HIBIKI SDS DAC декодер Система декодирования строки Fully Discrete SDM1024 5bit sdm SDS/BDS поддержка WCK/MCK часы Класс AB PLL включен.

HIBIKI SDS DAC декодер Система декодирования строки Fully Discrete SDM1024 5bit sdm SDS/BDS поддержка WCK/MCK часы Класс AB PLL включен.

Description

About configuration:

1. The classic version uses dual talema high-power ring cows, the digital and analog are powered separately, and the sound is warm and durable.

2. The machine is equipped with high-purity silver-plated wire to increase the amount of high-frequency information.

3. Six Mondorf electrolytic capacitors, each worth 160 RMB, provide transparent and brilliant high frequency.

4. The energy storage capacitor adopts the top elna style, and the sound is fast and charming.

5. Each SDS is equipped with three sets of two PLL architectures: LC wideband phase-locked loop provides a warm sense of hearing, and XC narrowband phase-locked loop provides accurate sound performance

6. SDS is compatible with both wck and mck clock standards.

7. SDS allows simultaneous input of 44.1/48k equivalent to dcs, and the decoder switches the sampling frequency by itself.

8. Two SDS exclusive decoding modules, four-channel true dual-mono true balanced architecture.

9. Each decoding array uses 36 switches with redundant switches to achieve true 5bit decoding.

10. Each channel decode array uses the LT6655 ultra-low noise voltage reference.

11. Decoding the analog output part, all use the low-noise transistors of the American diodes brand.

12. The analog circuit uses a completely feedback-free structure, and less feedback results in faster transients and more natural sound performance

Signature version configuration:

1. Upgrade high-power custom ring cow, separate power supply for digital and analog, and huge sound field.

2. The inner cable of the power supply is upgraded with pure silver cable to optimize the freshness of the sound.

3. Use the SDS decoding module of the top American Rogers ceramic substrate to improve the analysis quality.

4. The decoding module uses the fast clear capacitor to replace the elna, and the sound is more transparent.

5. The signature version is in stock from time to time, please contact the shopkeeper for inventory

BDS2 is affected by the price increase of chips, and the price has increased in an all-round way.

1. Optional SRC up to pcm3072

2. Native 1bit dsd decoding, support up to DSD1024 without converting PCM

3. Hardware dynamic shuffle can reduce msb error noise and distortion

4. The second-generation hardware deglitch, not only the zero-crossing part is properly handled, but the lsb part is equipped with a deglitch circuit, which is 64 times lower than the traditional r2r

5. The switching timing is precisely matched, the resistance array skew is controlled within 1ns, and the left and right channel errors can also be controlled within 1ns

6. Built-in three sets of high-precision PLL circuits ReVC1 and ReVX1, which can be switched between one-level or two-level phase-locked loops at will.

7. Support external dual clock input: compatible with wordclock and masterclock at the same time.

8. Continue the sound characteristics of BDS: full, thick and rich, which enhances the performance of the sound field and makes the sound more stable.

9. Compared with BDS1, it upgrades dual cows, cancels the DC module to improve the dynamic, adds DSD and DOP64 support, supports SRC up to PCM3072, and supports dual clock input.

The standard version uses dual talema transformers, FR4 military circuit board decoding module with silmic2 capacitors, and silver plating machine inner wire.

Upgraded custom transformer, ceramic decoding module and fast clear capacitor, with sterling silver machine inner cable.

HIBIKI SDS DAC decoder String Decoding System Fully Discrete SDM1024 5bit sdm SDS/BDS support WCK/MCK clock Class AB SDS PLL DAC

SDS I2S can only be compatible with PS Audio

The full name of SDS machine is String Decoding System. Because the decoding structure of SDS is a multi-bit resistor string, so it's called String.

SDS comes with an upscaling function, 5bit sdm 1024, no matter what is the sound source, you can upscaling with just one click. Currently 5bit sdm 1024 is the first innovation in the world.

Upscaling of SDS:

Turn on PLL means Upscaling turned on

Don't open PLL means NOS

On the current market, there is little difference in the effect of clock boosting on decoders that support external clocks. However, SDS has a special design of self-developed PLL, which greatly improves the clock boosting, and it is different from the current mainstream. We hope that the clock will with higher utilization.

It support both WCK and MCK of clocks, also support dual clock input.

