This review should have appeared more than a few months ago. When I reviewed Linn's Troika cartridge back in the Fall of 1987, in Vol.10 No.6, Audiophile Systems also supplied me with a sample of the Linn LK1 preamplifier and the LK2 power amplifier, which I had intended to review in the due course of things. As it transpired, however, I was less than impressed with the LK2, finding, as did Alvin Gold back in Vol.9 No.2, that while it had a somewhat laid-back balance, it also suffered a pervasive "gray" coloration, which dried out recorded ambience and obscured fine detail.
I suspect that Linn themselves also had had reservations over the LK2's sound; a revised version, renamed the LK280, was introduced a year ago at the 1988 SCES. When I learned that the '280 was to be launched, I naturally put the review on hold until I could get my hands on a sample of the new amplifier. As things transpired, this was not until December last year and I then got deeply involved in a number of other projects, most notably my quest to find good, affordable loudspeakers, which developed into a full-scale education into what makes speakers tick. It wasn't until late April of this year, therefore, that I was able to devote serious time and energy to this review.
This apology over, Linn Products is probably better known for their turntables, tonearms, and cartridges than for their loudspeakers and electronic components. They take all their products very seriously, however, and have put together a fine collection of some of the more inventive engineers in the UK. (It doesn't hurt Linn being based in Glasgow, Scotland, amid the UK branches of some of the more high-tech companies around, such as Hewlett-Packard.) Their preamplifier and power amplifier were launched in late 1985 and have become, I understand, steady sellers.
The LK1
Linn's LK1 is an unusual-looking preamplifier, its over-square, squat black chassis devoid of conventional knobs and switches. Instead, a recessed square section on the left-hand side of the molded front panel contains a keypad of rubber pushbuttons to select source, balance left/right/ volume up/down, mute, Record, and Memory.
Above the buttons, a green LED indicates power-on, with other green LEDs indicating the source selected, red LEDs indicating a source selected to be sent to one of two tape recorders, and a final red LED that flashes when the LK1 has received and understood a command from the infra-red remote control. (This duplicates the front-panel controls, though it also carries a number of additional buttons, presumably to cater to future, unspecified Linn products.) A microprocessor and associated circuitry carried on a vertical printed circuit board behind the front panel control all the input switching and offer considerable sophistication. A different source from that being listened to can be selected for recording; combinations of pushbuttons, using the first as a Shift key, can switch the balance control to full left or right and center it again; the "Mem" key can be used to store and recall current volume and balance settings for any source; and the controls, once set, can also be locked out to avoid them being inadvertently disturbed (at a party, for example). About the only operational idiosyncrasy—and that trivial—is that you cannot record from the MM input when listening to MC and vice versa. An on-board battery preserves memorized settings for several months.
The front-panel and remote controls set and reset DC voltages that control solid-state CMOS switches, these located on the main double-sided pcb that occupies most of the interior space. The advantages of such nonmechanical switching are twofold: first, there is nothing to wear out; second, the switches can be placed at the optimum point in the signal path, avoiding both the use of unnecessarily long or convoluted pcb tracks and of hardwiring looms that are time-consuming to put together in manufacture. Aluminum chassis apart, the LK1, for example, consists of just four assemblies: the switching pcb, connected by ribbon cable to the main audio pcb, which also carries the power supply and voltage-regulator circuitry as well as all the input and output sockets apart from the two pairs of gold-plated phono sockets for the MM and MC inputs, a small rear-panel pcb carrying the disc-input jacks, and a power transformer shielded in a steel can. The result is an electronic component that can be manufactured efficiently and easily tested automatically.
Following the circuit from the MC input sockets, the signal is carried by short twisted-pair connections to the board. Unusually, the input is AC-coupled via what appears to be a tantalum-electrolytic capacitor (though it could be a solid-aluminum type). A cascoded differential amp, using bipolar transistors, provides the first stage of amplification, with a servo circuit based on half of an LM358 dual op-amp chip (this a low-power type with high DC gain) rejecting DC. The signal from the MM input jacks is also carried via a twisted-pair connection to the main board, where it, too, is AC-coupled, this time via what appears to be a small-value polycarbonate cap. The input shunt capacitance is set by a polystyrene cap and the first half of the MM circuit seems similar to the MC, with lower gain, of course. Both MM and MC amplifier stages feed a DG309 analog switching chip, with then a further common stage providing, I believe, the final, low-frequency section of the RIAA equalization. Wire links in this section need to be snipped if a MC cartridge with a nominal output greater than 500µV/cm/s is to be used, and the equalized and amplified output signal is AC-coupled to the input switching chips, these another pair of DC-controlled DG309s.
