[ Home | Staff & Contacts | DIY & Tweaks | Listening tests | HiFi Playground | Music & Books ]

Yamamoto MGB-1 & Yamamoto MGB-2 Magnetic Floating Bases

[attracted by repulsion, dear reader?]
[Italian version]

Manufacturer: Yamamoto Sound Craft - Japan
Product: Yamamoto tuning MGB-1 Magnetic Floating Base
Price: 38,000¥ for four or 10,000¥ each
Approximate European cost: ≈311 € for four or ≈83€ each (YMMV depending on currency fluctuation & swinging import duties)
Product: Yamamoto tuning MGB-2 Magnet Floating Base
Price: 28,000¥ for four; 7,000¥ individually; each supports 0.9-3kg per foot
Approximate European cost: ≈228€ for four, ≈57€ individually(YMMV depending on currency fluctuation & swinging import duties)
Reviewer: Mark Wheeler - TNT UK
Reviewed: Winter 2009/2010

[attracted by repulsion, dear reader?]

"The old scribe's selling out now," accuse plebs, stage left, "He never used to review tuning and vibration control feet at these price levels. He's finally becoming seduced by the high end."

How can your old scribe resist products named MGB? Until my sons were too big for the little back seat I ran MGB GT's for 15 years, the later one evolving with the kind of upgrades and suspension tuning I apply to audio components; Bilstein adjustable telescopic gas strut conversion, lowered springs, anti-sway bars....

"Proof if ever it were needed!" assert plebs, stage left, "He's allowed the MGB nomenclature to bypass his cynic filters."

The Yamamoto PB-22, the least expensive tuning devices in the Yamamoto range were a great hit at the pocket money price (≈7.50€) and these Yamamoto MGB-1 & MGB-2 represent the other extreme of Yamamoto San's range of support devices. While the combinations of turned pins and sockets in ebony, cherry wood and/or brass are known to Mr Shigeki Yamamoto as 'Setting Bases', these magnetic elevators are known as 'Floating Bases' as they enable the audio component to float on a magnetic field. That's the theory anyway; in practice some form of tethering device will always be needed in a magnetic support to prevent the object being repelled away until the audio component slips sideways on the magnetic field and falls off completely. The tethering of the MGB-1 and MGB-2 is accomplished by a bolt secured by a nylon nut; the bolt's threads being lubricated from the surrounding ebony by a ptfe sleeve. It is possible to feel some variation in friction as the edge of the holes moves over the ridge and furrow of the bolt thread encased in ptfe and I do wonder if set up could be made easier if there were a brass sleeve (therefore non-magnetic) between the bolt thread and the ptfe sleeve, creating an even surface.

I did wonder if the slight notchy friction might tend to short circuit the magnetic circuit and thus a more flexible tether like a web of nylon filament might improve isolation. It might also ease set up as precise alignment to ensure every MGB-1 or MGB-2 bolt is parallel and free to move easily. Equally it would be useful if both the MGB-1 and MGB-2 were the same height. Then the MGB-2 could be mixed with the MGB-1 under chassis with unevenly distributed mass loads like valve amplifiers with a row of transformers at the back, as is common practice.

"Why bother with magnetic springs?" demand plebs, stage left, "Surely steel springs have done a fine job in subchassis for years?".

Magnetic Springs, Ohio residents might resent the question "Why bother with Magnetic Springs?", but in this context there are a number of advantages. The steel spiral of a coil spring is not a pure low-pass filter. The steel structure has its own resonant series due to its own mass, rather than the mass the spring is employed to support; this is the reason some subchassis turntables used to employ foam plugs within the spring. These resonances are a frequency selective short circuit. The spring rate of a typical steel spring only applies over a small range of movement too, whereas magnetic springs should be linear if the fields are suitably concentrated by magnet shape. The term neodium magnet used by Yamamoto San means neodymium magnet (NdFeB - increasingly popular in loudspeakers due to low mass) and seems to be a common spelling in some languages.

Yamamoto Specification
type MGB_1 MGB-2

dimension & units
Height 47-58.8mm 37-45

load per piece 2kg-10kg 0.9kg-3kg

diameter 59mm 49mm

Hence, set up proved to be almost as tricky as a suspended subchassis turntable. To get the best from these devices and to experience the effects I will describe, demands some skill, some expertise and a lot of patience.

