KLH Model Nine loudspeaker
J. Gordon Holt, June, 1975
Editor's Note: Although this product has been available for several years, it is being reviewed in considerable detail because it is a strong contender for the title of "Best Available Loudspeaker System, Regardless of Cost," and because we plan to review some of the other contenders for the same title within the next few issues. We feel that since all of these systems represent a considerable outlay of money, prospective buyers should have a thorough understanding of the merits and demerits of each system, so they will know what to expect from them in the way of performyince capabilities and operational requirements.
The following report is longer than usual, simply because the closer a reproducer comes to approaching theoretical perfection, the more it benefits from attention to small details in its installation and use.
In theory, most of the benefits accruing to the electrostatic principle are in the high-frequency range, where smoothness, extended range and good transient response are difficult to achieve in a dynamic speaker. This has prompted most designers to limit their electrostatics to the middle and upper-frequency ranges, while relegating the low end to a conventional cone-type woofer. The KLH Nine is one of the few existing designs which strives to span the entire audio range with electrostatic elements.
The complete KLH Nine system consists of two large, flat panels, closely resembling decorator-type room-divider screens both in shape and in appearance. Each panel is framed with finely finished wood, and the front surface is covered with boucle fabric (footnote 1). The rear surface is covered with a finely perforated metal screen, and a similar metal screen lies behind each grille cloth, to protect the radiating elements.
Each panel contains a complete two-way, full-range electrostatic system. A small square panel at the middle of one edge serves as a tweeter above about 2kHz, while the entire lower range is handled by eight large strips of diaphragm material that almost completely cover the remaining frame area. The two panels are mirror images of one another; one has the tweeter located at the left, the other at the right.
Each panel has its own polarizing supply (6kV), and hermetically sealed into a heavy metal case that attaches to the rear bottom of the panel and serves as a base to support the panel upright on the floor. Each power supply draws 5W at 115V AC, and 15' AC cords are supplied. The supplies are fully fused.
KLH suggests a choice of two basic setup arrangements: with the speakers side by side in a shallow V configuration, or separated by some distance, for wider stereo spread. Because of their two-sided radiation, however, they should not be located closer than about 4' from any parallel wall surface, as this will inhibit low-end radiation and may cause midrange standing-wave problems.
For the initial tests on our KLH Nine, we placed the panels together, using the special metal straps that are provided to couple the frames, and located the system near one end of a 20' by 13' by 9' room, in an area that has proven to be satisfactory with other speakers. An oscillator sweep failed to reveal any audible peaks or dips or harmonics, but we did find one spot near the low end where there were loud metallic rattles from the edge of one of the perforated rear screens. For some reason, slight twisting of the panel frame eliminated the rattles, and they have never since recurred.
High-frequency distribution was judged to be quite poor. Each small, flat tweeter panel radiates an increasingly narrow beam with a rising response at frequencies above about 2kHz. At 5kHz, the beam was judged to be about 30° wide, while at 10kHz it was barely more than about 10°. The speakers can be aimed so that neither beam reaches a listening area, and this will offset most of the highend rise, but unless some care is taken to do this, stereo imaging will be quite poor and the sound will be excessively toppish.
The system's overall response, off the tweeter axes, sounded almost perfectly smooth from well beyond audibility (14kHz in our case, and verified to beyond 20kHz with a condenser microphone) to around 40Hz. There was a slight high-end rise above about 3kHz, but for some reason, this did not add any spit or sizzle to the sound; it just seemed to impart an open, airy quality to it. There was no slow rolloff or humped response at the low end; it hit 40Hz and then dropped off very rapidly. (We should point out that, of the other systems that extend significantly below 40Hz, practically all of them have a peak at 40 or above, which acts to exaggerate lows and to mask the audibility of the extreme lows.)
With some experimentation, we were able to place the KLH Nine where its low-end radiation throughout the room extended down to a useful 35Hz, but were unable to elicit much of anything from it below that. There was no low-end fluttering; when it dropped out, it just went silent.
