• Constant Directivity WaveGuide
  
• Cabinets
  
• Crossover
  
• FST
  
• Neo panel
  
• Bass drivers
  
• VMPS Bass system
  
• Minimum phase

• Reverse Measurements 
  
• Constant directivity planars from VMPS
  
• Why VMPS is so darn peculiar in so many different ways?

• Money and the high-end - The price is the product
  
• Cone Talk - Materials
  
• Why suspend a spider?
  
• Neodymium

• We won the show
  
• Tips on biamping the ribbon monitors
  
• How to Adjust Bass Damping on Floor-Standing VMPS Speakers
  
• How to Turn Your Room into a Good Acoustic Environment

• The thick and the dead (Stereophile '89)
  
• Hi-Fi Retailing: The Avid Audiophile's Advisory (Stereophile '91)

Constant Directivity WaveGuide

CDWG is the the 17mm wide full range planar constant-directivity speaker/waveguide. Patenting of this technology is in the process. From the outside CDWG looks exactly the same as regular grilles.

The key to the invention lies in a peculiarity of planar panels: most of the output comes from the center of the transducer, or within about 12mm either side of the center vertical axis. This area is responsible for about 75% of the speaker's radiation over its operating range. 

It should therefore be possible to reduce the width of the panel by suitable means to a value which would enhance its directivity while reducing its sensitivity only slightly, 1 to 2 dB of SPL. This must be done without reflecting a lot of energy back into the diaphragm causing frequency response peaks/notches and coloring the sound unacceptably. In other words, one might suspend a disco ball in front of a speaker and improve its directivity, and in so doing destroy whatever good sound qualities the system already possesses. 

After much experimentation devised was a slotted waveguide the width of which was the same as the highest frequency for which Constant Directivity was desired. Selected frequency is 20 kHz and a width of about 17mm. The underside of the waveguide, which contacts the panel stators directly, features considerable absorption and some diversion. 

The question arises whether enhanced directivity is that important or merely an intellectual exercise. There was no way of knowing without building a CD speaker. The results of imposing the waveguide over an RM 30 or RM 40, as first test subjects, were spectacular. While the tonal character of the speaker was recognizable, there was much more palpability to the image, more of a "they are here" sense, more naturalness--all to a enormous degree. Incorporating the tweeter with its post-20 kHz bandwidth into the waveguide transformed the speaker system into a true full-range CD radiator. You can sit anywhere in front (or directly inbetween) the speakers and enjoy the effect.

Since on axis and off axis response are the same with CDWG, it is much easier to linearize the speaker in a given room, by placement, toe-in, and level control adjustments, or electronically. The Constant Directivity Wave Guide also eliminates lobing inherent to first-order filters in crossover regions, improving fidelity and linearity not possible by other means.

The price you pay is a 1.5dB reduction in midrange sensitivity, which can be compensated for with the level controls which are normally set about 4dB below max output, AND, most importantly, a 6dB/oct rolloff in the trebles above 10kHz, which can be compensated for by various means.

Because a CD configuration spreads the same amount of treble energy over a much wider angle, there is a gentle high-freqency rollof which requires compensation for measurably flat amplitude response. A 6dB/oct boost centered in the 8-10kHz range is necessary. This is easily accomplished by means of: 

1. increasing the tweeter level via a potentiometer such as that already in circuit with VMPS tweeters;
2. a standard Baxandall-type treble tone control like on preamps of old and some pre's and integrated amps of today; 
3. an equalizer with a 10kHz hinge point; 
4. a passive EQ which impresses an inverse rollin to the mids; 
5. a dedicated supertweeter operating at a higher level than the main tweeter, crossed over at 6 dB/oct at a higher frequency than the main tweeter; 

...and several other remedies such as the new digital speaker correction or room correction outboards (SOCS, DEQX, TACT and their kind). 

Many listeners will find the CD treble balance just fine as it is. CD trebles just sound smoother, more realistic and better defined, particularly in relatively undamped listening environments like almost all homes.

Goal was that all VMPS CD speakers work well with and without CD EQ, even if that only means turning up the tweeter level controls a little. The best way to bring up the trebles subjectively is to tighten the bass. We have engineered new low Q passive radiators and lowbass woofers for that purpose. Owners purchasing CD wave guides for their existing speakers will also get the new PR's. 

We find the sound of the "revoiced" CD speakers done so far to be punchier and clearer, with no sacrifice of extension. A felicitous byproduct of CD technology, this change should please everyone who tries it.