1. Modulator

a) Straight-up SDM1024, 5bits

As the DA core, we use a high-order sigma-delta modulator, both pcm and dsd are directly modulated to a 5-bit 1024 times rate. Generally speaking, modulation of 1 bit is DSD, but five bits are inherently more than 1 bit of information, which is 2 to the 5th power, which is 32 times, and the signal-to-noise ratio is also 30db. Generally, the internal modulation frequency of the DA chip is 256 times, which can quantify The noise is pushed to 50k, but the SDS modulation frequency is 1024 times, which can push the quantization noise to 100k, with more high-frequency information and richer details.

b) PCM/DSD digital signal processing

For PCM, we all upscale to 1024 times the rate, and for DSD, we do the same. Generally speaking, we can't edit DSD, but we reversely analyze the principle of DSD, and specially designed a modulator structure that allows DSD as an input to directly modulate DSD 1024 times. If we get the DSD128 information stream from hdmi or usb, SDS can be directly modulated from DSD128 to DSD1024 times without going through the PCM jump process at all, and there is no information loss in this process.

2. Digital Signal Processing

1.

0. Complete isolation technology

All input devices of SDS/BDS are equipped with ground isolation circuit, which can completely isolate the interference of the front-end ground wire to the machine. We use an expensive high-bandwidth digital transformer for each spdif input signal source. While completely isolating the audio bandwidth within 20k, it can also completely isolate the radio frequency signal interference above 100M. SDS/BDS can focus on optimizing the noise performance of the digital audio frequency band. . The ground wire of most HDMI ports on the market is directly connected to the front end, causing some users to complain that the HDMI sound is dry and lacking a sense of music, so SDS has made a special isolation circuit to shield the high and low frequency noise of the HDMI port.

1.

0. Ultra-low phase noise signal receiving circuit

We are well aware of the source of additive phase noise in signal transmission and how to avoid it. We have made a circuit that can reduce additive phase noise.

1.

0. Extremely tolerant digital circuit signal

SDS/BDS has a wide range of requirements for digital input signals, which can be seamlessly adapted from 100mv to 5v and sine square waves. We are very aware that the coaxial output capability of the old machine has been aging, and it cannot be well accepted by other machines. We are also very aware that some front-ends advertise high input voltage, so we have specially designed a set of digital that can be perfectly compatible with such a wide voltage range. Adaptation circuit. We take into account that some rising edge degradation and rounding that the signal may bring when passing over a long signal line lead to the inability to trigger, and we are also perfectly compatible with this situation.

In addition, our clock input port also uses the same circuit to ensure that some non-TTL standard clocks can also be used perfectly, especially in the RF field, high-precision and low-noise clock sources often have very low output voltage. As far as we know, some of the top excellent clock sources are often sine waves, and we can also support them.

3. Ultra-low phase noise clock architecture

0. Broadband Voltage Controlled Oscillator

For the first time, we have added an inductor-based voltage-controlled oscillation unit, which allows us to use a phase-locked loop circuit with higher bandwidth and faster transients in architecture. The faster the response speed of the phase-locked loop, the stronger the suppression of the front-end high-frequency phase noise; in addition, the high bandwidth can effectively introduce more front-end sound characteristics, and the excellent front-end with SDS/BDS can achieve better sound quality.

0. Narrowband voltage controlled crystal oscillator

In terms of indicators, the voltage-controlled oscillation unit based on quartz crystal is the strongest, but we are not completely stuck on the paper values. We are equipped with a low-noise, high-transient, and narrow-bandwidth phase-locked loop circuit. This circuit configuration can filter out most of the phase noise of the dirty front-end, so that only the front-end style is retained, and the overall sound is more three-dimensional. In the design of the voltage-controlled oscillation unit, we have also spent a lot of thoughts and selected the best architecture and parts so far to make the best voltage-controlled oscillation unit so far. Generally speaking, the control voltage of the voltage control unit is only 0-3.3v, and we extend it to +-15v, allowing our phase-locked loop analog circuit to have a higher voltage control range and resolution, thereby controlling the clock frequency with higher precision, and The higher slew range also reduces the impact of noise, which can kill two birds with one stone.

0.DSD/PCM independent FIFO data pool

We have separately configured two independent FIFO data pools for PCM data and DSD. This technology is particularly beneficial when storing large-scale data, PCM and DSD data will not be overwritten, and will not cause unpleasant sound breaks. In addition, the large-capacity RAM is also conducive to the time sequence reconstruction of the sound, allowing the data to be perfectly reproduced.

0. External clock source synchronization

We are very aware of the importance of clock source performance to sound. We not only provide a local clock source for reference, we also provide an external clock synchronization function, which is convenient for users to link the timing of multiple external devices to achieve lower errors and higher accuracy. Clock requirements. External clock synchronization has two advantages: the phase noise performance of the external clock is often better than the clock performance provided by the front-end audio source, so the use of an external clock can greatly improve the performance of the decoding array; in addition, the external clock synchronization can make the timing between multiple devices fixed, and the overall The noise spectrum tends to be constant, and the boost to the sound will be greater.

