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

The Birth of Septum

The Septum Arm Breaks Cover - Part 3

Products: Septum Pick-up Arm
Manufacturer: Audiomeca - France
Cost, approx: Septum 5000 Euro
Reviewer: Geoff Husband - TNT France
Reviewed: March, 2007

[Septum arm]
[Italian version]

Introduction

Having read both Parts 1 and 2 of this saga I'm sure you are ready for the final part - and so here comes the tale of Septum*1. It deserves as much time as Belladonna because it is just as unique - there's nothing like it on the market. Once again every aspect of its design has been arrived at for entirely logical reasons so that in reading this article you may well arrive at a greater understanding of what goes on in a pick-up arm. Hopefully in my ramblings you'll be able to at least follow some of Pierre's arguments.

Long ago I made a rather controversial statement in one of my articles "Recently I was in conversation with a manufacturer of turntables and arms. He said that in his opinion, given a basic level of competence in the turntable, the arm was responsible for 80% of the sound quality produced. During my series of turntable and arm reviews I've started to come to see his point, though I'd put the split nearer 60/40 in favour of the arm. Whether you agree or not, it's not in dispute that the tonearm has a fundamental effect on the sound produced by a vinyl front end."

This is of course heresy.

In the dark days of the '70's Linn strode the planet and the world was taught that the most important component was the turntable, upgrades such as power supplies more important than arms and cartridges. Following this Linnies would progress to a full dress LP12/Lingo with a cheap Japanese OEM arm (the Basic) before shelling out on the ITTOK. Perhaps this was the hierarchy that suited the Linn, but it became universal. So prevalent was this view that to this day you will rarely see reviews of high-end arms on 'budget' turntables. In all this time I've seen just one such review, of the new SME 5 (this in the 80's) on a Systemdek. The conclusion? That the combination was wonderful and it obviously showed the Systemdek to be much better than the writer imagined...

The aforementioned manufacturer was Pierre Lurne and my experience seems to back up his view. It's interesting that a manufacturer trying to convince people to shell out 15000 Euro on a turntable admits that the 5000 Euro arm that he also makes is the more important component.

Is this logical? Well at a simplistic level the turntable is most responsible for the low frequency quality of a turntable, the arm has some influence in the bass, then becomes increasingly important with it's "signature" going up to 2 kHz after which the cartridge is pretty much operating on its own. OK it's a sweeping generalisation but even if there's only a hint of truth there, the midband, where after all most of the music is, is dominated by the arm. Thus it's the arm that holds the key to music.

Theory and Design

Some of what follows has been covered in my review of the Romance/Romeo, my interview with Pierre Lurne and the article on arm design so please excuse any repetition.

Pure Mass and the Ellipsoid of Inertia

As with Belladonna the 'Pure Mass' principle is absolutely fundamental to the design. I've already partially explained this in Part 2, but here I'm just going to copy and paste what Pierre told me, all of five years ago.

"... a good turntable and arm should be neutral, have no personal character, no overt behaviour, to be "dead". The concept is defined in Physics as the "Pure Mass" concept. A pendulum for example, is not a pure mass by definition as it has its own resonance frequency - so many arms and turntables work as pendulums...

To know if a given body is a pure mass, and by the way, in order to know anything about its dynamic behaviour, one must calculate the "Ellipsoid of Inertia". Everything has its Ellipsoid of Inertia, a chain, a car, a pencil. My dear Geoff, I am pleased to inform you that even you have an ellipsoid of Inertia... What is it? To simplify - suppose you want to know the ellipsoid of a book. OK - start by taking a line or axis through the book's centre of gravity - any line, any direction but through the centre of gravity. Then you calculate the moment of Inertia of the book as regard to this axis. This will give you two values + and - either side of the axis - thus you have two points plotted. Now draw another axis through the centre of gravity and plot two new points. Keep doing this over and over again, or speed up thanks to integral calculus, and you will end up with a mass of points forming an ellipsoid, sort of egg shape - this will always be the result! More or less long, more or less flat but always that egg shape - an ellipsoid... That's physics, that's nature... And that egg is the mathematical representation of everything regarding the dynamic component of the given body. You will see straight away that with a very regular body, the egg will have a special form - a sphere called "the Central Ellipsoid of Inertia"...

[Ellipsoid of inertia]

OK so if we apply this test to a tonearm we'll get an ellipsoid around the arm's centre of gravity - approximately around the pivot of the arm. Now project the ellipsoid forward to the stylus so an ellipse surrounds the stylus and you'll have a representation of all the forces acting on the stylus, and by the way, to the arm cartridge combination. See diagram.

