D-2 Ver. 2 D/A Converter

Premium Silver/Premium Black

Version 2 is a dramatic evolution based on the knowledge gained from the 3 series.

Twin mono construction with New Type-R circuit

The greatest feature of the D-2 and the key to its sound quality was its large-scale non-feedback discrete amplifier, on a par with that of a power amplifier. The D-2 Ver.2 incorporates the latest version of this amplifier circuit, the New Type-R, developed for the E-3, which has been refined for use with DACs. This achieves a twin mono construction, with the power supply and other components located on the left and right.

Texas Instruments LMX2594 DDS

For the DDS, we have adopted the PLLatinumTMRF synthesizer LMX2594 from Texas Instruments, which boasts world-class specifications of -236dBc/Hz figure of merit and 45fs jitter (femtoseconds, where a femtosecond is 1/1000 of a picosecond), for the first time in an audio DDS. This ultra-low jitter DDS enables overwhelmingly clear and profound spatial depiction, whether using a highly accurate external clock or the built-in TCXO.

Supports 10MHz external reference clock input

The internal clock uses a high-performance TCXO boasting the highest level of close-in phase noise (-121dB at 10Hz offset). To avoid jitter caused by the IC, a high-frequency mechanical contact relay is used to switch between the internal and external clocks. When using an external clock, the internal clock is powered down to stop oscillation, eliminating mutual interference.

The external clock terminal uses a new SMA terminal with excellent high-frequency characteristics and a reinforced structure. This eliminates the need for the conversion plug previously required when connecting the X-3 clock generator. Using the X-3 dramatically improves the sound quality across the entire frequency range.

Resistor IV circuit using ES9038PRO

Two ES9038PRO amplifiers, boasting a powerful output current of 120mA, are used per channel. This allows for IV conversion using a single low-resistance resistor. This eliminates TIM distortion, a problem with feedback-type IV circuits, resulting in a completely non-feedback DA converter. The ES9038PRO also uses synchronous mode, which allows for the highest sound quality.

NOS mode adopted

In addition to the conventional FIR oversampling mode, we have also adopted the NOS (non-oversampling) mode, which can be considered a non-feedback mode in the digital domain. In this mode, the pre-echo and post-echo observed in the impulse response of FIR oversampling do not occur.

Impulse output waveform (analog output) in FIR oversampling filter

Pre-echo and post-echo are artificial "sounds" created by calculation from previous and subsequent data in order to interpolate data. This makes waveforms such as sine waves appear smoother, but the calculation algorithm can change the sound quality and add ambiguity to the time axis.

Impulse output waveform in NOS mode (analog output)

This waveform can only be achieved in combination with a non-feedback discrete amplifier that does not use analog filters. Because musical waveforms are a series of impulse waveforms of different heights, the NOS mode eliminates the ambiguity of time axis information, bringing a more realistic and natural sound quality and sense of space to the human ear, which is very sensitive to the time axis.

*DSD does not enter NOS mode.

Equipped with a 400VA bifilar-wound high-regulation toroidal transformer with 8 secondary windings

The AK4100 employs a power transformer with eight secondary windings that boasts regulation on a par with that of a high-power amplifier. Separate rectifier circuits for each stage prevent interference. Furthermore, unnecessary vibrations from the power transformer are vented to the exterior of the housing via spikes directly below the transformer.

Noise suppression without relying on filters

The D-2 uses a four-layer board with 70 micron copper foil and optimizes the artwork on the printed circuit board to eliminate noise at the source. Thorough noise suppression measures are implemented without relying on ferrite filters or other types that affect sound quality, and unwanted radiation is kept below the noise floor level in all modes.