Note that the foregoing observations about the KLH 9's low-end performance apply when the two panels are placed side by side. Unfortunately, this arrangement tends to kill most of the stereo effect except when listening from a distance of 8' or less, and when the panels are separated, to improve the stereo spread, the system's entire low end was noticeably impaired. Instead of linear response to 40Hz and a sharp drop-off below that, the whole bass range below about 70Hz exhibited a slow rolloff. There was still usable output at 40Hz, but it was weak, and there was virtually nothing below it.
Footnote 1: Pronounced "boo-clay," according to our office dictionary.
The effect, of this, in most rooms, was an impression of slightly excessive brilliance and low-end deficiency. Placing the edge of each panel right against a wall of the room helped to bring up the bottom, but not to the degree of linearity that was obtained with the panels together. And the against-the-wall arrangement necessitated putting the speakers too far apart for adequate center fill-in. On the basis of this, we placed the panels together for the rest of our listening tests.
To date, nothing we had found seemed to suggest that this system was worth anywhere near the $1140 the manufacturer asks for it. We changed our mind, though, when we heard music through it. This was the first time, since we started dabbling in high fidelity in the early 1940s, that we have ever felt we were really listening through a loudspeaker instead of to it; the system has a degree of transparency and detail that we simply did not believe possible.
Our long exposure to the openness and lucidity of Janszen electrostatic tweeters had prepared us for the kind of upper-range transparency we heard from the KLH Nine, but to hear lower middles and deep bass reproduced with the same entirely effortless clarity was a hair-raising experience. No matter where we placed the speakers, we could not get boomy bass out of them. Bass transients, plucked basses, and the impact transients of tympani were solid and entirely free from hangover, and when organ pedals and bowed double-basses alternated, there was simply no question as to which we were hearing. We could, almost literally, "count the cycles" of a bowed double-bass.
Only on very rare occasions on some original, ungimmicked organ tapes, for instance did we notice. that some extremely deep notes were coming through less strongly than on some other systems we've heard (the Janszen Z-600, for instance), but the superior low-end detail of the KLH Nine made us more than happy to forego part of the energy from one or two musical notes on one recording out of every fifty or so.
In listening to the KLH Nine, we were more conscious than ever before of the gimmicking that goes on in most commercial recordings. The system's remarkable reproduction of depth perspective makes it clearly apparent which instruments were close-miked and which were coming through the general-pickup mikes, and when an accent mike was suddenly turned up, the over-accented instrument would protrude from the mass of sound like a luminous proboscis. On the other hand, inherently natural recordings came through with greater depth and naturalness than we had thought possible.
Careful placement ofthe KLH Nine has a great deal to do with its performance more so in fact than with most speakers, because its two-sided sound radiation makes it responsive to acoustical conditions behind it as well as in front of it. It is possible to introduce audible response colorations by locating the panels where reflections from walls behind them tend to augment or cancel certain midrange frequencies. A few degrees change of angle or a few inches change of distance with respect to the nearest wall will generally lick the problem.
When using the hardware supplied for fastening the panels together, the radiators are angled slightly outward. This directs the tweeter beams away from the listening area and, in some locations in some rooms, broadens the apparent stereo spread by bouncing the treble beams off the room walls. It does, however, also tend to reduce the system's upper-midrange radiation to a slight degree, for these frequencies, too, are somewhat directional and tend to follow the paths of the treble beams. The losses are slight, but the system is so close to being perfectly linear that they may be quite noticeable in some rooms. If a bit more "presence" is desired, place the panels in a straight line, edge to edge.
Incidentally, when the panels are used together, they should be arranged so that the tweeters are located at the outer edges of the paired panels, for maximum stereo spread. When the panels are used spread apart, tweeter location doesn't matter (although their direction still does).
Efficiency of the KLH' Nine is very low, in the vicinity of 1%, which puts it about on a par with the best of the acoustic-suspension systems. The manufacturer recommends a minimum of 30Wpc, and to this we add a caution about the kind of amplifier that should be used.