The perfect arm would have the ellipse as a circle with the stylus tip in its centre. That means that the stylus will move equally easily up, down and left/right. Neutrality is complete.

If the ellipse is flat then the system is not neutral, it is easier for the stylus to go up and down than left to right. In this case the combination of the arm's effective mass and cartridge compliance will give various frequencies of resonance rather than one which is easier to control. This is the design weakness of all air-bearing parallel-tracking arms, the resistance to up and down movement will be much like a bearing tonearm, but side to side will be much more difficult as the totality of the arm must move - around 20 times more difficult than vertical movement. Depending on the eccentricity of the record (no record is perfectly centred) the arm movement can take any direction and speed each time adding a different resonance. At best such arms can be symmetrical, at worse just crazy - by that I mean that to have an ellipse so distorted with the stylus way off centre - it means that all movements are different, adding different resonances, different behaviours etc.

The question is, would you prefer a well designed arm with a round ellipsoid which moves equally easily in each direction, or one that prefers to go right than left and up than down with a crazy ellipsoid?

Just looking at an arm will tell you whether an arm is correct or not: If the arm is straight with the stylus well along the main central axis and a simple counterweight on the same axis this looks about right. If it has a curved armtube, a low centre of gravity, a low counterweight or a mass of parts then it is not correctly designed. A word of caution - the respect of these rules doesn't guarantee a good arm, it's still possible to make mistakes elsewhere (flexible armtube, poor joins etc) but it is certain that a good arm could be even better if it respected the basic laws of physics.

There are also secondary advantages to such a perfect system - if a parasitic force is applied to it (anti-skate design, lead out wire drag, any kind of vibrations coming from outside) then you cannot ignore it or get rid of it, but you can minimise the effect because it is applied to the system on its own, rather than working through another force (couple) caused by the poor ellipsoid - so you get a minimum value."

So there you have it from Pierre. It follows that many arms, low slung unipivots, strange shapes etc are going to break this "Golden Rule". Remember that this was written before Septum was a twinkle in Pierre's eye - now we have the ultimate expression of his ideas. Look at the pictures and you'll see clearly, that apart from the inevitable angled cartridge mount, the arm is totally symmetrical.

K.I.S.S - Armtubes, Bearings and Vibration Control.

Having read Part 2 of this saga you should know quite a lot on the control of vibrations, no-where is this more important than in the arm. The design of Septum is obviously a typical Pure Mass as described above, but it also takes the KISS principle (Keep It Simple Stupid) to it's logical conclusion. If we accept that every joint, every component that goes to make up an arm has the potential to resonate and to reflect vibrations, then all things being equal, the lower the parts count and the simpler the design the better the sound quality. At the heart of this principle - and the best example of its application - is the armtube.

Look at the pictures again and you'll see a unique armtube, it's made of aluminium, and the entire thing, from cartridge to bearing, is a single unbroken piece of aluminium. It is also tapered, the angle of the taper increasing towards the bearing housing - this forming a shallow "horn". This is the single most expensive and difficult to manufacture part of any arm. Look at all the arms on the market and very, very few use a one-piece armtube, no matter how expensive they are. It is just so much cheaper and easier to simply use a tube and attach it to the bearing housing at one end and the headshell at the other. There is however one huge exception and that is the family of Rega arms - how do they do it?

There are several ways of making a tapered one-piece armtube, but one has the potential to be cheap and that is pressure casting. There is however a major snag, and that is that setting up such fabrication is very expensive, but once moulds are made the tubes can be produced relatively cheaply. Rega took a big gamble with the RB300 investing a great deal of money in the process, but they did have a popular turntable that needed an arm and so the economies of scale added up. That the arm has been in continuous production for 20 years, and with I guess tens of thousands produced, that risk has been handsomely rewarded. SME chose the same route and again their high volumes made it economic. For all other manufacturers the sums just don't add up and so most glue or screw armtubes to headshells to bearing housings.

You can produce a one-piece armtube by taking a tube and cutting/flattening the tube to form a headshell, Alphason did this first with their titanium HR100, and it looks like Roksan do the same. But in both these cases the tube cannot be tapered and the thickness of the tube wall is constant. Lastly an armtube can be fabricated from a single sheet, cut into a kite shape then rolled and welded - this can produce a taper, but again wall thickness is constant and the 'join' may have detrimental effects.

The armtube of Septum is made in the most difficult and expensive way possible, but it is arguably the best. It is machined from a solid-aluminium, drawn bar. This involves many machining operations but produces an armtube where there is absolute control of thickness, taper and a very fine finish. Because it is a drawn bar the crystal structure of the arm is aligned along the length of the tube and this does give a preferred direction of travel for sound/vibrational energy. This crystal structure can also be more uniform than in a cast armtube. To my knowledge only the long-gone Mission Mechanic used such a tube.