  1. The number of MGB-1 or MGB-2 must be chosen to suit the mass being supported so that the weight and the opposing magnetic force are equal where there is movement each way on every MGB-1 or MGB-2
  2. The position of each of the 3 or four Magnet Floating Bases must be such that the distance separating the two magnet surfaces is parallel and equal
  3. The position of each MGB-1 or MGB-2 must be such that top and bottom are perfectly centred to minimise friction with the tie bolt

This last, condition 3, is where an additional brass sleeve would reduce the need for such accuracy, adding to the complication of meeting conditions 1 and 2. Hence each set up is time consuming and easily disturbed. The larger models, the Yamamoto MGB-1 have the additional feature of 3 domed ceramic discs inserted into the contact face of the top and bottom; experience with the Yamamoto PB-18 speaker Setting Bases suggests this is a desirable feature that added clarity to that product's performance. It proved impossible to tell if the ceramic discs on the MGB-1 made any difference as their different load ranges mean that different numbers of MGB-1 or MGB-2 Magnet Floating Bases have to be used under any given item of equipment. This introduces two variables and thus negates the experiment.

"So is it worth the effort?" ask plebs, stage left, perfectly reasonably.

The answer, like ALL audio tuning products, depends on the equipment being tuned. Once again valved equipment highlighted the effects strongly, as did electro-mechanical products like cd players and turntables.

Yamamoto MGB-1 meets Garrard401/SME 3012series I

The Yamamoto MGB-1 are the best isolation device I have tried with the Garrard 401/SME3012 series I. Using four MGB-1 within the plinth (a 1970's construction previously utilising furniture webbing as suspended isolation) spaced to provide a level platform from the sand filled plywood inner plinth compared to the previous best combination gained plus 12 and lost minus 3 using Mark's judging system. The previous best result was a mix of one Yamamoto VB-18 V-groove base under the stylus tracing arc combined with multiple isonodes. The VB-18/Isonode combination is slightly better than the MGB-1 (+1 on my Lickert Scale judging system) in just three parameters:

The first I suspect due to the VB-18 V-groove base and the latter two parameters due to the Isonodes. However, the set of Yamamoto MGB-1 gained similar slight +1 improvements in:

More significant than the 'just discernible' difference that ±1 represents, the Yamamoto MGB-1 succeeds in gaining +2 (significant difference) in the parameters:

I suspect a combination of VB-25 (which I have not tested, only the smaller VB-18) positioned under the playing arc and MGB-2 (the load would be much reduced by the VB-25) providing isolation at the corners would be unbeatable. Furthermore, the accurate location provided by the V-groove bass would remove the need for the tie bolts in the MGB-2, thus increasing isolation further. Until that is explored there is no real competition.

The Yamamoto MGB-1 moves into prime position as the most successful support for the Garrard 401/SME 3012 series I combination that I have tried.

Yamamoto MGB-1 vs ERaudio on Shanling CDT100c
parameter MGB-1 over ER stock feet over MGB-1

PRaT & emotion group
Rhythm 0 0

Timing -1 +1

Pace +1 -1
Tingle factor 0 -1
Extension +1 -1

pitch/tune +1 -2

speed 0 0

tautness -1 -1

midrange & vocals
pitch/tune +1 -1

intelligibility/clarity +1 -2

realism/timbre +1 -2

embodiment +2 -2

pitch/tune 0


clarity 0 -2

timing -1 0/-1

sweetness 0


hifi attributes
soundstage stability/realism +1 -2

soundstage width +1 -1

soundstage depth -1 nearer -2

soundstage height 0 -2

other essentials
fuzziness (lack of) +1 -2

virtuality +1 -1

micro-dynamics +1


macro-dynamics -1 0

totals +12 & -5 +1 & -29

Yamamoto MGB-2 vs ERaudio SpaceHarmoniser on cones
sited on Something Solid XR4

Moving through the reference gear Chez Wheeler (a mere audio revolt in the shadow of another Che) brings our ears to the valve outputs stage modified Shanling CDT100c that has proved very tuneable by support already. In order to begin with the best baseline performance possible the test is conducted on the already excellent Something Solid XR4 rack and ERaudio SpaceHarmoniser (on ERaudio steel cones) that have proved the best combination unter this player. Hence this test examines whether the Nth degree of extra performance is possible with these magnetic isolators. If they can perform here it will be a triumph, if they show no advantage we will move to another test using the Yamamoto Magnetic floating Bases on an ordinary audio frame and shelf.

