Meitner IDAT D/A processor

There are as many ways of designing a digital-to-analog converter as there are engineers. One approach is to select parts from manufacturers' data books and build the product according to the "application notes" provided by the parts manufacturers. This is the electronic equivalent of a paint-by-numbers kit.

A more creative engineer may add a few tricks of his own to the standard brew. Bigger and better regulated power supplies, careful circuit-board layout, tweaky passive components, and attention to detail will likely make this designer's product sound better than the same basic building blocks implemented without this care. Indeed, the vast range of sonic flavors from digital processors containing very nearly the same parts attests to the designer's influence over a digital processor's sound.

There is still another way of designing a digital processor: Junk the off-the-shelf parts, forget about "standard implementations," and trash-can conventional thinking. This designer is going to completely rethink the role of each stage in a digital processor and invent entirely new circuit topologies to overcome the limitations he sees in standard components and techniques. If our first engineer, designing by application notes, produces a product equivalent to a paint-by-numbers painting, this last designer is like the master who creates a whole new movement in art.

Such is the case with the new Meitner Intelligent Digital Audio Translator (IDAT) digital processor and its designer, Ed Meitner. The $15,000 IDAT represents a revolution in digital converter design; the unit brims with creative thinking and engineering innovation. From its custom input receiver through the unique DSP-based digital filter and the unit's eight DACs in a proprietary configuration, the IDAT is like no other digital converter extant. Indeed, the IDAT is perhaps the most technically sophisticated consumer audio product ever produced.

But innovation for innovation's sake is worthless; what really matters in audio reproduction is the musical experience. Novel topologies and inspired technical design may be interesting in their own rights, but must ultimately exist to better convey musical expression; they are the means to an end, not the end itself.

The technology
The IDAT's outward appearance is as unusual as the technology inside. The unit comprises four chassis, one housing the power supply, the other three joined together to form the Translator itself. All the metalwork is machined from solid aluminum blocks and covered with a black powder-coat finish.

The Translator is divided into three sections: the lower, base portion houses the digital circuitry, and the two upper sections hold the left- and right-channel DACs and analog stages. The audio sections' front panels are dominated by two large beveled glass squares that show off the unit's four Motorola DSP chips. These glass squares form the front of the two audio modules, which rest on a base containing the input jacks and the digital circuitry.

A chrome bar beneath the left window allows the user to select between the three digital inputs (coax, Toslink, and AT&T ST-type optical) and invert absolute polarity. A row of eight LEDs indicates the selected input signal's sampling frequency, which input has been selected, absolute polarity inversion, and if the C-Lock circuit (described later) is locked to the incoming data. Unusually, these LEDs are blue rather than the familiar green, orange or red. Blue LEDs are a recent phenomenon and are still very expensive. I found them a refreshing change.

The Translator's rear panel holds both unbalanced analog outputs on RCA jacks and balanced outputs on XLR connectors. The panel's lower portion is fitted with three digital inputs and an identical set of digital outputs. Digital outputs allow connection to a digital recorder such as DAT or CD-R.

As I removed the power supply's cover, it hit me that the power-supply chassis is identical in size and construction to one of the Translator's audio modules. The outboard power supply uses two transformers, one for the analog DC supplies and one for the digital DC supplies. Incoming AC is filtered at the IEC jack. No regulation is performed in the outboard supply; all regulation occurs next to the supplied circuits in the Translator. Two captive "D"-type connectors terminating 48"-long cables supply DC to the Translator.

Once inside the Translator, the DC is regulated by banks of three-pin IC regulators. A row of seven regulators is mounted on the single large printed circuit board that consumes the Translator's lower section. These are sandwiched between the chassis and a large bar for greater heatsinking. The seven regulators are strictly for the digital power supplies; virtually every digital circuit has its own power-supply regulation stage. The digital board contains two additional regulators.

