The DAC2 is an audio digital-to-analog converter. Most people focus on the word "digital" and assume that all of the "magic" happens in the digital processing, but nothing could be further from the truth! A look inside most audio converters would show that about 90% of the components are analog! This application note takes a look at the analog processing in Benchmark's DAC2 D/A converter.
If an audio conversion product (A/D or D/A) is 90% analog and 10% digital, it stands to reason that the performance is closely tied to the quality of the analog processing. For this reason, two converters with identical digital chip sets can have wildly different performance. The DAC2 has an ES9018 8-channel D/A conversion chip and this same chip can be found in some of the very best stand-alone D/A converters. The ES9018 can even be found in some premium-quality Blu-ray players. The mere presence of the ES9018 does not mean that these products will sound identical. Many people assume that the converter chip determines the audio quality, but there are vast differences between the measured performance of the DAC2 and these other DACs and Blu-ray players. It is safe to say that 90% of the measurable performance differences are directly related to the differences in the analog processing inside these products. The photo below is color coded to show the major subsystems in the DAC2 HGC.
The remaining 10% of the measureable differences can be attributed to differences in the digital processing. The digital processing in ES9018 is very versatile, and the processing can even be offloaded to external devices. The DAC2 performs much of its digital processing outside of the ES9018. These digital processing differences will be discussed in part 2, while the analog processing difference will be discussed here. Part 3 will discuss power supplies.
D/A converters must include most of the following analog subsystems:
The performance of these analog subsystems are determined by the following:
The analog outputs on the ES9018 D/A converter IC can be operated in "voltage mode" or "current mode". The current mode provides the best performance (lowest distortion), but it requires many more analog components. The Benchmark DAC2 series converters use the ES9018 in current mode, which means that the DAC2 has external current to voltage converters (I-V converters) at the outputs of the ES9018. The performance of these I-V converters is critical to the overall performance of the DAC2. Benchmark spent a great deal of time designing this critical stage. The components, impedances, and gain of these I-V converters were carefully considered and computer modeled. This task is complicated by the fact that D/A converter chips (such as the ES9018) have a substantial amount of radio-frequency (RF) energy on the output pins. The I-V converters must maintain stability in the presence of this RF energy, and this RF must be filtered out before the analog signal leaves the box.
Analog low-pass filters are required to remove the RF and ultrasonic energy at the output of the D/A converter chip. Some of this energy is a by-product of digital sampling. High-frequency images of the audio information are repeated at integer multiples of the oversampling frequency. These images must be removed by an analog low-pass filter. This filtering process changes stairstepped digital signals back into the continuous analog waveforms that were captured by the A/D converter in the recording studio. The DAC2 uses the ES9018 with a high oversampling ratio and this provides an increased separation between the musical signal and the first ultrasonic image. This wide separation makes it easier to remove the unwanted ultrasonic and RF interference without impacting the audio signal.
The ES9018 has balanced outputs. It would be tempting to connect these directly to the balanced XLR outputs on the back of the box. Many products use this cost-saving shortcut, but a careful examination of these outputs will show that this is a bad idea from a performance standpoint. The balanced outputs of the ES9018 have a very strong common mode distortion component. This common-mode distortion signal can be completely removed by using precision-trimmed differential amplifiers. We use 0.1% high-precision metal film resistors to create an outstanding set of differential amplifiers. These well trimmed differential amplifiers are very effective at removing the common mode distortion produced by the ES9018.
Please note that common-mode distortion is not unique to the ES9018. All D/A converter chips tend to have high common-mode distortion at their outputs. The balanced outputs provide a mechanism for removing distortion that occurs inside the chips. Differential amplifiers can reject the common-mode distortion while passing the audio. Unfortunately most audio products omit the differential amplifiers as a cost-saving measure. Precision differential amplifiers are one of the distinct advantages provided by the DAC2 family of converters.
The outputs of D/A conversion chips are usually limited to a voltage swing of only a few volts. Benchmark operates the ES9018 in a current output mode which eliminates the need for a voltage swing at the output of the D/A chip. Our I-V converters maintain a constant voltage on the D/A outputs while sensing variations in output current. The I-V converters immediately convert this current to a professional studio-level voltage signal. The internal signal path of the DAC2 uses levels that are between 18 dBu to 24 dBu at 0 dBFS. These high-level signals are well above the noise floor of the analog signal path and this allows the DAC2 to deliver the full specified SNR of the ES9018 at the balanced XLR outputs. The DAC2 has lower noise than most products that use the ES9018. Again, the difference is due to the analog processing.
The RCA and XLR outputs on the back of the DAC2 are designed to drive long lines and a wide range of impedances. The XLR outputs feature low-impedance passive attenuators (pads) that can be used to reduce the studio-level outputs of the DAC2 to the lower levels required by many power amplifiers and powered speakers. Note that the Benchmark AHB2 power amplifier is designed to accept full studio-level balanced inputs, and the pads are not required when driving an AHB2. Most other amplifiers will require the use of the internal 10 dB or 20 dB pads. The pads optimize the gain staging between the DAC and the amplifier and place the volume control in a reasonable operating range.
Some outboard D/A converters include volume controls and analog inputs. The Benchmark DAC2 converters feature a generous mixture of analog and digital inputs. All of these inputs can be controlled by the volume knob. The DAC2 is designed to connect directly to virtually any power amplifier. No other device is needed between the DAC2 and the power amplifier.
