Balanced R2R DAC ACSS 8xPCM1704 32bit 384kHz FPGA Amanero AccusiliconAudio-GD offers with its HE-7 MK3 a massive R2R DAC benefiting from a high-end design, based on very high quality components and an extremely well thought architecture. It features a conversion stage based on eight PCM1704UK DAC chips, discrete audio circuits, Accusilicon femtosecond clocks, FPGA-based clock management, isolated Class A power supplies in a regenerative system. Its main features are:Use of a servo stabilized discrete power supply to power the digital circuitry to ensure extremely low noise and low operating temperature. The clock and analog circuitry is always powered by a Class A regulated power supply for best sonic results. Possibility to enjoy an analog type sound restitution thanks to a selectable option on the front panel. Both the USB and HDMI inputs are equipped with isolators and two sets of linear power supplies to power the USB and HDMI modules separately before the isolators, to avoid source interference. The USB uses a bidirectional transmission isolator, which not only transmits the I2S signals to the FPGA processor, but also receives the synchronous clock signal sent by the FPGA processor. The USB interface itself is no longer equipped with clocks. The synchronous clock is used to achieve a more accurate transmission and to take the sound quality to a whole new level. The HDMI module is equipped with an independent isolator to improve the sound quality of the HDMI input. Display of the sampling rate of the input signal. The 10MHz clock input allows the use of an external clock generator. The FPGA operates using a parallel data processing mode. The I2S signal allows the transmission of data in series. Each data bit requires one clock cycle. One frame of data from the left and right channels requires 64 clock cycles. Stability over 64 clock cycles is therefore required. Parallel mode requires only one clock to transmit and process the 32bit data from the left and right channels, which greatly improves processing speed and stability. The I2S input data (USB and HDMI) is reorganized into two parallel 32bit data sets upon input. Once the SPDIF signal is demodulated, it is also sent to the next level of processing, via two parallel 24bit data sets. The parallel processing mode results in a clearer, more neutral sound with better dynamics and a more analog signature. The clock management design of this architecture makes the clock operation more table-like, which brings greater transparency and detailed sound. Analog regenerative power supplyThe integrated regenerative power supply works like an electrical generator. It blocks most of the disturbances in the power line, providing a clean power supply to the amplifiers and thus reproducing a neutral and analog sound. This architecture allows the AC current to pass through the regenerative input transformer, which is then converted to DC and directed to a Class A power supply and then to the regenerative wave generator and the gain stages. A regenerative regular wave generator produces a very low distortion 50Hz wave that passes through the gain amplifiers and output stages. The regenerative transformers produce a clean AC power supply that feeds the separate Class A power supplies for the left and right channel preamplifiers. The Class A parallel connection power supply has a very high input impedance to prevent interference through the PSU from affecting the signal, and a low output impedance with very high speed and linearity. R2R PCM1704UK DACThe R2R digital to analog conversion stage of the HE-7 MK3 is based on no less than eight PCM1704UK DAC chips. The R2R architecture does not convert the clock signal into an output signal, is not sensitive to jitter and provides a more accurate output signal than a Delta-Sigma DAC. The PCM1704UK chips support high resolution audio streams, up to 32bit 384kHz PCM. PCM1704UK DACs do not support DSD playback. DSD files must be converted to PCM from the source. FPGACPLDThe HE-7 MK3 integrates an FPGA and three CPLDs to separate the different configured circuits and avoid interruptions. The internal components are fully controlled by a complex software layer. A significant advantage of using an FPGA is the ability to easily upgrade it to improve the performance of the device, bringing flexibility to the device. In the HE-7 MK3 the high performance SPDIF interface of the FPGA replaces the traditional SPDIF interface chips whose performance is inferior to that of an FPGA. The reclocking process with FIFO design is applied to all inputs. This way, the output data remains fully synchronized with the clock signal to reject jitter. The HE-7 MK3 features 2x / 4x / 8x oversampling modes and built-in digital filters, as well as four different NOS modes, to suit your listening style and system. Discrete output stagesThe last signal stage is the analog output stage, which can have a considerable impact on the sound quality of the DAC. After digital to analog conversion by the R2R DAC modules, the analog signal is carried by fully discrete matched transistor output stages. The DC-coupled design uses no surface mount components. The output buffers are single-ended FETs. Two stages are connected in parallel to achieve a very low output impedance. All output stages operate in pure Class A, without any feedback, to achieve the purest and most transparent sound reproduction possible. Optimized power supplyThe HE-7 MK2 features an impressive power supply stage. It uses four high-performance R-Core transformers with a total power of 295W, as well as over 70,000F of filter capacitors to ensure a clean and stable electrical signal. 10 sets of discrete servo stabilized power supplies and 7 sets of ultra-high speed linear stabilized power supplies provide separate power to the digital circuitry. The Class A parallel regulated power supply supplies the regenerative stage and each analog section separately.

Galerie