In each of the two audio modules, six regulators supply the DAC and analog output stages. I was curious about the row of six-transistor/470µF capacitor pairs on the DAC board that appeared to be related to the power supply. It turns out that these are "capacitance (C) multipliers," a simple yet ingenious circuit using emitter followers used on the power-supply rails. Here's how they work: The regulated DC appears at the base of the transistor, which has a 470µF cap between base and ground. This capacitance is effectively multiplied by the transistor gain, resulting in higher effective capacitance on the supply rail. The transistor emitter, through which a constant and fixed current is drawn, then becomes the DC supply rail for the analog output stages. The result is reportedly a supply rail with a lower and more stable output impedance. Each of the six analog supply voltages is conditioned with a C multiplier. With the outboard chassis, 21 regulation stages, and 12 C multipliers, the IDAT's power supply is impressive.

Ed Meitner's ability to avoid compromises in existing designs by inventing completely new and innovative topologies is exemplified by the IDAT's input receiver. This is the circuit that receives the incoming data stream from a CD transport, generates a new clock based on the incoming signal, and formats the data for presentation to the digital filter. The input receiver is a critical sonic component in the digital reproduction chain: it is here that jitter is introduced in the clock signal, resulting in audible degradation (footnote 1).

Rather than use an off-the-shelf input receiver, Meitner designed a new circuit that operates on a principle completely different from the standard method of recovering the clock with a Phase-Locked Loop (PLL): the receiver looks at a part of the S/PDIF data stream called the "preamble." This portion of the data stream charges a capacitor, which changes the squarewave to a ramped voltage. The peaks are then detected and a new clock is generated from the ramp peaks. This technique reportedly results in relative immunity to jitter in the incoming data stream—jitter is not propagated to the recovered clock. This technique of looking only at a small portion of the incoming data stream to recover the clock is in sharp contrast with the usual technique of locking to the signal with a PLL.

The input receiver is a 4" W by ¾" H by 2 3/8" D potted module containing two circuit boards. This large module dominates the digital electronics board that consumes the Translator's base section. Museatex intends to make an IC of the circuit so that it may be incorporated in less expensive products. Most of the digital board's components are surface-mount devices.

In addition to using a unique topology, the input receiver uses C-Lock R (for Receiver), a version of the C-Lock circuit that reportedly results in a virtually jitter-free recovered clock. Indeed, Meitner claims that with a C-Lock R digital input stage, sonic differences between transports are reduced. (When he visited Santa Fe last summer, Ed demonstrated the IDAT using the digital output of a CD Walkman.)

The most sophisticated of the IDAT's many innovations is its proprietary digital filter. Unlike any previously designed, the IDAT's filter sidesteps the limitations inherent in both off-the-shelf filter chips and existing DSP-based custom filters.



Footnote 1: See "The Jitter Game" in the January 1993 Stereophile.
COMPANY INFO
Museatex Audio Inc.
(Company no longer in existence)
ARTICLE CONTENTS

COMMENTS
christopher3393's picture

Interesting news,,, for less than 1% of the world's population. Glitter people: talk amongst yourselves.

John Atkinson's picture

Quote:
Interesting news...for less than 1% of the world's population.

Check the date of the review: March 1993. I find it notable that, when it comes to measured DAC performance, what it took heroic engineering and a megabuck price tag to achieve 20 years ago can now be obtained for <$1k. See, for example, www.stereophile.com/content/musical-fidelity-m1-dac-measurements. But sound quality? Probably not.

John Atkinson

Editor, Stereophile

 

hollowman's picture

I think it's important to realize that many of the MEASURED parameters/variables (i.e. OBJECTIVE phenomenon) used in audio measurements are less than 80 years old, and even fewer for digital audio science. I recall, several yrs after CD was consumerized (1982-3), "linearity", oversampling, and "jitter" were "discovered" -- and add to these, now we have pre-ringing, minimum phase and apodizing filters. And, of course, none of these predictably correlate with "good" sound. So what does? ... well, it may take another 20 years for clearer answers!

