C.E.C. TL 1 CD transport

I find it astonishing that two products built on completely opposing engineering principles can both have musical merit. Design goals exalted by one company are considered anathema by another, yet both components produce superb sonic results.

201cecx1.jpgThe two products I'm referring to are the $8500 Mark Levinson No.31 Reference CD transport I reviewed last month and the new C.E.C. TL 1 belt-drive transport that's the subject of this review.

No two transports could be more different. Where one design team went to great lengths to minimize a certain parameter, the other company went through equal effort to maximize that parameter. The two transports are based on completely different design philosophies. And, though they certainly don't sound alike, both are worthy high-end products. The respective designers managed to optimize their designs after starting with very different sets of fundamental beliefs.

Technical description
Before telling you how the TL 1—the world's first belt-drive CD transport—works, I'll give a little background on C.E.C. It's a large Japanese manufacturer that makes OEM products and components for brandname companies. Although the C.E.C. name is new to consumers, its products aren't. C.E.C. has a long record of making turntable parts, particularly motors and bearings. For example, they manufacture: the bearings in the SME tonearms; the motor, printed circuit board, and power supply in the SME turntable; and the Oracle turntable's main drive bearing. For most of its 20-year existence, C.E.C. made low-end turntables for the Japanese audio giants. As turntable sales were replaced by CD-player sales, C.E.C. got into the CD transport market. They are now one of the largest suppliers of transport mechanisms in the world, and make complete CD players for some companies. Although the bulk of C.E.C.'s design and production goes to the mass market, there are a few dedicated tweaks in the company who created the TL 1 belt-drive CD transport.

Nakamichi was in the same position in the early 1970s—its cassette-deck mechanisms were used in many brand-name components, yet no one had heard of the company. That changed when Nakamichi introduced the revolutionary Model 1000 cassette deck under its own name. I suspect that the TL 1 is only the first of many high-end products we'll be seeing from C.E.C. Just as the Nakamichi 1000 was an ambitious flagship product intended to make a name for the company, so, too, is the TL 1 transport.

The TL 1 is a monumental from-the-ground-up design effort. C.E.C. makes virtually everything inside the TL 1, from the chassis down to the spindle shaft and bearings. The belt-driven transport mechanism was built exclusively for the TL 1. Other than the various "big-company" drives, the only other "from scratch" transport mechanisms of which I'm aware are those made by Linn (in the Karik), Teac/Esoteric (Esoteric P-2 and Wadia WT-2000), and Pioneer (their stable-platter mechanism).

The TL 1 has a massive internal build concealed beneath gorgeous metalwork. Its 37-lb weight, fortress-like build, and stunning appearance combine to suggest that the TL 1 is a serious product. The unit's rounded corners, sloping front and sides, and top-panel sliding window give it a futuristic yet elegant appearance. In fact, the TL 1 is one of the best-looking audio products I've had in my listening room.

The top panel holds a sliding glass door that opens to reveal the disc mechanism. A large and very heavy clamp is placed on the disc to hold it on the spindle. Closing the door triggers a microswitch that spins the disc and reads the disc's Table of Contents (ToC). This switch is also a safety interlock: The laser won't turn on until the door closes (footnote 1).

Front-panel controls are minimal. Only Play, Stop, Pause, Track Skip, and a Power On/Off button are included. A greenish-blue fluorescent display shows track number and elapsed time. A button on the remote control switches between elapsed track time and total remaining time on the disc. The remote also includes programming functions, a numerical keypad for direct entry of track numbers, and audible search.

Four digital outputs are provided on the rear panel: coaxial on an RCA jack, coaxial on a BNC jack, ST-type optical, and TosLink. A coaxial digital interconnect, terminated with RCA plugs, is included with the TL 1. The interconnect has an inductive ferrite capsule around one end, similar to TDK's Digital Noise Absorbers and AudioQuest's RF Stoppers. Interestingly, this inductor is made by Taiyo Yuden, the Japanese company heavily involved in recordable compact disc technology (footnote 2).

I couldn't get all the way inside the TL 1 to see the mechanism and electronics: It's built like a tank. Removing the thick outer metalwork revealed a tightly built inner enclosure that houses the circuits and transport. Consequently, I could get only a peek at the innards. The sturdy inner chassis is made from extruded aluminum between 2mm and 8mm (5/16") thick. The chassis floor is 5mm (nearly ¼") thick. Damping material covers some of the inner chassis parts. The TL 1's solid construction is impressive.

The TL 1's power supply consists of a large transformer with four secondary windings, two of them center-tapped. These feed three full-wave rectifiers and one half-wave rectifier. Six regulation stages are used, all of them based on standard three-pin devices.

A printed circuit board in the rear right-hand corner contains the servo and decoding electronics. The main chip is a surface-mounted Yamaha YM7121B. This device decodes the EFM signal output from the laser photodetector, generates servo signals, and performs CIRC error correction. The laser head itself is made by Sanyo.

During the Audio Engineering Society Convention in San Francisco last October, JA and I visited Parasound (importer of the TL 1) and saw the unique belt-drive transport mechanism that had been removed from the chassis. The entire mechanical system is suspended on a subchassis by four large springs and rubber cushions. Two large motors flank the mechanism—one drives the spindle, the other moves the laser pickup assembly.

As I considered the TL 1's belt-driven operation, I wondered if there were engineering justifications for belt-drive (as there are in an LP turntable), or if a belt-drive CD transport is merely a marketing ploy. Is belt-drive an advantage in a CD transport, as it is in a turntable? I had some skepticism based on the vastly different requirements of turning an LP and spinning a CD. Let's take a brief look at these very different rotational drive systems.