Four pairs of line-level signals—Tuner, Aux, and two Tape recorders—enter via Linn-manufactured, pcb-mounted XLR sockets and are also taken to the input switching chips. The selected source signal then passes to the active volume control, this offering 256 discrete steps—enough that the action of the control sounds to be continuous—implemented by more DG309 ICs and an Analog Devices AD7528 DAC, here used as a switched resistor ladder. (Feed a digital word into a DAC's input data terminals, drive the chip's voltage reference pin with the analog signal, and the output should be a signal current proportional to the value of that word.)
The final output stage feeds two sets of outputs in parallel. As noted, with the exception of the disc inputs, all the in/out socketry consists of XLRs, two 5-pin male types carrying stereo in/out signals for the two tape loops, and 3-pin males provided for the two additional line-level inputs and the two pairs of outputs. These are thus nonstandard in two ways: in usual professional practice, a male XLR socket denotes an output rather than an input; and Linn wires its sockets to carry a pair of unbalanced signals rather than a single balanced signal, with pin 1 the ground, pin 2 the left channel, and pin 3 the right.
This may seem a pain in the you know what, but good XLRs offer at least three advantages over RCA connectors: they make the ground connection before the hot, thus avoiding the possibility of blowing up loudspeakers and amplifiers; they lock, providing a mechanically secure connection; and the contacts are both gas-tight and self-wiping, ensuring a consistent contact. It is not surprising that they have become the connector of choice for all pro-audio work where people's livelihoods depend on their connectors. If you are worried about soldering your own plugs and cables to use with the LK1, isn't that what dealers are for? In any case, Audiophile Systems can supply adaptor cables to enable the LK1 to be used with source components and power amplifiers fitted with RCA sockets.
The LK1 is exquisitely made, though my sample did feature a couple of electronic afterthoughts in the form of extra resistors soldered on the top of the main board and a little piggyback board on top of one of the chips. The case gets quite warm, apparently due more to the power transformer than to hot-running circuitry, so plenty of ventilation should be arranged. My only quibbles concern the number of line-level inputs (too few), the volume control—there is no visible feedback of selected volume to aid resetting to an exact level (though the memory does help here, if you remember to use it)—and the buttons on the remote are undifferentiated, making it hard to use the thing intuitively. It also took me quite a while to realize that locking out the controls using the remote only affects the remote, leaving the front-panel switches operative, although locking the controls with the front-panel switches does also shut out the remote.
The LK280
The LK280 is a more powerful development of Linn's original LK2 power amplifier. Of similar height and depth to the LK1, it is wider. The only control is a power on/off switch on the right of the front panel. Construction is totally dual-mono, apart from the centrally sited toroidal power transformer, and even that has dual secondaries, one for each channel. Complete left and right channels of the amplifier—diode bridge, twin reservoir capacitors, voltage regulators, output transistors, heatsink, and input circuitry (everything except the output sockets)—are each carried on an individual vertically mounted pcb running the full depth of the chassis. The rear panel carries a 3-pin male XLR socket at each end, these the same as featured in the LK1 and mounted directly on the pcb. One has pin 2 connected to define that channel as "left," the other pin 3 connected for "right." Two input leads are therefore required. The outputs are taken from 4mm sockets, these taking only banana plugs—they are spaced too close to take dual bananas—and are very tight-fitting with conventional plugs. A heavy internal wire connects the negative terminal to the junction of the two reservoir caps, while the positive terminal is connected to the two flat-pack output transistors with another heavy piece of wire. The series-pass voltage-regulator transistors appear identical to the output transistors and are in intimate thermal contact with them.
As with the LK1, the '280's modular construction means that assembly can be quick and testing automated. According to their literature, Linn's Quality Control procedure involves testing every LK280 board with a Hewlett-Packard computer that measures the value of every component and checks it against specified tolerances. The computer then powers up the board and runs a full function test. The boards are soak-tested at high power for two days (!), then retested. Additional tests notfeatured on the production line include thermally cycling the amplifier from –20°C to 80°C (–4°F to 176°F) while operating; storing the '280 at temperatures of 100°C to 180°C (212°F to 356°F!) for up to two weeks (at the higher temperature, the amplifiers apparently look like they have been in a fire but still work); overloading the amplifiers into all manner of impedances, from a short circuit to a purely reactive load; and, toughest of all, being used on Linn's loudspeaker production line!