[Armtube cutaway]

OK so the armtube is very flashy and expensive, but it's important to understand why that expense is justified. That sophisticated shape with no joints is very stiff, far more so than a standard tube and though I'm only spending one line on the subject it is important! The taper also reduces standing waves, spreading them in a way that a parallel sided tube cannot, this is just one way in which the tube controls vibrations, but there are more (and they're a lot more difficult to explain!).

You're going to have to close your eyes at this point and do some imagining (read the article first!).

Imagine you are standing in the doorway of a small, square room. In your hand you hold one of those super-bouncing balls. Now the tricky bit - for the analogy to work you need to imagine that there is zero gravity...

You take the ball and you hurl it at the far wall. What happens? It probably bounces straight back and smacks you in the mouth:-) Now throw another and another, each time the ball is quite likely to do you serious injury. Try throwing the ball at the side wall - a couple of bounces and that ball comes right at you again.

All very entertaining, but this is a crude demonstration of what happens in the Arm. You are the cartridge, picking up vibrations from the disc (those not converted into electrical energy and passed to the amp) and the balls represent that vibrational energy going into the headshell. The balls that return to cause you damage are the equivalent of the vibrations which bounce around the headshell and come back to hit the cartridge, vibrate it and in turn pass those vibrations as electrical energy (noise) to the amp. Some energy will pass from the headshell into the armtube, but much of it, (how much depends on the armtube material*2), will remain bouncing around in the headshell.

These are very bad vibrations, the worse kind of noise. Because they are music related, and come very quickly after the initial reading of the disc, they blur the sound, rounding leading edges and reducing dynamics. They contrast with record surface noise, which is not music related and so, within limits, doesn't spoil the music.

If you look at the diagram you'll see what happens in a solid block when sound energy is introduced*3. At each boundary the sound is reflected. If you want another analogy, imagine throwing a stone into a small pond, the ripples widen until they hit the bank then return.

Now go back to the doorway in your gravity-free simulation - see fig 2. Imagine that the room is much, much longer, 20 times as long as wide. Now hurl those balls into the room. Yes they will come back to you, but after a lot more time and with diminished force. This is what happens in a one-piece armtube, and remember that as the energy travels through the aluminium it will be dissipated as heat in the aluminium itself. The longer the vibration is actually in the armtube the more vibrations are attenuated in this way. This is the case for a one-piece armtube.

Now lets go for the last time to that doorway. In front of you is that long room, but now the room becomes wider and wider down its length - fig 3. Now the vibrations not only travel much further to the back wall, but also there is much more material for them to bounce around in and dissipate.

[Vibration pathways]

Septum though has a taper shaped as a horn and so the amount of vibration trapped, or which undergoes many more bounces and spends more time in the armn away from the headshell is hugely increased see Fig 4 (obviously exagerrated).

[Septum bounce]
In Septum this taper extents to the thickness of the armtube as well, with the walls being far thicker near the pivot. This giving more space for the vibrations to be converted to heat. In a conventional tube with uniform thickness the wall must be thin down its length otherwise the effective mass of the arm will be too high. With the Lurne taper the mass is concentrated around the pivot where it will do least damage. The "gilding of the lily" comes last - Pierre has anodized the surface, providing an almost diamond hard finish where high frequencies can travel easily (they stay near the surface like "skin effect" in cables), and of course the tapered armtube gives a much bigger surface for those vibrations to travel in...

All very clever, but inevitably some energy will reach the bearing end of the armtube and it's important that as much of this as possible is passed out of the armtube and not reflected back towards the cartridge. How much is passed on is dependent on the material on the other side of the join and in Septums case this material is that magic material - Lead *4. Sitting at the end of the armtube is a disc of the stuff. Because it is near the pivot axis it adds very little effective mass and yet it means that as much energy as possible is removed from the armtube. It also of course does the same for vibrations coming the other way, and so acts as a kind of isolator for the arm.

The armtube also forms the bearing carrier and is then bolted directly to the counterweight stub, which carries a symmetrical counterweight made of...Lead (cleverly disguised). The counterweight is actually held onto the stub by a single point. The idea of this is that it reduces one weakness of traditional arms. If you attach a big lump of brass on the end of an armtube the result is that most energy from the armtube will just bounce straight back. The solution of "decoupling" the counterweight is flawed because by putting a lossy rubber or similar layer between the weight and arm we introduce another resonance as the arm can bounce around relative to the weight. On Septum that Lead counterweight allows the maximum amount of vibration to pass easily into it where they are transformed to heat.