Cabinets

VMPS cabinets are normally made from MDF (medium density), except RM/X which uses even heavier HDF (high-density) and ST III which bass column has synthetic granite front plate. 

RM/X cabinet is a piece of art in cabinet making, both technically and cosmetically. Extremely complicated CNC carved shape of 14cm thick HDF frontplate ensures completly stable board for bass driver and an "invisible" surrounding for array of six Neo panels. FST tweeter is housed in completely separate chamber which is angle adjustable. Cabinet is finished in 7 layers of automotive lacquer.

New MLS cabinets, available for 626R, RM30, RM40 and LRC, are made from HDF, and are heavier and more extensively braced than standard cabinet. Finishes are beautiful, in choice of light and dark genuine wood veneers and high-gloss piano black. All MLS cabinets are delivered in custom-built wood crates.

Standard oak cabinets of RM2 have massive genuine oak sides and edges.

Midrange chambers are damped with natural lamb wool which is the top-quality choice for damping rear midrange radiation. There is no doubt in my mind the lambswool is a step forward in clarity and naturalness. Lamb wool is also not a good choice for bass section, being not reactive enough for good low frequency absorption. For bass we use fiberglass.

Soundcoat is ceramic compound which is in liquid state when applying and hardens when dry after a couple of days. It is very effective in damping unwanted cabinet resonance, around 10dB in audio range above 200Hz. Newest option for even beter results are BlackHole 5 self-adhesive plates.

Crossover

VMPS uses first order networks and in-phase drivers because that emulates live music best. Material used as standard is first rate (Erse "Super Q", Solen, Bennic, Axon, Vishay), and absolute best capacitors from Auricaps and TRT are offered as option at additional cost. We build passive crossovers to very tight tolerances. For us that means four decimal places, or 1/2000th of 1%. Each crossover is trimmed to exact value, no tolerances. This means, for example, 1.600uF, not 1.6uF. You only get repeatability with high levels of precision.We laugh at the tolerances (5%, 1%) even the best competitors use. All connections are soldered point to point, no printed board is used. This solution does not look very good, but sounds much better.

Our "Super Q" Erse coils are superior to any aircore, since their DCR is about 35% of an equivalent gauge aircore of the same value. It has nothing to do with cost, the Erse's aren't cheap either. DCR is the most important characteristic of a series inductor as long as it does not saturate or suffer from hysteresis. The very large laminar core of the Erse's eliminates both problems and sound just wonderful, which is why we use them.

Crossover includes two L-pads for level adjustment of midrange panels and tweeter. We use them instead of fixed resistors because L-pads allow a 1/20th of 1dB change in level to be effected by the user. Often such a change is all that is needed to tame hot trebles or a forward midrange. A fixed value does not allow the user to adapt the music to his tastes. Also, resistors are not made with sufficient tolerances to give us the precision we feel necessary for level changes. You can't get this level of precision with a stepped attenuator or a resistor network. In fact even 1% tolerance Vishays (which we use elsewhere in the xover) do not permit this level of precision, which can be very important when you switch equipment or IC's/speaker wire.

Also, fixed parts change value with age and there's nothing you can do about that. L-pads change too, but all you have to do is move the wiper over one winding and you're back in business. Fixed values are invariably determined by measurement, i.e. whatever meassures flattest with your mic and test gear and environment. However, bass below 200Hz is boundary-dependent; fixed values do not allow for changes in room or placement. You're going to be wrong virtually 100% of the time, no matter what you think your measurements tell you is best. Our L-pads allow great flexibility and microadjustments that nothing else approaches. Rather than a compromise, the Lpad gives the user essential freedom to adjust the system to taste. The level control adjustments are essential to obtaining optimum performance from any of our systems. This also comes from general knowledge of basic acoustics and psychoacoustics.

Crossover and midrange/high frequency drivers are connected with custom silver/teflon wire. Standard wire for bass connection is Monster copper wire. Option for this purpose is Bolder silver wire.

Binding posts we use are selected because they have been proven best for a number of years and are best suited for accepting spades. They are solid brass/gold plated, with 6mm shafts, solid metal nut that can be tightened with a nut driver. There is no hole through the center of the shaft (for the bare wire) which weakens the shaft and causes breakage.

FST

"Free swinging tweeter" is the latest true-ribbon tweeter used throughout VMPS line. It is one of the best tweeters available today. It's advantages over previous spiral ribbon tweeters were easily audible and cost is not overly prohibitive. We use custom version (8 Ohm, custom transformer and diaphragm) of similarly looking Aurum Cantus tweeter. The FST tweeter has 30 degrees of vertical dispersion, less than 25 mm dome tweeter, but does not become directional at 11 kHz like a 25 mm dome. Horizontal dispersion is broad and even up to 25 kHz.