The DAC2 features our unique Hybrid-Gain-Control™ system which separates analog and digital inputs into two entirely different volume control systems. We use 32-bit digital gain control for digital inputs, and an analog potentiometer for analog inputs. We do not convert analog inputs to digital, nor do we pass the output of the D/A converter through an analog volume control. Nevertheless, both types of inputs are controlled by a single motor-driven potentiometer. Analog input signals pass directly through this pot while digital input signals follow a shorter analog signal path that bypasses the pot. For digital inputs, the position of the pot determines the multiplier for the 32-bit digital gain control that feeds the 32-bit input of the ES9018. This hybrid system provides two signal paths; one that is optimized for analog inputs, and one that is optimized for digital inputs. This hybrid system delivers the lowest possible noise and distortion for each signal type.
Remote control is achieved through the use of a motor drive. The analog and digital gains track very precisely as the knob rotates. This makes it easy to switch between analog and digital sources without readjusting the volume.
We don't believe in just tacking on a headphone jack. Benchmark has a long history of building stand-alone headphone amplifiers. Our HPA2™ is actually a small audio power amplifier with a near 0-Ohm output impedance. It is capable of driving difficult loads and it can deliver large voltage swings. Internal jumpers can be configured to optimize the output level to your favorite headphones. The HPA2™ is designed to provide the same measured performance as the XLR and RCA outputs and we find that many of our professional customers use the headphone outputs for critical listening.
The HPA2™ easily drives two pair of headphones and this is one reason that we provide two headphone jacks. The other reason is that one jack mutes the XLR and RCA analog outputs while the other does not. Headphones only? Use the left-hand jack. Headphones and speakers? Use the right-hand jack.
All D/A converters require a reference voltage. The analog outputs of the ES9018 are derived from a 3.3 volt reference supplied to 8 separate input pins (four for the left channel plus four for the right channel). The cheap and simple solution is to tie all of these pins together and connect them to the same 3.3 volt supply that powers the rest of the digital system. This simple solution is cheap and widely used, but it fails to fully leverage the potential of the ES9018.
The DAC2 has a very low noise reference voltage system that separates the left and right reference inputs. The system starts with a precision voltage reference followed by a multistage analog lowpass filter. The output of the filter drives two separate active buffers. These buffers separately drive the left and right voltage reference pins. Each reference pin also has a dedicated filter capacitor. This elaborate system extracts the best possible performance from the ES9018.
No discussion of audio converters would be complete without mentioning clock generation, clock recovery and jitter attenuation. An SPDIF coaxial digital input or a Toslink digital input can be buffered and sent directly to the input pins on the ES9018. This low cost solution is popular with many manufacturers, but it is not the method used by Benchmark. The Benchmark UltraLock2™ jitter attenuation system is entirely external to the ES9018. All incoming clocks are fully isolated from the low-jitter fixed-frequency crystal oscillator that drives the D/A conversion. This oscillator is located less than one inch away from the ES9018, and the clock connection is well shielded and isolated from the surrounding circuitry through the use of a 6-layer printed circuit board.
All digital inputs on the DAC2 are asynchronously upsampled to 211 kHz to remove all traces of interface jitter. Benchmark's UltraLock2™ upsampling system provides uniform jitter-free performance with all types of digital interfaces (coaxial, XLR, Optical, and USB). Cable length, cable type, and interface format will not cause an increase in jitter at the D/A chip in the DAC2. External jitter attenuation devices will have no positive or negative effect on the clock performance of the DAC2.
The ES9018 is a very flexible D/A converter chip. It offers simple solutions that require very few external components, but it also offers opportunities to extract much higher performance. Consequently there are wide variations in the measured performance of products that use the ES9018 converter chip. Never assume that the sound of a converter is determined by the D/A chip!
"Inside the DAC2 - Part 2 - Digital Processing" discusses the unique digital processing found in the DAC2.
"Inside the DAC2 - Part 3 - Power Supplies" will have more information on the power supplies in the DAC2.
Application Note subscribers will receive these application notes when they become available.
Myth - "Damping Factor Isn't Much of a Factor"
Myth - "A Damping Factor of 10 is High Enough"
Myth - "All Amplifiers Have a High-Enough Damping Factor"
These myths seem to trace back to a well-know paper written by Dick Pierce. His analysis shows that a damping factor of 10 is virtually indistinguishable from a damping factor of 10,000 when it comes to damping the motion of a loudspeaker cone. This analysis has been examined and repeated in many more recent articles, such as a well-written post on Audiofrog.com by Andy Wehmeyer. Articles such as these are often cited as evidence that amplifier damping factor doesn't matter. The mathematical analyses are correct, but the conclusions are incomplete and misleading!
How fast things can change!
It is March 23, 2020 and we are currently battling the worldwide COVID-19 pandemic.
This application note will be a departure from normal. I will make a few observations about the current situation and then look at the nuts and bolts of how we reconstructed our operations in less than 48 hours. Benchmark is 100% operational, but nothing looks the same as it did last week.
- John Siau
As an engineer I like to use "rules of thumb" to make quick estimates that help to explain the physical world around me.
These rules of thumb are easy-to-remember approximations that eliminate the need for complicated and needlessly precise calculations.
If you feel discombobulated by the complexities of high school physics, there is hope! I encourage you to step back and take a fresh approach.
If you learn a few simple rules of thumb, you can unravel mysteries of the physical world, amaze your friends, and yourself.
In this paper I will present 15 simple rules that I find useful when working with music and audio.
- John Siau