Unusually, the voltage supplies to the output stages are fully regulated. (The only other domestic power amplifiers currently available in the US that I am aware of that use full output-stage regulation such as this are the Mark Levinson No.20.5 and the top Naim and Exposure designs.) Full protection is provided by a custom thick-film hybrid circuit that monitors the current in the voltage regulators, shutting them down if it appears that the amp is about to be overloaded. To reset the protection, the LK280 must be turned off and left for about five minutes.
The Sound of the LK1
The LK1 and LK280 were used on and off for everyday listening over a period of six months, with formal auditioning of both pre- and power amplifiers divided into four distinct sessions. First, the LK1 was used as a phono preamplifier, then as a line preamplifier; third, the LK280 was substituted for my regular amplification; finally, the combination was assessed as a whole.
Comparing the LK1 with the Vendetta Research SCP-2, it was apparent that the latter was in a different class from the Scottish preamp, both in the delicacy with which it presented the music and the overall soundstage and in the manner in which it presented detail within that soundstage. This is rather an unfair comparison as the Vendetta now costs some $2250 and is dedicated to MC-cartridge replay.
More relevant is the comparison with the PS 4.6/M500 combination, which competes in the same approximate price region as the Linn. The initial auditioning involved comparing the PS with its active line section in-circuit. There was no question that the Linn LK1 bested the Californian preamplifier in this condition. Listening to the Arturo Delmoni solo violin recording that Kavi Alexander produced for Apogee, for example, revealed that the PS sound was fizzier than the LK1, with less body and more of a rosiny edge to the sound of the violin. However, I have never been that impressed with the sound of the PS 4.6's line section, and continued the auditioning with the PS set to its "Straightwire" setting, where the output is taken from the wiper of its volume control, bypassing the active line circuitry.
Now the sound from LP was slightly better through the PS. Both preamps presented an excellent sense of space, but the PS had just that more "palpable presence" (to borrow a cliché from Sam Tellig, the Audio Anarchist). Whether vocalist or instrumentalist, they sounded just that bit more "human" through the 4.6 when compared with the LK1. Listening to the excellent collection of songs by Noel Coward and Flanders & Swann from the King's Singers (EMI EMC 3196), the LK1 could be heard to dilute the individual identity of the different voices a little, as though a slight layer of thin gauze was being interposed between image and listener.
Moving on to the 1979 Proprius recording of Bach's "Wachet Auf" cantata (released on Meridian E77016)—a beautifully natural sound—confirmed the impression that voices had a rather lighter tonal balance through the LK1 than the PS in "Straightwire" mode, but that the sound was less open overall. Similarly, the individual tonalities of violin and oboe on this recording were rendered more alike through the Linn than through the PS Audio.
The low-frequency performance of the Linn preamp, however, was excellent. Despite lacking a little weight in absolute terms, the bass was beautifully defined. I slapped the MFSL re-release of Emmylou Harris's Quarter Moon in a Ten Cent Town(MFSL 1-015) on to the LP12, specifically the Jesse Winchester song "Defying Gravity." It's probably because this album was half-speed mastered, but the lows have always had a flabby quality to them. Via the LK1, the bass guitar and bass drum were, relatively speaking, as tight as a nut, with considerably less overhang to the sound when compared with the PS 4.6. Dynamics, too, were excellent, there being a good feeling of "slam" to the sound of rock recordings in general.
To sum up my impressions of the sound of the LK1's phono section, the complete preamp from MC input to line-stage output has a less HF-prominent nature than the complete PS Audio 4.6, and gives a better sense of space. Its bass is a little lightweight, but is very well defined, and rock program reproduces with a good sense of pace and dynamics. Using the PS preamp in its optimum configuration, however, reverses this ranking order, and reveals that the LK1, while extremely detailed, is a little thin in the presence region, accentuating the throat character of female voice. Recorded tape hiss, too, came over as a little "whiter." The highs, however, were not as dry as the Hafler, which also presented less of a sense of space than the Linn.