Like all good electrostatics, the KLH Nine does not need the hardness of a typical transistor amplifier to give it adequate crispness, and it reproduces every nuance of distortion fed to it. We know of only two manufacturers whose amplifiers and preamps fully meet the necessary standards: Dynaco and Marantz. There may be others, but we haven't found them yet. And if you're wondering about some of the new transistor units, find out what their measured IM distortion is below the 5W power level, If it rises to above 0.2% at any point below 5W with a 16 ohm load, forget it. If it doesn't exceed this limit, it may be suitable, but we couldn't guarantee it.
Finally, a word of caution. The crossover networks in these speakers can be damaged by attempting to drive them at high levels with continuous sinewave test tones. Response tests should be conducted at levels of 10W or less, and tones at this level should not be allowed to continue uninterrupted for longer than about a minute. If either panel suddenly starts making loud fluttering noises at low frequencies, the crossover network is being, or may already have been, damaged. The KLH Nine will safely handle the full output of a 70Wpc amplifier on music and natural sound material, and will produce full concert-hall volume, as heard from a row-H to row-M seat, with a dual 50-watter in a room of moderate size.
But be careful with continuous tones; much of the "horrible" sound people have reported from these speakers is due to a damaged crossover network. The power supplies and radiators, on the other hand, appear to be at least as rugged and dependable as conventional dynamic speakers. We have heard of, only one case of electrical breakdown, and that system had been used where it was repeatedly exposed to conditions similar to those in a steam bath.
This is the most nearly perfect loudspeaker we have ever heard, and we've heard every likely contender that's in commercial production. It does not favor some instruments over others, but seems to make everything sound almost perfectly natural.
It is not entirely without its own sonic character, though. Although there is nothing about its sound that we would describe as a coloration, it does tend to move things slightly back from the listener, "placing" them a bit farther behind the surface of the speaker than is the intent of the recording. The effect is much like that of moving back about ten rows in a concert hall, and since either spot can be a perfectly good seat from which to listen, it is fair to say that this characteristic of the Nine does not impair the realism of the sound.
This is not the kind of loudspeaker that everyone will like on first hearing. Its major points of superiority its transparency and detail are not readily appreciated until you've lived with them for a while, and it demands better associated equipment and more careful room placement than many users are willing to provide for it. Neither is it a very satisfactory wide-stereo reproducer in many listening rooms.
Because of its low efficiency and its tendency to move the sound slightly away from the listener, the KLH Nine is not recommended for the person who likes to re-create the extremely high level and intimacy of a very close-up seat in the concert hall, or who plans to be listening to it from a distance of 20' or more. But for the person with an average-to-large living room, whose preferred concert seat is between Row F and first balcony, the KLH Nine can come closer to re-creating the illusion of concert-hall realism than any other commercially available loudspeaker system, regardless of cost.
JGH summarized the KLH Nine in October 1968 (Vol.2 No.10):
KLH Nine: Low efficiency, highly reactive loading characteristic, and superb transient response make this speaker extremely critical of the driving amplifier and input signal cleanness; drivers well-blended; tweeter highly directional, best heard off-axis; sound very smooth, liquidly transparent, likened to a spotlessly clear window on the sound; "Row-M" perspective places sound rather behind panels; somewhat cold sound; two panels (one complete Nine system) essentially flat to around 40Hz with panels together, gradually rolled-off below 60Hz with panels separated for wider stereo separation; with usable bass to around 32Hz in a large room, plus somewhat warmer sound; imaging with two panels (apart) good but optimum listening area restricted; listening area fairly large with four panels, but imaging only fairly good.
This is probably the most nearly perfect loudspeaker we have tested until this time.J. Gordon Holt
Sidebar : Specifications
Description: Full-range electrostatic loudspeaker. Amplifier power requirement: 5OWpc minimum, 75Wpc maximum. Impedance.: 16 ohms nominal (30 ohms max), primarily resistive. Polarizing supply: separate for each speaker, 115V AC, 5W.
Dimensions: each panel, 70" H by 23½" W by 2 7/8" D.
Finish: Choice of mahogany, walnut or oiled walnut.
Price: $1140/pair (1966); no longer available (2006).
Manufacturer: KLH Research & Development Corp, Cambridge, MA (1966).