Product: magnetic levitation antivibration base Relaxa 1 by SAP
Manufacturer & distributor (Italy & Countries with no official dealers): SAP - Italy
USA distributor: Blue Cow Audio 6305 Willowfield Way - Springfield, VA 22150
Approx. price: 600 $/620 Euro (your mileage may vary)
Reviewer: Lucio Cadeddu
Reviewed: December, 2001

Many audiophiles with a technical background have thought, at least once, of designing a magnetic levitation support for their HiFi components. It is sufficient to have a look at newsgroups and HiFi Forums to see how many have tried this (unsuccessfully).
The pluses of such an approach are two-fold: first of all, a completely "levitated" base will completely decouple the HiFi component from the HiFi rack or table, minimizing vibrations by contact. Secondly, the idea of levitation has always fascinated scientists and techno-minded researchers since it reminded one of the greatest man's dreams: to fly.
Sadly, the task isn't all that easy: first of all a full magnetic levitation with passive magnets isn't easy to achieve (actually it's impossible, see later) because the suspended base tries to escape from the equilibrium position with apparent no reason, secondly you need strong magnets and, hence, magnetic fields. These may be not welcomed near phono cartridges, phono stages and CD players, for example.

The first problem (unstable equilibrium) is well known since, say, 160 years or so. The well known Earnshaw's theorem (1842!) states that if you have a static field (let's say, if your magnets are passive and static) then the equilibrium of any charged particle into the field created by the magnets is unstable.
This means that any tiny perturbation from that position will make the particle move away from there (actually, escape quickly!) [S. Earnshaw, "On the nature of the molecular forces which regulate the consitution of the luminferous ether", Trans. Camb. Phil. Soc., 7, pp 97-112 (1842) ].
Later, the result has been extended by Braunbeck (1939) to the case of a dielectric body into an electric field or a ferromagnetic or paramagnetic body into a magnetic field [ Braunbeck, W. "Free suspension of bodies in electric and magnetic fields", Zeitschrift für Physik, 112, 11, pp753-763 (1939)].
Moreover, Tozoni [Tozoni, "New stable magnetodynamic suspension systems", IEEE Trans. Magn., vol. 35, no. 2, p 1047-1054] proved that the Earnshaw's result changes if one introduces non-linearities into the magnetic field.

There are several ways to escape from Earnshaw's limitations: you can use a spinning system (gyroscopic effect) like the ones used on magnetic bearings, you can make the whole system "active" by means of electromagnets controlled via software and position sensors (and feedback), more or less like on MagLev Japanese trains and/or you can make use of superconductors (Meissner's effect) at temperatures near the absolute zero or near -100 C with high critical temperature superconductors.
The one and only passive magnetic suspension that doesn't need any mechanical constraint to avoid unstable equilibrium is the well known toy called Levitron where you need a spinning top to avoid the trouble (hence, Earnshaw's theorem is still valid, since it required STATIC charges).
An improved version of the Levitron has been used - recently - to build a friction-less magnetic bearing.
Anyone interested in reading more about the topic can find almost everything searching the web with "Levitron" as keyword.

Long story short, the magnetic levitation without points of contact and static magnets is impossible to attain.
At the end of the 19th century the first "hybrid" magnetic suspensions were designed and they made use of mechanical constraints (guides) to avoid the unstable equilibrium problem. And this is exactly what Eng. Puppin has designed for SAP when thinking at a magnetic levitation antivibration base, the Relaxa 1 under test.
So here we have it: a magnetic levitation base with eight neodymium magnets, coupled one facing the other (repulsive mode), and 4 linear guides (rails + bearings) that keep the suspended base securely locked into an equilibrium position.
The second problem to solve, as anticipated, was the magnetic field interaction with phono cartridges and electronic circuits. Using 8 strong but small magnets, SAP declares to have been able to reduce the stray magnetic field to a negligible value even close to the magnets.
Indeed, following the measurements provided with the Relaxa 1 official technical papers, the stray field at 2 cms (1 inch) from the magnets is 80 gauss along the x-axis, 44 gauss along the y-axis and 180 gauss along the z-axis and it rapidly decreases with distance.
The Relaxa 1 is a patended design owned 50% by E. Puppin (Politecnico di Milano) and 50% by V. Fratello (SAP's CEO).