[ ----------------------- SIDEBAR: ------------------

Real science often involves slow, empirical processes and techniques. Acceptance of this is what I find lacking in the so-called "objective" community; IOW, the HydrogenAudio-type "objectivists" are impatient, and put all their immature eggs in the basket of academic or textbook science. But textbooks are moving targets, subject to perpetual revision, since MOST Natural phenomena are, at the fundamental level, uncharted/undiscovered gray areas. That is, until scientists focus in on bits and pieces of these areas and have them make some mathematical sense in a dynamic TOTALITY, and/or in multiple dimensions (like transducer waterfall plots). In essence, hard-core "objectivity" proponents are dimensionally or temporally trapped -- unable or unwilling to think proactively and outside the academic box (e.g., HydrogenAudio objectivists can't answer this simple question -- "why do so many educated & scientific-minded audiophiles, engineers, researchers, physicians, as well as musicians, composers and even the general public not put too much faith in blind/ABX testing, etc?" -- w/o resorting to an overly simplistic reply like  "it's due to their cognitive dissonance" or "it's a psycho-acoustic trick because ABX tests prove otherwise"). It's important to realize that blind/ABX tests are, ultimately,  subjective human choices. And this, IMO, rightly crushes much of what the objectivists propose & argue for/against.

--------------------------------------------------]

Anyway ... all that said (and back to the topic!). First, many feel that single-bit (of which Delta-Sigma DACs are a variety [Delta-Sigmas comprise the majority of DACs used today by major manufs like Wolfson, AD, TI, CS, etc] have been --until very recently--sonically inferior to older multibit DACs. Indeed, a few "high-end" D/A's and some DAPs (HiFi Man) use older DAC chips like Philips TDA1543, TDA1541 (and non-oversampling at that!). So it may be that, for mostly economic reasons, DACs took a few steps backward for a while (esp. in the early 90's Bitstream era), and only the latest generation Sigma-Deltas are coming back to speed [the reasons behind this are complex and interdependent, including: older multibits were massive power hogs (e.g. > 200mA current draw!!) and $$ to manuf; the 1-bits were/are cheaper and more pwr efficient -- yes, pwr hungriness = higher manuf. $$ as the PS components/heatsinks needed to support these chips add to the cost/complexity, esp. in portable and automotive use]. Another way to think about it: first, contemplate the $$, hot-running,  complex/"hybrid" technology of the Meitner IDAT -- it took all that $$, ANALLY tweaked/optimizing-every-stage, volume-and-mass BULK of the IDAT to achieve decent SQ in 1993: What many SOP-sized monolithic DACs do today at much-smaller physical and pwr-consumpstion scales... and prices, of course.

I think non-digital components in D/A processors have also made significant gains. One clear winner in analog evolution is the op-amp. I recently installed some of the latest Analog Devices OPAs into a portable iPod-like DAP diy project. And they clearly blew away much-more-$ OPAs of only a decade ago. Ergo ... if a late-model D/A (or other digital/computer-audio component) does sound better than antiques like the Meitner, the reason may not be entirely "digital" after all.

audiodoctornj's picture

As a new EMM Labs dealer, we displayed the first and only example of Ed's latest ideas in digital at the recent New York Audio show, the DAC 2X SE, which at $15,000.00 certaintly isn't inexpensive, howerver, Mr. Meiter again shows why he is a genius at digital design.

On Saturday we had the pleasure of being able to play this dac with HD tracks 24/192 masters via USB into the DAC 2X Se, and the sound was remarkable, it was like you had the actual musicians in the room!

To be honest I don't know if it was the DAC or the true 24/192 track or the combination but the sound was transcendental.

The DAC 2X SE is completely new and use improved versions of Ed's unique technologies, we had an indepth conversation with Meitner's head of sales who told  us that both the Meitner MA 1 Dac and the DAC 2X SE share none of the previous algorithms which Adreas Koch worked on five years ago, so hopefully consumers will get to understand that Meiter's current products are unique in their implimenation of Ed's original technologies which he has been refining for years, and Meitner is back at the forefront of digital design, yet again.

 

silvertone's picture

 

It would make a fascinating comparison to review the new DAC2X, being at the same price point from this product twenty years ago.

John, anyway you can get your hands on one of these units?

Warm regards.

Oscar

X