Instruction Manual

The Technical Stuff
__________________________________________________________
- 33 -
errors caused by transients, overshoot, intermodulation distortion, etc. are big and must be corrected for by
huge amounts of negative feedback. However, the intermediate circuit elements within the op-amp don’t
get corrected immediately because these devices don’t have the required infinite frequency response
assumed by the classical circuit models. To make matters worse, high loop gain requires more circuit
elements to provide the added gain needed for the “corrective” negative feedback, which exacerbates the
problem. It’s like throwing gasoline on a fire to put it out.
In addition, standard IC op-amps use class A/B biasing, which amplify the positive and negative half-
cycles using different parts of the circuit. But because every gain stage in the 2192 is pure Class-A biased,
the amplifiers are always drawing current from the power supplies, and they amplify both the positive and
negative half-cycles of the signal. The Class-A “constant-current” mode means the circuit doesn’t suffer
from “supply-droop” distortion when handling low frequencies and sharp transients, and there is no cross-
over distortion inherent with class-AB IC op-amp designs. Our Class-A, all discrete op-amp design
approach uses low-gain op-amps with matched, high precision components, and minimal component
stages. We use only enough negative feedback as we need to provide stability, low output impedance, and
good linearity. The goal was to have the analog signal paths (from the line-ins to the A/D converter, and
from the D/A converter to the line-outs) be as good as they can be. The analog circuitry is fully DC-coupled,
meaning that there are no capacitors in the signal path to introduce phase distortion, which can smear
high frequencies and take the punch out of low frequencies.
The 2192 analog circuitry is also fully dual-differential, meaning that there are independent, identical
circuits processing both the + and - sides of the differential signals. This reduces distortion, and improves
common-mode rejection and dynamic range. It also improves imaging by eliminating cross-talk between
channels. Another major factor that affects dynamic range is noise immunity. Any analog circuit that
requires a low noise floor must be designed as a differential circuit so common mode noise is eliminated.
We designed the 2192 to use not just fully-differential analog circuitry, but dual-differential circuitry. This
means we use two fully-differential circuits for each channel of signal for even greater noise immunity.
We also use high quality field-effect transistors (FETs) with matching characteristics. FETs are a special
type of transistor that share the good-sounding characteristics of both tubes (i.e., low-order, “musical”
2nd and 3rd harmonics) and standard bipolar-junction transistors or BJTs (clarity, transparency, fast
transient response). If FET circuits are designed correctly, they don’t share any of the negative aspects of
tube or BJT circuits (noise, harshness, slow transient response). However, if used incorrectly, FETs can be
affected by a type of parasitic circuit capacitance (called the Miller Capacitance) which reduces transient
and high-frequency response. We use a special biasing technique in our FET op-amps to eliminate this
problem.
Another important design decision was to avoid the use of any soft clip or limiting circuitry in the 2192.
Many converters that incorporate such “safety-net” approaches use variations of essentially fuzz-box
circuits to round the signal so it’s already clipping when the converter clips. Unfortunately, the distortion
these circuits introduce isn’t very musical. We designed the 2192 to be both as accurate as possible, and
as musical as possible. These are conflicting goals at high signal levels because transients can easily
overload even a “perfect” converter. This is simply because digital signals don’t have infinite headroom.
This isn’t as much of a problem in analog circuits because analog clipping introduces harmonics that are
usually musically related to the signal. Digital is different. Because digital signals are sampled in time,
digital clipping is modulated by the sampling rate, and this introduces non-musical tones called aliasing.
The problem is caused by the abrupt transition between accurate signal representation and a full-scale
square wave. We call the gain region between nominal and full-scale levels the “transition” region between
clean and clipping, and the way the signal approaches the impassible digital full scale level is critical for
reducing aliasing and improving signal clarity.
Our approach to soften digital clipping is to use what’s always worked better: the transparent compression
and harmonic “bloom” that occurs naturally in low-feedback Class-A circuits. We carefully selected the