The Relaxa 1 has been designed to work as an antivibration base within a prescribed range of applicable loads, from 2 kgs (4 lbs) till 25 kgs (50 lbs), and, according to the manufacturer, any audio or video appliance can benefit from the antivibration effect of the base (CD & DVD players, amplifiers, preamps, VCR's etc.).
The Italian suspended base has been awarded twice at the last CES with the Best Innovation 2001 and Best of CES by TechTV awards.
Its retail price may vary depending on the Country you live in, starting from 620 Euro/600 $, which is the official list price here in Italy.
Aesthetically, the Relaxa 1 looks cool and it certainly has that kind of magic that may attract more than an audiophile.
The quality of the finish is very good, with minor faults for the steel guides and bearings (more on this into the Some advice and complaint section).

Does it work?

The Relaxa 1 is a very special device and the effect it produces varies greatly depending on the HiFi component you use. For this reason the listening tests took me much more time than planned.

I've tested the Relaxa 1 with different HiFi components: CD players, turntables, preamps, and integrated amplifiers and the results always varied from "yes-yes" to "no-no", including "so-so" situations, where some parameter benefited from the use of the base, some other did not.
For example, on some el cheapo CD player - where only insane audiophiles may want to use a 600 $ antivibration device - the result was pretty good in terms of transparency and richness of detail, especially in the mid-high range. Surprisingly, the already so-so 3D soundstage collapsed and you can only imagine my surprise, since I was expecting a BIG improvement of this parameter.
Cheap CD player image poorly and anti-vibration tweaks normally tend to improve the situation. Not so with the Relaxa. Quite strange. I have no logical explaination for this.

The best effects were achieved with semi-rigid chassis turntables: the sound became cleaner, with plenty of detail and harmonics and even the 3D virtual soundstage became way wider, more focused and realistic. Thumbs up.
Anyway, the final result on turtables greatly depends on the kind of chassis, suspension and arm/cartridge resonancy frequency. You should consider the Relaxa 1 has a resonancy frequency (unloaded!) as low as 2 Hz and it decreases with load.
Hence, effects (and benefits) with analogue turntables should be evaluated personally, case by case.
It seems the semi-rigid tables benefit from the use of the Relaxa 1 so one can guess even rigid ones should. As Rega owners know pretty well, it is hard to improve the performance of their t-tables, as they are an almost magic-mix of compromises and parts that work perfectly together to achieve a result that is far better than the mere sum of its parts.
On preamps and integrated amplifiers, the results are predictable and quite good, though it is not a kind of day-night difference. Let me remark that I have been able to test the Relaxa 1 only with light preamps and amps, avoiding power amps and tube gear because of the load limit of the Relaxa 1.
I'm pretty sure that, considering how microphonic and prone to suffer from mechanical vibrations the vacuum tubes are, the benefic results achieved with solid state gear (smoother sound, wider soundstage, cleaner treble and mids) could be greater with tube gear.

Finally, as we have always remarked here on TNT-Audio, anti-vibration devices with damping or not (the Relaxa 1 is damping-less), normally modify the way a HiFi component sounds. There's no such a thing as a free meal and it is always a give-and-take game. You lose something, you earn something. HiFi designers and expert tweakers are well aware of this non-written law.
Actually, it is not the minimization of the vibrations that counts, the interaction of the different resonancy frequencies being far more relevant. We all know every HiFi component vibrates and reacts differently both to airborn and (contact) vibrations.
It is quite impossible to predict how an anti-vibration device works with a component. The Relaxa 1 is a device that resonates at 2 Hz (which is good) but my tests have proven this is not always welcomed by HiFi components, one has to try and listen to independently, as every system behaves differently.

