he different between dual ES9038 Pro and single ES9038 Pro is large:
The different not only the DNR, but the dual ES9038 Pro working under mono mode, the channel crossing is improve and offer a larger and exact sound stage.
Dual ESS9028 Pro, working under mono mode, its DNR had 3DB better than single ES9038 Pro .
The FPGA is important in the D-27:
The FPGA is Programmable Array Logic.
Nowadays the FPGA had applied in a lot high end grade DACs, like the popular ROCKNA WAVEDREAM DAC.
Its internal hardware construction can through the complex software design and arrangement , be upgraded through the software upgrade.
Upgrade to the firmware is the same as same as upgrade to the hardware. That design is much more flexible . It can improve on sound quality and add the newest functions , it can never become out of date.
What the FPGA duty in the D-27:
1, Separate the data feed to dual ES9038 Pro for working in mono mode.
2, Clock re-clock process and FIFO design for all inputs , the output data can exact synchronisation with clock signal, reject jitter.
3, Built in high performance SPDIF interface , bypassing low performance SPDIF interface of the ES9038 Pro .
Full discrete output stages:
The signal last stage is the analog output stags , they greatly effect the whole DAC sound quality.
The output stages working after ES9038 pro, they are full discrete transistor built with hole parts but not SMD parts .
The high speed ACSS amps work as the signal gain and processing , they are of non feedback design and working in current signal mode . the dual ES9038 pro output [current signal] goes through the ACSS amp .I don't like other design with multiple conversion to the signal between current and voltage.
The output buffers are the pure class A design , built with 4 pcs 150W (in total 600W ) on every channels, to drive the most headphones without difficulty .
In our design output stages working in pure class A , and without any negative feedback for the most pure and life like sound reproduction .
There are 4 pcs OPA built in as DC serve , so the DAC can working well without couple caps and output the most neutral sound.
Heavy power supplies design:
The DAC has 3 high quality low noise, low flux leakage, R-cores transformers.
In total 130W power to supply all digital parts and the left and right analog boards.
The digital parts DC power is distributed by 15 ultra high speed low noise PSUs, they are double stages PSUs .
The analog parts DC power is distributed by 4 groups PSUs ..All are pure class A low noise regulated power.
This results in ultra-high speed and ultra-low noise performance. Clean and independent power to achieve highest quality of sound .
The volume control is also different from conventional technology. Conventional volume control technology implements volume control the analog input stage through a volume pot to reduce the signal . The volume pot quality effects the sound quality directly , the low quality volume pot cause loss of detail and couse channels imbalance and sound stage distortion . Even the high performance volume pot still cause loss of detail . Whatever quality volume pot still can cause the S/N degradation .
But in the D-27 , we applied the I/V conversion volume control , the volume control is a variable passive I/V conversion placed at the ACSS amp output , where the output is the current signal mode not in the voltage signal mode . The volume control is I/V conversion, and the volume depends on the current (I) conversion to the voltage (V). (Like R-2R D/A chips output passive I/V conversion) It can keep the signal frequency band flat and not lose any detail. It want degrade sound quality regardless of volume level . After the volume , there are four groups diamond non-feedback buffer output stages offer with very low output impedance .
The volume control quality is. even more important in the real balanced gear . It must guarantee the four channel signal control to be exact and achieve the performance of the real balanced gear . If the cool and hot of the balance signal can't be exact, the balance output will has large distortion and cause the sound quality and performance to be even worst than a single end gear . ( whole idea behind balanced design is wasted)
In D-27 is applied a four channels digital control relays based volume control. .D-27 has total four channels of the Vishay resistors to control volume and avoid channel imbalances . This achieves the best performance and sound quality.
D-27 has 100 steps volume control with range that achieves 90DB (usually the step volume have around 70DB range ), very smooth control range on first 50 steps , along with 13DB high gain can be very easily adjusted for both high / low sensitivity headphone and speakers. While users change the volume, the different relays break or close, it may has a slight switch sound on output but disappears when you stop adjusting volume.
Supper Exponential volume characteristic Volume steps comparison to normal volume knob
The overall arrangement :
The DAC had separate the digital, Left analog, Right analog and power transformers parts, every part is separated by 5MM thick aluminium boards to avoid distortion.
The Left and Right analog parts have been symmetrical placed beside the digital parts , to keep the wires and distance same . This allows sound stage to be more precise .
The sound quality adjustments :
Some people may don't like the sound of ESS chips. They are consider to sound artificial . We want to change this perception .
At first we want to let the unit fully burn in, then adjust the parameters layout and design to lower the distortion to absolute minimum . This is easy part..
Then we have to spend a long time on listening tests and constantly comparing sound quality . Than we adjust the parameters to change the Odd and even harmonics to let the sound be life like and palpable.
At the end the DAC may not have least amount of distortion but has the most neutral sound with life and realism.