Yamamoto MGB-2 meets Shanling: Summary

Phew!!! what the table on the left means is when supporting my extensively modified Shanling CDT100c that the Yamamoto MGB-2 bettered the ERaudio SpaceHarmoniser (on ERaudio steel cones) by 12 points (spread over 11 parameters) and was slightly inferior by 5 points over 5 parameters. The quality of embodiment, that sense of believability that a human body is producing human voices on recordings, is where the MGB-2 beats the ERaudio by two points indicating a substantial improvement and the ERaudio SpaceHarmoniser excels here too so the MGB-2 is at the top of a lofty league. In all the other parameters the differences are subtle but repeatable in reverse. The Shanling was close to bottomming out on three MGB-2 so this test tried both three and four MGB-2; surprisingly there is no discernible audible difference, which is counter intuitive as (if vague recollection of school physics is accurate) the simple harmonic motion equation's use of Hooke's law K states that the resonant frequency of a system of four will be higher than that of three, dictating the frequency at which the system acts as a low-pass filter.

The ERaudio SpaceHarmoniser is much more expensive than either three or four MGB-2 so the Yamamoto Magnetic Floating Bases demonstrate a clear and cost effective advantage.

Both the tests annotated in the table on the left were confirmed by reversing the condition and checking that the figures apply equally in reverse. This check is conducted without the notes from the previous test visible (although memory obviously undermines this attempt at reduced bias) and where a difference in results repeatedly occurs both figures are given, otherwise results are averaged. This is obviously time consuming when items require careful set up, but one should never trust results that from audio reviews that are not conducted with similar rigour. [rack maxxed]

Comparing the trio of Yamamoto MGB-2 with the stock composite feet on the Shanling CDT100c demonstrates a total improvement of 29 or 30 points, including a difference of +2 in 11 parameters! Interesting the one area with a just discernible edge for the stock feet is overall timing. Treble timing alone is better on the MGB-1 by the same small magnitude, and global top-to-bottom timing suffered the same subtle blurring in the ERaudio comparison. The Yamamoto MGB-2 is clearly a great match with this very transparent valve output stage cd player, lifting its performance by a greater magnitude across more parameters than an equivalent priced cable upgrade could achieve. On an ordinary audio rack or coffee table this score would increase by more than 50%. Recommended.

MGB-1 & MGB-2 meet other hardware

Results are less clear cut with clanky cased solid state equipment. The legendary Accuphase T101 tuner was pressed into service to hear Lenny henry (better known as a stand-up comedian) in the title role of Othello in a BBC Radio 4 production of Shakespeare's play. I have seen Lenny Henry live less than 3 metres from me on the stage apron and heard him many times on radio and TV (including his debut) since my teens as we are similar ages from the same part of the country. I know what Lenny henry sounds like and the vocal embodiment improvement noted with the MGB-1 prompted me to try this as a test piece. Halfway through the play I removed the Isonodes from under the weighty Accuphase and placed various combinations of Yamamoto MGB-2 Magnetic Floating Bases. I left the BrightStar Little Rock on top of the tuner. The MGB-2 were slightly inferior in several respects under the Accuphase. I suspect it is more important to damp the clangy steel casework with lossy materials like the Isonodes than it is to provide improved isolation from structure borne vibrations using sophisticated apparatus like the MGB-2.

Positioning the Yamamoto MGB-2's under the Avondale Audio AAA5 did not maintain its trademark PRaT excellence compared with Avondale's own aluminium and carbon fibre feet, which cost just 45 (≈50 €). Given that the raison d'être for the Avondale is it's Flat Earth supremacy, this renders the MGB-2 unsuitable with this player even though it does improve some other aspects of performance. Under the recently reviewed AAAAVT SM-12B the Yamamoto MGB-2 the performance gains of a set of floating feet costing more than half the cost of the pre-amplifier they support is not a wise spend; the money would be better spent on either a better pre-amplifier or on component upgrades, circuit modifications and chassis isolation and damping (perhaps some vibration isolation on the mains transformer).

