Audio-GD - R7 2020 - Discrete Balanced R2R Ladder FPGA Accusilicon DAC - PCM / DxD / DSD
Balanced discrete R-2R digital analogue converter equipped with advanced high speed ladder correction including optimized FPGA FIFO buffering and re-clocking. All parts and components are selected based on excellent sonic and measurements performance. Components used are e.g. WIMA caps; DALE and Vishay resistors, KOA transistors and NOVER capacitors. The design (e.g. low noise power supplies; discrete analogue output stages) were being selected after many years of experience and practices and are considered to perform and provide finest sound quality. Up to date technology supporting high resolution including native DSD, DxD and hyper-modern inputs like Amanero USB and LVDS I2S inputs.
Unique R2R correction technology. Digital signal processing driven by high-speed FPGA supporting FIFO buffering and re-clocking. Software up-gradable design enables future new capabilities and improvements.
New in R-7 2020 version
Supports a 10 MHz master clock clock input
FPGA can now process data in parallel mode (Oversampling Mode), improves clock stability. NOS mode works with serial data processing.
IIS input (Include USB and HDMI-IIS) are dual 24bit parallel data, and DSD dual 64bit parallel data.
Parallel processing improves sound transparency and detail.
Menu Controlled System
All functions operable from front panel for user friendly experience. In total 7 different sound characteristics: 4 different algorithms in NOS mode and 3 types oversampling modes. (Jumper-less version!)
Firmware ugprade port
Micro USB port on back panel. (Update without opening the unit)
3 Different firwmwares available
We offer R2R7HE to R7HE upgrade! Audio-GD - R2R7 (HE) to R7 (HE) Upgrade
REVIEW BY 6MOONS: "In conclusion, we were very impressed with this DAC. For the price, it offers very good sound which is unusually adjustable to individual hardware conditions and taste."
Software up-gradable design. wo high performance low jitter Femtosecond Accusilicon clocks, operating at 98.xxx and 90.xxx MHz level to support todays and future (higher) sample rates. A true balanced analogue implementation
- 4x R-2R DA-7 V2 DAC modules
- True balanced design
- Digital DSP for optimal signal processing
- Discrete push-pull R2R DAC design
- Internal OCC wires & low-jitter TCXO clocks applied
- Discrete analogue output stages
- DC coupled design (No capacitors in signal path)
- Non- feedback ACSS analogue output control
- Power supply consist of 21 regulated low noise regulators
- Over 50,000uF power capacity to purify the power supply
- High-grade matched transistors per channel
- USB Amanero Combo384 interface with galvanic isolation
- High Res support: DSD64 - DSD512 & DxD
R7 is characterized by its coherent, transparent and natural sound. Striking is the spatial sound image. Rich in detail and strong authority. You'll forget that you're listening to a digital source.
USB Singxer F1 Option (Replacing Amanero)
Enjoy DSD512 and best sound quality with Singxer F1 USB module installed (Replacing Amanero)
MAGNA HIFI is offering a significant upgrade of the internal USB interface by replacing the Amanero USB interface for a more sophisticated Singxer F-1 XU208 board using the Crystek CCHD-575 ultra low phase noise clocks. Besides the sound quality improvements it enables native DSD512 support! We'll make sure the Singxer F-1 is properly installed and using internal ultra clean power source from the internal DAC to ensure best sound quality
- 1 - AES/EBU ((3-PIN XLR, 110-ohm)
- 2 - S/PDIF (RCA, 75 Ohm)
- 3 - 10Mhz Master Clock Input (50Ohm)
- 4 - S/PDIF Optical (Toslink)
- 5 - I2S (HDMI, LVDS)
- 6 - USB (Amanero, USB class 2, 32bit)
- Micro USB - firmware update
- ACSS balanced
- XLR balanced
- RCA single ended
R-2R LADDER DESIGN
The R-2R DAC has become popular and affordable. From DIY kits and completely up to of the shelf high-end products.
In the lower end DIY market, the R-2R design is often based on old technology; designed long time ago by MSB, often includes basic R2R ladder design only and not include the wonderful correction design of the modern MSB technology. Previous less good designs used data shift registers logic chips in serial mode to convert the data to an analog signal. The stubborn R2R issues cannot be avoided this way and performance is almost entirely depending on the accuracy of the ladder resistors.