Some advice and complaint

Admissible load.
The maximum load you can put on the Relaxa 1 is 25 kgs (slightly more than 50 lbs). This is enough on many applications but too low for power amps, heavy turntables and CD players, tube gear. Normally high-end stuff tends to be massive and heavy :-) so the use of the Relaxa 1, though logical considering its price tag, proves to be difficult (or impossible) with proportionally priced HiFi components. Of course, you do not want to spend 600 $ on a suspended base and 400 $ on a CD player. For sure, a 1,000 $ CD player sitting on soft (1$) doorstoppers will outperform the exotic combo :-)

Size.
The Relaxa 1 isn't exactly small, being 50 cms (20") wide, 38 cms (10") deep, 6.5 cms (2.5 in.) tall (when unloaded) and its weight is around 4 kgs (8 lbs). This means it is rather difficult to place it "into" a standard 19" HiFi rack and one is forced to put it on the top shelf, where CD players and turtables sit.
Even if you succeed placing it on an intermediate shelf you should be careful since the height of the HiFi component plus the height of the Relaxa 1 can easily exceed the distance between the shelves. It happened to me.

Price.
After all's been said and done, one needs to evaluate what he has got for 600 hard earned bucks: 8 small neodymium magnets, a class 1 metacrylate platform and an aeronautical compound rigid base. That's it, plain and simple.
Of course, you need to take into proper account the cost of the research behind the device but I still consider it too expensive, in view of the sonic results that are far from being universally applicable (more on this later).

Isolation performance.
Considering the technical papers that come with the Relaxa 1 and the practical tests I've performed, its effectiveness to avoid VERTICAL vibrations on the base passing to the suspended top platform is stunning. You can be pretty sure that if your HiFi rack vibrates on the vertical plane only, this vibration won't reach the component that sits atop of the Relaxa 1.
Good, but still not enough. Horizontal shocks hitting the rack or the base itself, due to the presence of the vertical guides, are perfectly transmitted to the HiFi component. A simple test with a playing turntable tells the whole story. A small hit on the rack becomes a big bump on the speakers, unless the turntable makes use of a good suspended subchassis itself.
With a soft, horizontally free to float damping device this problem is absent.
It could be interesting to investigate how many vibrations that reach the HiFi components from the environment (actually, from the loudspeakers) are of the "vertical" kind only.
This is not a fault of the design, just a unavoidable drawback because of the Earnshaw's theorem requiring the vertical guides :-)
The Relaxa 1 designer states this clearly in the technical papers, where he refers only to "vertical environmental vibrations".
If I've understood well, the guys at SAP are working on a new kind of magnetic suspension (like a foot) that will be free to oscillate even along the horizontal plane (perhaps by modifying the rails).
Finally, let me state clearly that, unlike most anti-vibration devices in the HiFi market, the Relaxa 1 is a damping-free platform. This means the mechanical energy isn't transformed in heat (only negligible amounts due to the deformation of the platforms), as it normally happens with devices that work by deformation.

Mass balance.
Another drawback of this design is its sensitiveness to uneven loads. If the HiFi component is perfectly balanced (from the point of view of masses) putting it perfectly centered on the top platform of the Relaxa 1 will make the suspension "sag" in a purely symmetric way at the 4 corners.
Sadly, in the real World, HiFi components are everything except mass-balanced, actually they weigh more at one side, where there's the power transformer, for example. Even worse, think at a turntable with the motor housed into a separate chassis...then try to balance it on the Relaxa 1. Not easy. Nearly impossible, actually.
This means that, after you've spent the best years of your life trying to level your turntable or CD player....your efforts vanish thanks to the Relaxa 1. Oh yes, you can try to level it again by means of washers under the component's feet but let me tell you this is a no-no and for various reasons.
First of all, after you've spent the savings of a whole life on an exotic esoteric HiFi component, you wouldn't be happy to have it levelled as it was an old grandma's desk ;-)
More seriously, adding different materials under the feet of the component will modify the mechanical coupling of it with the top platform - unevenly - if you use a washer on one side only (trying to counterbalance the Relaxa lowering effect).
Moreover, if you've attended a physics class when you were at primary school, you should know that if the top platform is unevenly placed, the load on the steel rails is uneven and so the friction on the bearing increases....not good (the vector component of the load is not perfectly vertical).

