explain Bruno Putzeys' influence.
His work helped move the conversation about high-end Class D toward more specific questions. How much control does the loop retain throughout the audio band? Are the output filter and load interaction included in error correction? What happens near clipping? How do noise and distortion behave across the conditions in which an amplifier will actually operate?
Those questions are more productive than trying to decide a product's quality from an operating-class label or a fixed preference for more or less feedback.
For Maroon Audio, using Hypex and Purifi platforms connects our products with this engineering history. The modules supply the power-amplification foundation. Maroon integrates analog input buffering, gain matching, power supplies, cooling, and mechanical construction into complete amplifiers. The two module suppliers remain independent companies, and a designer's reputation does not replace the work required in a finished product.
Our respect for Bruno is grounded in the method visible in his work: identify an error, understand it, and keep looking for a better solution. It offers a more useful standard than surrounding sound with mystery or accepting that a circuit is already good enough.
His lasting achievement is helping Class D earn a place at the center of high-end audio through its engineering performance.
Sources and editorial note
IEEE Spectrum: Bruno Putzeys—The Sound of Music
SoundStage! Ultra: Searching for the Extreme, Part One
Sound & Vision: Bruno Putzeys—Head of the Class (D, That Is)
Hypex Electronics: Company History
Purifi: Building Purifi, from a Cold Office in Siberia
Purifi: Combating Hysteresis Distortion, Part 1
This feature draws on published interviews, company biographies, and technical writing; it is not an interview conducted by Maroon Audio. Listening anecdotes are attributed recollections, and assessments of influence are editorial judgments. The linked sources provide the basis for technical attributions and historical details.
More in this series
How Much Amplifier Power Do You Really Need?
How Modern Class D Amplifiers Have Improved
Class D Does Not Mean Digital: Analog Inputs and Preamp Matching
Understanding Hypex: From UcD to NCORE and NCOREx
Understanding Purifi: Why Eigentakt Focuses on Error Control
Browse the System & Craft Journal