So we turn from trying to improve on the already superb Something Solid XR4 rack, and two lesser audio racks to the Ikea Hol table as a relatively inert (by furniture standards) real wood table of comparatively high mass (compared to a really good specialist audiophile turntable table) beneath a weighty valve amplifier. This combination is more vibration sensitive than a lightweight valve amplifier (as described in my first TNT-audio article Box clever perched on the same platform. Incidentally I tried the remote transformer of that power amp on a single Yamamoto MGB-2 and the superleggerra body of it on 3 more. Sadly the amplifier body was too light to compress the magnetic springs and the single MGB-2 too unstable to avoid the bolt thread binding on the magnets. I tried all possible combinations available in stock for this review as magnetic isolation is still novel enough (the Relaxa shelf being the only established product I know of using this technology) to deserve the most thorough examination. We should despair of those reviews of tuning devices elsewhere, that claim to establish opinion without proper evidence, having only the briefest of trials under a couple of components in their main system (often unfamiliar review components). Sadly this is because these reviewers regard high end components as sexier and fail to understand the significance of vibration control (understood perfectly by manufacturers like Naim who use suspension systems and non-magnetic masses in engineered internal solutions). Those reviewers still regard wire as more important (and possibly sexy, cable is perhaps slightly more phallic than feet) than mechanical tuning and vibration isolation so their half page of A4 isn't worth the paper or bandwidth it consumes. That, dear reader is the justification for nearly 7000 words (over half a Master's degree dissertation) on a single product. And the properly assembled qualitative data presented in the accompanying tables.

However, this test offered hints of greatness that leads us to the final and ultimate test.

Yamamoto MGB-1 vs ERaudio platform on non-audio table

My Assemblage SET300Bsignature has undergone extensive subtle upgrading and tube-rolling and fully justifies the 2500€ that would be needed to build a replica. The Achilles' heel of the SET300B is its rigid steel chassis. This is a bad mix with microphonic valves, and octal base versions do tend to microphony more than b9a types (which tend toward poor crosstalk between twin triode sections by comparison with their octal electrical equivalents) and the GiS Alnic Audio Absorb Gel Isolators still haven't appeared in octal versions despite a photo to the contrary on their European importers' website. Hence, the MGB-1 have a tough task beneath this heavyweight power amplifier, and four of them sink to just 2mm. The question is whether the Yamamoto MGB-1 will have any effect on a really heavy power amplifier. The TJ Meshplate 300B output valves are much more microphonic (and euphonic) than the Western Electric pair. So the WE's are pulled and the TJ's inserted for the first run through. Interestingly a validation check with the WE's produced remarkably similar results well within test-to test variability.

Yamamoto MGB-1 vs ERaudio cones under SET300B
parameter MGB-1 over EAR feet

PRaT & emotion group
Rhythm 0

Timing +1

Pace 0
Tingle factor +1
Extension 0

pitch/tune +1

speed 0

tautness +1

midrange & vocals
pitch/tune +1

intelligibility/clarity +2

realism/timbre _2

embodiment +2

pitch/tune 0

clarity +2

timing 0/-1

sweetness +1

hifi attributes
soundstage stability/realism +1

soundstage width 0

soundstage depth 0

soundstage height +1

other essentials
fuzziness (lack of) +1

virtuality +2

micro-dynamics +1

macro-dynamics +1

totals plus 21 & minus zero

Turning to the Assemblage

The synergy between this SET300B and the ERaudio SpaceHarmonisers (which sadly no longer seem to be at the same URL, even though they'd been in the violin business for years before they started making audio platforms) makes any competitor start at a disadvantage. This prompts interesting debates about tuning feet and platforms; are they reducing noise thus elevating the signal to noise ratio or are they tuning noise like a filter? That wideband noise is constantly being generated is beyond doubt, from the acoustic noise in the room (from the loudspeakers plus ambient noise) ricocheting around the various structures in the room whose own resonances randomly emphasise or reduce particular frequencies. This acoustic onslaught strikes every component in the audio chain from the printed circuit boards, the parallel casework (of magnetic or non-magnetic materials), the active components (valves, transistors or ICs), the passive components, including capacitors that emulate condenser microphones and inductive resistors vibrating in the magnetic fields from other components Resonance Reproduction tuning platform was tried on top of a carefully space cluster of 4 Yamamoto MGB-1 magnetic bases. Only a slight lift of 2mm between the magnet faces was obtained with this very heavy combination and a slight lift in soundstage tangibility and scale resulted, compared with the same MGB-1 magnetic bases, similarly spaced, beneath the SET300B above the larger SpaceHarmoniser platform. I had originally established in the Yamamoto PB-22 review that PRaT was slightly better preserved with the ERaudio steel cones beneath the ERaudio SpaceHarmonisers, so I finally tried the following combination:

"Wha-at?" exclaim plebs, stage left,"The old scribe's gone too far this time!"