Foam pads surround the opening of the diaphragm, slot-loading the tweeters for better horizontal disperson. Foam pads won't be necessary when all VMPS speakers are converted to CDWG.  

Neo panel

The star of the VMPS ribbon speakers are their midrange panels, a rebuilt Level 9 products. "Neo" comes from Neodymium, element that is a part of magnet material (NeFeB). Level 9 built about 300,000 of these panels most of which ended up in computers, where they worked OK. In high fidelity and automotive applications they failed at a high rate due to a mechanical design flaw, which requires us to rebuild each one. With our rebuilt they have been durable and good sounding. They are the only such panels which extend down to 200 Hz and below, a critical range. Neo panels are available only at VMPS loudspeakers. Similar competing products are not nearly up to it, not having enough upper bas response, or having large nonlinearities etc.

Bass drivers

We make our own bass drivers for decades and do not use anything short of best possible for the purpose.  That means custom made Woven Carbon Fibre cones in various sizes, natural rubber surrounds, 300W voicecoils, 2.3kg magnets. Driving our woofers with 20Hz signal to loudness of 115 dB we can see (not just theorize) cone travels of at least plus/minus 15 mm. Also, we are unique in suspending the spider on bass drivers. Check the articles Cone Talk - Materials and Why suspend a spider? to learn more on bass drivers construction. 

All drivers has phase plugs instead of usual dustcaps. Appropriately designed phase plugs are superior to any kind of dustcap, inverted or not. We haven't used dustcaps on our drivers for more than 10 years. The phase plugs are precisely the length and bluntness needed to move the woofer acoustic center forward to match the planar sections.

In place of noisy ports at large speakers there is a tunable passive radiator (PR). It is plenty fast for music and has the heft, extension and output levels for HT. Sealed designs either cut off an octave or two higher than the PR design, or are very inefficient. Not what you want.

A PR is basically a tunable vent and allows you to match system Q to the series R of your speaker wire and output impedance of your amp. A necessary feature for good bass. While we tune the PR at the factory for what sounds right on our gear, yours and your room are certainly different. Just follow the setup instructions.

So, all our drivers are custom and almost all are made in US, by VMPS or custom for VMPS. There is no clones of our speakers out there.

VMPS Bass system

The Passive Radiator (PR) located in the bottom of the cabinet already has a certain amount of mass attached to the center, by the factory. By adding or subtracting mass from the PR, it is possible to make system tonal balance warmer (higher Q) or tighter (lower Q). This is accomplished by changing an inert mass such as Mortite rope putty, the substance attached to the PR. You can buy additional putty at most hardware stores but your speaker is delivered over-damped (a bit too much putty), so that in almost all cases, you will tune the speakers by removing mass from the PR. Mass is accessible by inserting your hand into the slot formed by the base and the bottom of the cabinet. Removing a very small amount, no more than 1gram of putty, will be sufficient to make the adjustment.

Moving mass of PR is very low. Since a PR is driven equally over its entire surface by the active woofer's backwave, the diaphragm will move pistonically even if it is not rigid. The PR cones are treated paper. Paper is fine as long as you don't have to listen to their high frequency noise and distortion products. Facing the PR down and slot-loading it out the front filters such products out nicely. 

We invented the slot-loaded PR and decided not to patent it, since we would spend our life in court defending against copies. We knew we were on the right track when we saw a Klipsch monitor at the AES in 1984 with a slot-loaded PR.

Bottom of the Larger subwoofer that forms a slot-loading with the floor,
showing 38cm PR with added mass at the center of the PR

Minimum phase

Minimum phase response means that the envelope of frequencies arrives at the ear at the same time and in phase. 

To do this we use first-octave crossovers (phase linear) and wire all drivers electrically in phase. Phase plugs in the woofers move their acoustic center forward to the blunt tip of the phase plug, and into the same plane as the mid and treble ribbons. Since these are planar they do not need to be moved back relative to the woofer's acoustic center. 

There are "time aligned" speakers that wire midranges electrically of inverted polarity relative to the woofer and tweeter, which results in measurable better amplitude response. Of course it also destroys the integrity of the music signal outside the crossover region. Phase changes 180 degrees as the signal exits crossover and into the passband, a highly deleterious and audible effect.

If you have more technical questions, please check FAQ or send us an e-mail.