Universal Audio WebZine
Volume 4, Number 8, October 2006
[UA Universe] [Ask the Doctors] [Artist Interview] [Analog Obsession]
[Support Report] [The Channel] [Plug-In Power] [Playback] [Featured Promotion]
[Graphic-Rich WebZine]
[Back Issues] [UA Home]


Ask the Doctors: A Few Questions about the DCS Remote Preamp
By Scott Silfvast and Rob Silfvast

This month, we ask Scott and Rob Silfvast a few questions about the new DCS Remote Preamp.

What is transimpedance?
Fundamentally, a "transimpedance" circuit is one that converts current to voltage. The term applies to the DCS Remote Preamp design because a transimpedance gain stage lies at the heart of the current-feedback instrumentation amplifier employed in our design.

What made you decide transimpedance is the most neutral way to go?
This was a natural choice because a transimpedance-or current-feedback-amplifier has a superior ability to respond identically, and very accurately, to all frequencies throughout the audio spectrum. Its internal architecture is a fundamentally different from that of a conventional, voltage-feedback amplifier, so that it provides much more "raw amplification power" at higher frequencies. To explain this, we need to dive a bit deeper...

 “Transformerless input coupling combined with the transimpedance gain stage provides a clean transparent signal path without any coloration in a manner that doesn’t blow the project studio budget.”

A current-feedback amplifier employs a low-impedance feedback terminal, driven by an internal voltage source, to keep itself within its linear operating range. The current provided by this voltage source is multiplied through a high-gain transimpedance stage to produce an output voltage, and this action forces the feedback current to zero. Figure 1 shows a basic schematic of a current-feedback amp. In the presence of dynamic input signals, this low-impedance source can charge and discharge capacitance in the surrounding (feedback) circuitry virtually instantaneously, allowing it to track signals very accurately no matter how fast they change. In contrast, a conventional (voltage-feedback) amplifier has a high-impedance feedback terminal, so it cannot drive capacitance as quickly when tracking dynamic signals. This can cause its output to lag behind the ideal response. Even when the external capacitance is kept to the lowest possible amount, the difference can be dramatic.

Figure 1: Basic current-feedback amplifier architecture. Note the feedback terminal (In-) driven by the emitters of Q1 and Q2. Signal gain is provided by the transimpedance stage inside the blue outline

With its inherent ability to react more quickly, a transimpedance stage is able to overcome two major problems inherent in conventional amplifiers: gain-bandwidth trade-off and slew rate limiting.

Gain-bandwidth trade-off is a phenomenon in which an amplifier has less and less available open-loop gain as the frequency of the input signal increases. Even in high-performance conventional (voltage-feedback) amplifiers, it is typical for this roll-off to begin as low as 10 Hz. The amplifier's overall closed-loop transfer function, which is determined by applying negative feedback around the amp, is restricted by the open-loop frequency response, and becomes degraded as the closed-loop gain approaches the open-loop gain. Most affected are phase and transient response, because the amplifier essentially loses its power to immediately react to fast changes in the input signal, in much the same way as a car with a weak engine cannot accelerate quickly at 60 mph. By eliminating this gain-bandwidth trade-off, the transimpedance circuit has plenty of "raw amplification power" across the entire audio spectrum and beyond, so it can impart exactly the same transfer characteristic to all frequencies arriving at the input. The result is excellent transient response and zero signal coloration. Figure 2 illustrates this important difference between conventional and transimpedance amplifiers.

Figure 2: The benefits of transimpedance architecture: more raw amplification power at high frequencies

From the graphs, we see that the benefits of a transimpedance architecture, operating in the audio spectrum, are most significant when high amounts of (closed-loop) gain are demanded from the amplifier. That's why it makes so much sense for a microphone preamp. For amplifier stages requiring unity gain or small amounts of gain, a high-quality voltage-feedback amp is often a more appropriate choice due to its lower cost and power consumption.

Slew rate limiting occurs when an amplifier cannot change its output fast enough to respond accurately to a sudden change on its input. While related rather closely to the gain-bandwidth trade-off phenomenon described above (which affects signals at all amplitudes), slew rate limiting can impede an amplifier's performance in the presence of large signal transients. The absence of this limitation is another benefit of the transimpedance architecture.