In the High-End of the shelf (finished products) market, the R2R design is usually much more complex when outstanding performance is offered. A basic R2R ladder is simply not sufficient to achieve the desired sound quality. Some manufacturers are using shift registers design to realize ladder compensation. A less complex, unfortunately also a less performing design based on traditional logic chips working in serial mode.
A far better design switches resistors in parallel mode; an ultra-fast FPGA chip controls and corrects the R2R ladder. The parallel design mode controls every bit respectively and therefore achieve unprecedented performance. (In parallel mode only 1 clock cycle is needed to output all data; serial design mode needs at minimum 8 up to 24 clock cycles) The parallel design is particularly complex, however when it is done properly it can correct every bit of the ladder. (Photo below shows a FPGA design with R2R ladder). With incredible speed and accuracy the R-28 will correct the unavoidable imperfections of the basic R2R ladder caused by tolerance of resistors; avoid glitches at ultra-high speed to deliver unrivaled musical performance.
TOLERANCE OF RESISTORS IN LADDER
Some manufactures claim to achieve good performance becuase they use lowest tolerance ladder resistors. The following calculation example will cover this. At 16 bit the tolerance of 1/66536, 0.1% (1/1000) it is by far not enough to achieve decent performance. Even using resistors with 0.01% (1/10000) tolerance, lowest tolerance available today,cannot deliver the desired performance This is just 16 bit, imagine a 24 calculation! To process 24 bit resolution it would require resistors with a tolerance of 0.00001%. This is purely theoretical; if this tolerance would exist the discreteness of the switch logic chips will already have too much internal impedance and will shake-up such design.The tolerance of the resistor will never solve Imperfections of a ladder.
The solution lies in the correction the ladder and not only depend on the tolerance of resistors. It’s a combination of both: Ultra-low tolerance resistors controlled by a correction technology using very high speed with parallel FPGA processing.
PROS AND CONS OF DISCRETE R-2R LADDER DAC
- R-2R will not convert the clock signal into the output signal.
- R-2R is less sensitive to jitter
- The output signal is more accurate with R-2R
- Modern Delta-Sigma chips offer exceptional THD; R-2R offers satisfactory THD results.
- Accuracy and glitches of the ladder resistors require complex technology to enhance
ULTRA FAST SIGNAL PROCESSING
R2R7 FPGA responsibilities:FPGA stands for Programmable Array Logic. Nowadays the FPGA is widely adopted in audio devices. The internal hardware design is fully controlled by complex software. A huge advantage is the fact the software in the FPGA can easily be upgraded offering new capabilities or improve the performance of the device without replacing any hardware. Versatile and future proof design.
- High performance SPDIF interface, replacing traditional less good performing SPDIF interface chips like DIR9001, WM8805 or AK411X,etc.
- Full re-clocking process with FIFO design applicable on all inputs. This way the output data keeps fully synchronized with the clock signal to reject any jitter.
- Built in 2X, 4X and 8X oversampling and digital filters and on top of this 4 different true NOS (only analog 6dB filtering) modes. To completely configure the sound according to your taste.
ANALOGUE OUTPUT STAGES
The analog output stages are as important, they have a tremendous influence on the final sound quality. After d/a conversion by the R2R D/A modules the analogue signal is transported by fully discrete matched-transistor output stages; DC-coupled design with first class through-hole components. No SMD components are applied in the analogue section.
The high speed unique ACSS ((Audio-gd Current Signal System) output stages are non-feedback and current driven. Unique performance because almost all other designs need to convert the signal multiple times from and to current and/or voltage, resulting in less detailed and less transparent sound-stage The output buffers are single ended FET; two stages in parallel to reach very low output impedance. All output stages are in pure class A without (negative) feedback to achieve purest and a real live sound reproduction. The 4 OPA op-amp's are DC servos, this way no coupling-capacitors are needed and DC output is automatically biased. The listener will be presented a strikingly transparent and neutral sound.
The DAC consist of 3 tuned 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 DC power is distributed over 19 separate power regulators. All are pure class A low noise shunts; fed by 3 groups linear power supplies. Resulting in ultra-high speed and clean power for all individual parts.