We all know that belt-drive LP turntables are generally better than direct-drive models. The belt isolates the motor's vibration from the platter. Belt-drive is a natural choice in turntables because the platter spins at a constant 331/3rpm—there is no need to control the platter rotational speed with a servo. The motor just sits there, spinning at one speed. If a direct-drive motor is used, not only are the vibrations from this motor directly coupled to the record and stylus, but so is the servo-control frequency, which is generally somewhere in the low treble.

CD transports are very different, however. The motor's speed varies greatly, according to what portion of the disc is being read. At the inside radius, the disc spins at about 500rpm. At the outside, it turns at about 200rpm. This changing speed results in a Constant Linear Velocity (CLV) as seen by the playback laser. When playing a disc from start to finish, the motor gradually slows down (a CD is read from inside to outside). But when asked to skip a track, the motor and servo must respond by rapidly increasing or decreasing disc speed.

Moreover, a CD transport's rotational speed is usually under tight servo control. The data stream read from the disc produces a clock signal (the 4.3218MHz bit rate resulting from EFM encoding) that tightly regulates the rotational speed. Accurate speed is essential for correct data recovery.

In short, conventional thinking suggests that a CD transport's electronics should have a close grip on the mechanism's rotational speed. This goal would seem to be at odds with putting a belt between the motor and spindle, because the belt would apparently bog down the servo's control. It would be like driving through an obstacle course in a car with lots of play in the steering wheel. Just as the steering-wheel play would produce a lag time between turning the steering wheel and the car's reaction, the belt would produce a delay between the servo correction signal and the transport's response to that signal.

Moreover, the TL 1's 1-lb disc clamp would also appear to slow the servo response. As discussed in my review of the Mark Levinson No.31 Reference CD transport last month, a lightweight clamp has many advantages in regard to rotational servo performance. Low rotating mass can be controlled much faster than high rotating mass. These contrasting design philosophies were dramatically illustrated every time I made a comparison between the TL 1 and the No.31. Picking up the No.31's ultra-lightweight clamp after handling the 1-lb TL 1 clamp was quite a shock (and vice versa). Further, the TL 1 was slower in reading the disc's TOC and accessing tracks.

C.E.C., however, makes a compelling case for belt-drive and high rotating mass in a CD transport. A belt-driven CD transport has the same advantage enjoyed by belt-drive turntables—isolating the disc from motor vibration. In addition, the very heavy clamp acts as a flywheel, smoothing out variations in the disc's rotational speed. With such a large rotating mass, the disc would be less prone to speed fluctuations and thus wouldn't need tight servo control. The TL 1's designers gave up the ability to quickly respond to changes in rotational speed in exchange for inherent speed stability. Instead of making many fine and tight speed adjustments, the C.E.C. mechanism makes fewer, slower corrections to keep the disc spinning at the right speed. Consequently, the mechanism can use a very low-torque motor. This approach is opposite to that used in the No.31 and many other direct-drive transports.

The TL 1's brochure includes a comparison of the rotational servo-correction waveforms of three types of transports: 1) direct-drive with no flywheel and a medium-torque motor; 2) direct-drive with a large flywheel and a high-torque motor; and 3) belt-drive with a large flywheel and low-torque motor (the TL 1). The TL 1 had the highest-amplitude servo-correction signal (as would be expected for controlling the high-mass flywheel), but was also the smoothest in shape. The two other waveforms were lower in amplitude but less like a sinewave. I was surprised C.E.C. touted this as desirable: Transport designers usually optimize the mechanical design for lowest servo-correction signal amplitude.

C.E.C. makes other claims about the superiority of belt-drive and high rotating mass. First, the high-torque motors used in direct-drive designs reportedly radiate high levels of electromagnetic noise; C.E.C.'s low-torque motor is said to be quieter. C.E.C. also claims that the flywheel's high inertia reduces disc vibration and smooths out the drive motor's cogging effect.

Not only the spindle is belt-driven; so is the laser pickup. A second motor flanking the transport mechanism drives the laser-head assembly. This is claimed to reduce radiated electromagnetic noise near the disc. The laser sled is connected to a toothed gear driven by the laser-head motor.

These aspects of the TL 1's design—belt-driven spindle and laser head, high rotating mass, and low-torque motor—are unconventional, to say the least. In short, the TL 1 is like no other CD transport. But how does this contrary design sound?

My first listen to the TL 1 was with it driving the No.30 processor via ST-type optical interface. Because I was familiar with the No.31 transport, the TL 1's presentation took me by surprise. These were two very different-sounding transports, but the TL 1 appeared to have some very special qualities of its own. Extended auditioning confirmed these first impressions. Indeed, I came to like the TL 1 the more I listened to it.

First, the TL 1 had an extraordinarily smooth and analog-like presentation. To call the TL 1 laid-back is an understatement. The music was gentle, sweet, and presented with a more distant perspective than I'm used to hearing. There was a feeling of ease, relaxation, and warmth uncharacteristic of digital. Even compared to the very refined No.31, the TL 1 had a softer, less incisive presentation. The TL 1's overall character can be summed up as unforced and easygoing.

Footnote 1: A word of caution: If the CD is placed in the transport and the door closed without the disc clamp being in place, the motor still starts and hurls the CD around the transport's interior.—John Atkinson

Footnote 2: Taiyo Yuden joined with Sony and Sonic Solutions to develop a CD recorder that could be connected to Sonic Solutions' Macintosh-based digital-audio editing system. The recorded CD would replace ¾" U-Matic tape for CD mastering.—Robert Harley

Chuo Denki Co., Ltd.
US distributor: Parasound Products
950 Battery St.
San Francisco, CA 94111
(415) 397-7100