Now, this "out of level" problem could be negligible on preamps and amps...but it becomes very relevant on turtables and CD players, which are, incidentally, the components which SHOULD benefit more from the Relaxa 1.
With lightweight CD players the problem is even worse, since the stiffness of the magnetic suspension is inversely proportional to the load. This means that with a light load the suspension is very soft and more prone to suffer from an unbalanced load.
This was a well known problem of hybrid magnetic suspensions, considering the stiffness variation is NOT linear with the load (it's quadratic, actually).
This, in man of the street words, means that the suspension of the Realaxa 1 behaves differently depending on the applied load. The manufacturer claims it is a negligible difference...but I'm not all that sure.

Magnetic field influence.
This was one of the main problems, considering the magnetic field generated by the neodymium magnets could interact with phono cartridges and electronic circuits. According to the technical papers supplied by the designer the stray field near the platform is negligible and decreases rapidly with distance (for values, refer to the introduction above).
Using two different MM phono cartridges I've detected no negative influence of the field, indeed. Anyway, long-term effects of the exposure to low-value magnetic fields should be investigated, especially when dealing with delicate magnetic devices such as MC or MM cartridges. It is well known that magnets tend to demagnetize when exposed to repulsive interacting fields.

Friction and craftsmanship.
Just two words on the steel rails used to constrain the top platform in its equilibrum position. They are said to be of the low-friction kind...actually they work as a stiff connection between the base and the floating platform and, even worse, the DO have friction.
It is sufficient to push the top platform against the base with the hand to "feel" the movement of the bearings rolling on the steel rails. Not satisfied with this "manual" test, I've verified the action of the bearings by means of a sthetoscope. Then the NOISE (steel against steel) of the bearing became pretty obvious (and strong) both at the "guided" corners and even at the free floating ones (the noise spread all over the top platform, actually).
The guides (rails) are claimed to be stainless steel but, on the two samples I've received for tests, after a couple of months, steel oxide build-up was clearly visible, hence making the smoothness of the bearings operation even worse.
This also means the quality of the "stainless" steel is poor. If you are familiar with REAL stainless steel (marine applications, for example) stuff you probably know this is non-magnetic, in the sense that a magnet doesn't attract it. The steel guides of the Relaxa 1 DO attract magnets.
This is an old trick I've learnt from the guy who takes care of my sport car. When he changed some steel screws on the rear plate of the car he used stainless steel ones for increased durability and coolness factor. How did he choose them? With a magnet. Those which weren't attracted by the small magnet were OK for my car. I love this guy :-)
Now, I understand the stainless steel rails of the Relaxa 1 have been CNC-machined (sort of) in order to achieve maximum smoothness. This process has probably destroyed the stainless steel properties of the rails. Unavoidable? No, not on a 600 $ device. Just use titanium or metal matrix rails: non-magnetic and rust-free for a lifetime.
Also, IMHO, non-magnetic rails are the way to go, considering the strong magnetic field near them. This magnetic field could eventually make the guides flex, thus increasing friction.

Conclusion

The SAP Relaxa 1 is/should be every audiophile's secret dream: a full floating suspension for HiFi components. Actually, because of Earnshaw's theorem, the designer has been forced to add steel rails to the magnetic suspension....thus locking it on the horizontal plane and making it useful to minimize only vertical vibrations propagating from the base.
As explained, the Relaxa 1 is NOT an universal device. Where it works, it almost makes wonders, where it does not work, it is a no-no.
So, careful personal evaluation should always be perfomed before buying, especially considering the amount of cash involved.

Now I'll wait for an electro-magnetic software-controlled feedback-activated levitation platform...and I'd expect to hear some audiophile whining because of the use of feedback, normally considered as bad sounding :-)