Bypass testing on a preamp's line stage is a particularly revealing, even cruel, way of assessing overall quality. The LK1, however, did quite well on this test. (Its inverting nature meant that I had to reverse the speaker connections when it was in the loop, a most awkward procedure.) Yes, it did have an identifiable character, but it modified the nature of the sound to a relatively small extent (though it did have more of an editorial nature than the tubed Conrad-Johnson PV9 that I reviewed in the May issue). Specifically, the LK1's line section appeared both to reduce the size of the performing space a little and to bring forward individual images within that image. The Michael Tilson Thomas CD of Gershwin piano works, for example, (CBS MK 39699), though mainly featuring the Rhapsody in Blue and Second Rhapsody, also includes a number of solo piano works. The sleeve is ambiguous about where these were recorded, either the Dorothy Chandler Pavilion in LA or RCA's Studio A in New York. If the latter, then artificial reverberation must have been used as the piano seems to be in a large hall, particularly via the Line Drive. With the LK1's line section in circuit, the hall noticeably becomes smaller, particularly on the first Prelude, which has a number of percussive climaxes where the ambient die-away is clearly audible.
This was just as apparent on my own recording of Anna Maria Stanczyk playing Chopin on the HFN/RR Test CD. The CD, made from a digital tape, even under the best circumstances always sounds a little drier, a little closer, than the analog tape I ran at the same time. The LK1 takes the sound slightly further in the same direction, as well as reducing the image depth.
The piano recordings also hinted of a slight degree of added warmth in the lower midrange, which was confirmed with tracks from Linda Ronstadt's What's Newalbum, recorded with the late, great Nelson Riddle (Asylum 9 60260-2). I love the songs on this album so much that I can almost forgive Ms. Ronstadt for the fact that no one ever taught her how to breathe, how to effortlessly float a melody. But I can't forgive engineer George Massenburg for the "swallowing the microphone" perspective he chose for her voice, which adds a phlegmy edge to the sound. This edge is accentuated via the LK1, but only to a moderate degree, and the lower mids are definitely warmed up a little. There was also a slight loss of top-octave "air" with the LK1 in-circuit.
All in all, this is pretty good line-section performance for a reasonably priced, solid-state preamplifier.
Finally, a point that came up in my review of the Hafler Iris last month was the possibility that alien remote controls might inadvertently trigger false responses. I fired every infra-red remote I could lay my hands on at the LK1 and pressed every button. Nothing. No reaction. Like the Hafler, the LK1 is probably burst-proof in this respect, therefore. One thing did occasionally intrude, which was that changing the volume occasionally produced a faint background of soft clicks, especially at low volume levels, as the resistor ladder changed its overall value. This, of course, ceased when the volume had been set to the desired level. (For those concerned about such things, it appears to take about 12 seconds for the volume control to go from full down to full up.)
LK280 Against the Adcom
Comparing the LK280 with the Adcom GFA-555 proved interesting in that the two are almost typical examples of their pedigrees. The Linn is archetypically British—oops, almost said English—in that it is as physically small as its designers have managed to make it, and is relatively low-powered, in terms of output voltage swing if not of current delivery. Finished in plain black, with rounded edges, it completely lacks features, its sophistication all internal. It uses bipolar output transistors with a fully regulated power supply. To manufacture the LK280 and its cousin preamplifier, Linn Products established a new automated factory on a green-field site in Scotland.
The Adcom is as American as apple pie in that it is large, is intended for 19" rack-mounting, has sharp edges to its enclosure and heatsink fins, is very powerful, and, in my opinion, is ostentatiously styled, considering that it too is finished in black. While not as full-featured as some designs, it does include distortion-monitoring LEDs, and the ability to be transformed into a bridged-mono amplifier. The Adcom features a muscular but unregulated power supply with some 60,000µF of capacity. Perhaps sadly, it is also typically American in that, although designed in the US—I understand that Nelson Pass had a hand in its creation—I believe the GFA-555 is actually manufactured in the Far East (footnote 1).
The sounds of these two amplifiers are extremely different, perhaps also typifying the difference between the two schools of design. The Adcom packs a tremendous wallop in the bass, with noticeably more extension than the Linn. Bass guitar and bass drum had considerable impact, whereas they were more polite via the Scottish amplifier. The highs, too, were different, the Adcom either being clearer or too bright, depending on your taste, the LK280 softer, less lifted up in the top octave. For my tastes, however, the American amplifier comes over as being a little too brash, too forceful, with less detail resolved despite the impression of clarity.
Take Thomas Dolby's Aliens Ate My Buick (EMI Manhattan CDP-7 48075-2). The first track, "The Key To Her Ferrari," is an involved production, with just about everything, even the unmistakable voice of Robin Leach, thrown into the mix. Via the Adcom, you are struck by the sheer forcefulness of it all, the power of the bass and the clarity of the highs. But through the LK280, though it doesn't sound as "loud" (even with levels matched exactly), there is actually more detail apparent, presented more subtly. Buried in among everything in the intro to the crazed, quasi-technoswing arrangement are what sound like bongo drums. They are quite audible on the GFA-555 but are not well-differentiated when it comes to determining their position. The Linn presents the bongos with just a little more space around them.