Are there any transformers in the signal path?
No. The DCS Remote Preamp's microphone and DI signal paths are transformerless. The goal was to provide a third sonic option alongside UA's world-class Precision Class A and Classic Tube preamp offerings, which both use very high-quality transformers to couple input signals to the amplifier stage. The sonic signature of a premium-grade transformer is highly desirable for many recording applications and it comes with a premium price tag-partly due to the high cost of the transformer itself. Alternatively, the DCS Remote Preamp was designed to provide amplification without any sonic signature or coloration while offering two complete preamps with DI in a project-priced product. Transformerless input coupling combined with the transimpedance gain stage provides a clean transparent signal path without any coloration in a manner that doesn't blow the project studio budget. In short, the DCS Remote Preamp allows the natural sound of the mic or instrument to be amplified and recorded directly, while the option to add coloration later in the mix can be implemented in DSP with great products like the UAD-1 and its suite of plug-ins.

How did you achieve these kinds of specs at this price?
As mentioned above, the elimination-by-design of expensive transformers was one of many cost-reducing solutions we used to meet the challenge of making world-class preamps available to the desktop market.

We started by making the product extremely compact, which significantly reduces the cost of the steel enclosure and printed circuit-board materials. With small board sizes, we could afford to use four copper layers (with internal power and ground planes), which facilitates very compact circuit layouts and excellent noise immunity. These surface-mount boards are designed for quick, easy and reliable assembly using standard automated assembly machines, keeping labor time and cost to a minimum.

We have been designing digitally controlled analog circuits since the mid-1980s, and over time we have compiled a sizable bag of tricks from which to draw. For example, we have learned how to employ very cost-effective silicon-based switches in ways that totally avoid distortion and cross talk. Our precision, digitally controlled circuits are designed such that no labor-intensive factory calibration steps are required. Some of them actually calibrate themselves on power up. The only factory cal is the centering of the needle on the VU meters.

Wherever possible the design incorporates mature industry-standard components. Many standard parts are becoming increasingly useful in professional audio products as the electronics industry continues to mature with multiple vendors competing on price and quality.

We chose to use ubiquitous CAT-5 cable for the interface between the Base Unit and Remote Controller. It has twisted pairs to keep audio and logic signals isolated from each other, plus it's dirt cheap. Not only does this save cost on the bill of materials, but it allows users to substitute a cable of any length (up to 300 feet) with inexpensive, off-the-shelf cables available at all kinds of stores like Radio Shack and Home Depot.

What's good about the DCS Remote Preamp's low-cut filters in comparison to other designs?
Most notable is that our filters work in conjunction with the preamp stage to provide at least 5 dB of extra low-frequency headroom. In typical designs, the low-cut filter is placed after the preamp, so that unwanted low frequencies are fully amplified prior to being filtered out. This is not ideal because in order to avoid LF clipping, the overall signal gain must be reduced, and the result is a smaller signal with less dynamic range at the preamp's output. In the DCS preamp, the filtering occurs prior to full amplification, and this blocks LF energy closer to the source, allowing the desired signal to make use of the full dynamic range of the preamp.

Another advantage is that we provide three different cut-off frequencies in addition to the off setting. At each of the three settings, the roll-off characteristic is 18 dB per octave, so these filters are surgically accurate and very effective. The 100 Hz setting is perfect for bringing out clarity in vocals tenor and above, drum overheads or any instrument that "lives" in the midrange or above. The 70 Hz setting is great for passing guitar, cello or low male vocals, while blocking underlying "mud." The 30 Hz setting accommodates most bass instruments and kick drums while effectively blocking rumble. Finally, if you want to capture all the boom and rumble of a room, or perhaps a deep bass drum, you can turn the filters off and pass all frequencies above 5 Hz.

What else is special or unique about the circuit design as compared to other designs?
There is a unique front-end circuit for the DI (line/instrument) input. It is a balanced, high-voltage, discrete transistor stage that provides very high-input impedance (4 M_), while allowing unusually large signal swings of ±18 volts on either the hot or cold legs along with excellent low-noise performance. This circuit operates on the internal 48V phantom-power supply, and includes a servo that maintains constant biasing conditions over the entire signal swing, virtually eliminating distortion.

The 48V phantom-power switching uses a 2-second ramp period rather than suddenly switching on or off. This can save speakers and eardrums from the destructive effects of huge transients that occur when someone inevitably switches the phantom power with the preamp unmuted. Furthermore, when the ramp generator reaches 48 volts, it functions as a second stage of regulation to help ensure very clean and quiet phantom power.

There are several more unique circuit designs and product features we could get into that relate more to utility and functionality than sound quality . . . but maybe that's a whole other interview for a future WebZine!

Do you have a question for the Doctors?