4. Digital filter

1.

0. No ringing before and after

Usually DAC will have digital filters to reduce the quantization noise brought by the sound source, but this type of digital filters will bring greater problems in the analog domain. This type of digital filter can help the DAC work at an ideal frequency such as 256x or 1024x of SDS, but digital filters in the market tend to have ringing effects on the edges of the waveform. Originally a transient note, after passing through a digital filter, it will turn a tone into a tonic and multiple minor notes, so we decided to present the most complete and original information of the sound source to you, and we designed a unique digital processing In this way, the original single tone is presented to you intact.

5. Decoding Array

a) 36 programmable switch resistor array

We designed 36 independently controllable switch resistor arrays for DA arrays. The SDM modulator we designed uses pure 5bit or 32 switches, usually 32 switches can complete the task, but we added four redundant switches to smooth the mismatch noise, this design can fight for more DAC 10% of the amount of information.

b) The analog phase error is more accurate

For audio playback, phase is the most important factor we care about. We have configured a phase calibration function for all four channels. Generally speaking, the DAC usually cares about the phase accuracy of the decoding array, but often forgets to consider the phase error of the analog part, which makes the sound field of the DAC become small and the details are blurred. We especially added a phase calibration function for the analog filter, and reduced the phase error of the left and right positive and negative channels to 2 nanoseconds, which is 2 billionths of a second.

6. Fully discrete class AB simulation without feedback

a) Fully discrete Class AB analog circuit

The analog output part is independently researched and developed by us for many years, and its architecture is a fully discrete complementary input class AB, high-bandwidth design. We think Class A amplifier has the advantage of his timbre, but we prefer Class AB because the latter has the fastest transient response speed and can provide excellent current output ability, which is excellent for coping with the capacitive reactance and inductance of the wire. select.

b) Completely no feedback architecture

In theory, negative feedback can bring better distortion performance, but global feedback cannot well control the back end due to the impedance and capacitive reactance caused by frequency, which is more likely to cause unpleasant distortion. We especially changed the negative feedback to a non-feedback architecture, so that the output part of the DAC decoding array and the separation circuit can be well isolated, which can effectively make the decoding array work in an ideal state, and allow the filter network and the back-end wire to perform Electrical isolation, in this way, no matter what kind of wire or later stage, there is no way to affect the decoding array and filter network, so that the circuit can run more smoothly.

In addition, the most important point is that the no-feedback architecture can bring faster and more transparent transient performance, which is especially suitable for the expression of micro-dynamics.

7. Ultra-low noise regulated power supply

A special power supply architecture is used to obtain a full-frequency -150db power supply noise floor, which makes more details stand out

To

8. Subjective tuning direction

Technology is for sound quality. The group owner was also burned in the early ten years before designing the machine, and they have handled a lot of good and bad. After BDS, the group owner also got a lot of tuning experience. In addition, I have also referred to some representative models at home and abroad, such as:

Meitner MA1

MSB platinum III

MSB analogue + PSU

Weiss Minerva

Weiss dac202

Soulution 560

Antelope amari

Merging Anubis

Denafrips Terminator plus

Dcs delius+purcell

Dcs Puccini

Although the number is not too much, I will slowly buy more models at my own expense in the later stage and come back to play, and learn from their tuning ideas.

To

My own listening view is biased towards Germany and Austria. I hope that the transparent sound field frame will be the first, and the instrument details can be fully filled in. Although everyone has different understandings of German and Austrian music, I believe that equipment that can play back German and Austrian music should meet this requirement. Among them, soulution has the best grasp of the overall framework and level. Secondly, the dcs and msb models have been integrated into the tuning of SDS.

To

In the positioning of musical instruments, SDS also strives to be the most accurate. Just like weiss and Dengfeng, the front, back, left, and right, up and down levels are very distinct, and the level and intensity of the piano tone are often different for many times, which puts very demanding requirements on the noise floor of the DAC. As we have obtained from internal tests, weiss and Dengfeng are very outstanding in terms of noise performance.

To

Compared with some older machines, the sound is often slower. That is because the simulation technology they use is slower, which makes people feel that the transient is insufficient and the sound is lifeless. SDS deliberately optimized the speed of the sound to make the whole sound faster and the scene to explode more vigorously.

9. Machine size

330x330x65 mm

Self-developed crystal oscillator module and PLL.

Resistance module

6 capacitors