In fact, it was this ability to better differentiate small details within complicated mixes that ultimately decided my preference for the LK280. Elsewhere in the arrangement, Dolby talks us through the subject of his obsession—"And then I saw her...she was a bright red '64 GTO with fins and gills, like...some obscene phallic symbol on wheels..."—his voice accompanied by pitch-shifted clones. Again, the Linn amp better differentiated the distinction between the original voice and those clones.
But again I must point out that this preference will be an individual affair. My wife Laura agreed with me on the nature of the differences between the two power amplifiers but preferred the Adcom for its clearer high frequencies, its "less muddled" presentation, and its greater bass weight. Me, I had to disagree. The stunning performance from the Atlanta Symphony Orchestra and Chorus of the not-widely-enough-known Requiem by the French composer Duruflé (Telarc CD-80135, coupled with the Fauré Requiem) just had more of a three-dimensional quality via the Linn, even though the Adcom underpinned the choir and orchestra with the organ pedals that much more effectively. In addition, the GFA-555's more forward upper treble seemed to accentuate what can best be described as a "fffy" quality that hangs over the strings on this recording.
It was the Linn by a nose, in my opinion.
I also briefly compared the LK280 with the original LK2. No contest. The newer amplifier, while still laid-back, perhaps a little polite, is fundamentally musical, while the LK2 is, well, typically solid-state, that "gray" coloration overlaying everything to the detriment of the musical flow. You should note, by the way, that the LK280 runs extremely hot. If you intend to hide it in a cupboard, be sure it has sufficient ventilation.
The Sound: Together
Then I hooked up the pair as a system.
Odd. From all my auditioning of the individual components and my notes on the individual aspects of sound quality, I had expected that, in harness, the sound from CD would be best served by the combination. Yet with CD, it seemed that the rather restricted image depth offered by the LK1's line stage failed to be compensated for by the LK280's good performance in this area. With LP as source, however, there was depth aplenty. The sound was "typically British" in that it lacked extension at the band edges, with restricted low bass and rather dull highs, but the music as a whole flowed, even bopped, with a powerful presentation of upper-bass and tenor instruments, cello and bass guitar, for example. I got the impression that the complete Linn LP-playing system and the Linn amps formed a synergistic combination.
No, it still didn't approach the sheer refinement coupled with musical verisimilitude offered by the (much more expensive) Vendetta/Mod Squad/Krell combination that has evolved into being my reference. But there was a rightness to the sound of the Linn combination driving the Celestion SL700s that led night after night to one LP following another. Whether it was Prince's Purple Rain (Warner Bros. 25110-1), Randy Newman's curiously evocative score to The Natural (Warner Bros. 25116-1), or the Rostropovich performance of the solo Britten cello suites (Decca/London SXL 6393)—to name three records that followed in quick succession the night I finished tapping out this review on the keys of the trusty Toshiba laptop—the overall sound was musically "right."
Conclusion
Exquisitely made but unusually styled, Linn's LK1 and LK280 can be used as stand-alone pre- and power amplifiers. I imagine, however, that they will almost always be used as a pair, due to the idiosyncratic XLR connectors, the '280's low input impedance, and, last but not least, the fact that they do appear to form a musically synergistic partnership. As Len Feldman said in the April 1989 Audio, perhaps the LK1 and LK280 should be regarded as a two-chassis integrated amplifier. But at $2545 total, this is one expensive integrated amplifier, and it is fair to point out that the hairshirt audiophile can achieve better performance in at least one of the separate areas of sound reproduction by carefully choosing separate, "universal" preamps and power amps. (Use Stereophile's biannual "Recommended Components" listings as a starting point.)
Achieving greatness in one area of reproduction at the expense of others is not necessarily the best way of achieving musical satisfaction, however, and the LK1/LK280 scores in its ability to present a satisfyingly whole sound. The combination is not so much intended for audioholics who need to mix'n'match components on a monthly basis, but for music lovers who require "Fit and Forget" high-quality amplification.
Footnote 1: Read Fred Warshofsky's The Chip War (Scribner's, New York, 1989) for a fascinating if disturbing account of how the US consumer electronics and semiconductor industries reacted to Far Eastern competition by deciding not to compete at all. See also The Economist, May 20, 1989, p.91, for a frank discussion of the same subject.
0 nhận xét:
Đăng nhận xét