Indeed your old scribe has gone too far with two SpaceHarmonisers, four steel cones, four Yamamoto MGB-1's and four EAR vibration absorbing feet, as the difference is a negligible plus 3 and minus 1 between this and the condition tested in the table (pun alert). This is neither cost effective nor elegant and therefore not worth the effort. The tuned quality of the ERaudio sound is elevated by this mixture of cone decoupling and magnetic isolation, I suspect in part due to the ceramic inserts in the Yamamoto MGB-1 bases allowing the SpaceHarmonisers to do their thing. However, it just not worth the effort or money for 2 Spaceharmonisers, 4 ERaudio cones, 4 Yamamoto MGB-1's and the 4 EAR feet fitted as an expensive upgrade from the SET300B deep rubber original ferules. It merely serves to demonstrate how far it is possible to go while still making gains, however slight. Returning to a single more inert lower platform, from Something Solid in place of the lower ERaudio platform changes nothing consistently.

Way back in the annals of this vibration control Odyssey I quoted William Blake's idea that we cannot know how far we should go in any endeavour until we have gone too far and then return to the ideal position. The plebs chorus correctly identified when I'd gone too far and listening proved no advantage despite doubling the spend and pile of shelves.

[magnetic personalities?]


Yet more Yamamoto Sound Craft tuning devices; the word "footers" is both clumsy and inadequate to describe the complex functions of these multi-part isolation products that do what they claim. Whether they are worth the very high (by my real world standards) exceptionally high cost of 10k¥ for MGB-1 biggies that support 2-10kg per device, or 7k¥ for the smaller 0.9-3.0kg supporting MGB-2 depends entirely on the resolution of the system in question. Their capacity to wring extra nuances of performance from high resolution thermionic hardware, in combination with other tuning products throughout the system, is unmatched in my opinion. Compared with magic wire at extraordinary cost these magnetic suspension devices deliver much more bang for your buck, your €, or indeed for your Yen. Compared with the supposedly magical performance of exotic tuning devices manufactured on mountain tops by gurus formerly employed as unicorn farriers the Yamamoto MGB-1 and Yamamoto MGB-2 deliver genuine real world isolation via their magnetic circuits plus whatever benefits that a combination of ebony and hard ceramics is supposed to deliver.

If the, probably unresearched, 10% expenditure allocation to cables has any validity in outcome experience, a similar, if not greater, proportional allocation to vibration control must be in order. My extensive experience since the start of the vibration control and tuning series is that vibration control and tuning is often much more cost effective than cable tuning. Hence, high resolution devices like the modified Assemblage SET300B (suffering from its steel casework despite extensive damping measures) that cost approximately 3000€ (with couture passive bits and NOS and premium valves) do noticeably benefit from three, four or five MGB-1's at 82€ apiece (but 311€ for a set of four) represents almost the cable 10% rule but delivers much greater benefits. Given the price of the wires each side of one's power amplifier probably exceed 3 times this amount, the combination of a set of Yamamoto Sound Craft MGB-1 and a good plank (audiophile or well chosen wood) will make more difference. Lesser chassis with clangy casework are better served by blobby devices like the BrightStar Isonode or the Jade Absorbers and a Little Rock or a brick on top.

The Yamamoto MGB-1 and MGB-2 Magnetic Floating Bases produce dramatic differences that are noticeable with electro-mechanical devices like turntables and cd players. In value terms the Yamamoto MGB-1 and MGB-2 Magnetic Floating Bases justify themselves by their performance with high end cd players, unsuspended turntables in high resolution valve based systems.

Music enjoyed during this review

On Vinyl:

  • Stevie Wonder: Hotter Than July
  • Orbital: Are We Here?, 12" vinyl edition LIARX15
  • Orbital: Snivilisation 4 sides of glorious vinyl, after the decade that misic forgot 80s Orbital's Chime was the first new and imaginative music, an intelligent techno that announced a new decade of great music to come
  • Jimi Hendrix: Live at Woodstock, recent triple vinyl reissue box set
  • Neil Young: Vol 1-4, vinyl box set and a lot more I forgot to write down! On cd:
    • Zzebra: Zzebra/Panic
    • Arvo Pärt: Alina,
    • Yucca Flats: Garden of Weeds
    • Damage Control: RAW
    and all recent cd reviews...
    and on FM, BBC Radio 4 production of Othello broadcast 27th February 2010

© Copyright 2010 Mark Wheeler - mark@tnt-audio.com - www.tnt-audio.com

[ Home | Staff & Contacts | DIY & Tweaks | Listening tests | HiFi Playground | Music & Books ]