Unipivot

Septum is a unipivot, attached directly to that armtube. Knowing what we already do about the way vibrations bounce around at every boundary it's obvious that for Pierre the idea of a mass of ball bearings, each presenting boundaries is simply not the way to make an arm. Add the difficulty of making the bearing tight enough to reduce bearing "chatter" (rattling) and yet free enough in all planes and the unipivot, when properly done has much to recommend it. Vibrations still left in the arm have one simple path down through the unipivot and in the case of Belladonna down through a dedicated spike. Oh and at the base you'll find another disc of Lead sitting between the arm base and arm tube...

[Pivot detail]

With Septum, Pierre has the unipivot optimised to the very limit of his ideas, with the one-piece tapered armtube, symmetrical Ellipsoid of Inertia and Pure Mass.

Details

There are some audiophiles who have taken the KISS principle to the limit with their arms, sawing off as many extraneous parts as possible to lower the parts count - they report improvements each time, this may be illusional, but with the Septum this isn't necessary because the arm can be stripped down to its essentials very easily. The VTA ajuster allows the arm to be simply moved up and down to find the optimum height by turning an adjuster. It's an excellent system, but when the arm has been set the whole VTA apparatus can be taken off in seconds leaving the arm mount as simple as possible. Likewise the anti-skate, finger lift and cueing device can be removed. The latter is made entirely of Metacrylic in order to reduce the ringing found with metal lifts, and it is very pleasant to use.

The anti-skate thread is attached exactly level with both the pivot and the Centre of Gravity. That means that it pulls only on that plane. Contrast this with most such threads, which are attached at the top of the arm tube. In these cases the anti-skate will be trying to twist the arm as it pulls. On a unipivot with a very low C of G, or on a gimballed arm this effect will be masked, but it will still be there, especially on eccentric records (all of them...). Septum has to be designed right otherwise the error would show up plainly.

Similarly the arm lead-out is always a problem, as it will inevitably offer resistance or bias to the arm as it is bent as the arm moves. In Septum that lead out is below the pivot and as close to it as possible so that the effect is minimised. Other arms may well have problems with such lead out wires, but because they will return to a preferred position this effect is masked.

Azimuth is adjusted by simply turning a socket screw to move a weight inside the balance bar - very precise and foolproof, and cartridge alignment is just provided by slotted screws. This may seem slightly crude compared to the SME 5's sled ajustment of overhang, but the SME assumes standard geometry and perfect alignment of the stylus/cantilever, both big assumptions.

[Septum top]

In Use

First off this was a prototype and Pierre was very apologetic about its fit and finish. As far as I can see both are excellent as they are, and the production model, if better, will be a real Rolls-Royce job.

Now the surprise. It doesn't feel like a unipivot in use. It doesn't wobble or feel unstable, simply because with its Pure Mass it just moves where you want it without rocking about. Most unipivots give the feeling of being very "free", if you balance them at zero downforce they appear to float beautifully, but in fact they feel that way because they always want to float to a particular position. The Septum felt much more like a gimballed arm being very stable.

The various adjustments, VTA , azimuth, anti-skate and downforce are simple, in fact as easy as any arm I've tried, certainly those put off by fiddly unipivots have nothing to fear here.

Compatibility

Septum is a substantial arm but by no means a heavyweight. It's designed to work well with most quality cartridges, but more importantly it will fit just about any turntable - some recent designs are so heavy that they upset many suspended decks, or are just too large to go under dust covers and the like.

Conclusion

Septum is an original, there's nothing quite like it on the market, and yet it comes with a design history going back over 40 years to Pierre's first "commercial" arm. But though unique it is widely compatible and easy to use. As for sound I've already explained why I cannot give a listening review, but in one sentence: I found the prototype to be big, bold and dynamic yet capable of subtle detail.

I'd just like to thank Pierre Lurne for his infinite patience with me for the last couple of months in opening the secrets of his new babies. And on the subject of secrets, one could look upon the Belladonna/Septum as being similar to an open-source computer project. Pierre's primary concern during these articles, especially those concerning Belladonna, was to put his ideas to a wider public, all the details of the designs are here - if you want to build a top DIY turntable, or if another manufacturer wants to use the same principles then that's fine by Pierre - he's a man dedicated to improving analogue reproduction first and foremost. That said he's also insistant that these are not his 'inventions', merely the siple application of high-school physics.

*1 - In case you're wondering, this is the seventh ("sept" - seven in French) commercial design from Pierre's drawing board.

*2 - To enable the text to flow I've rather glossed over this aspect, but it is hugely important. All materials have an "acoustic impedance", a function of their internal speed of sound, their density and elasticity. If two objects of identical acoustic impedance are joined then in theory 100% of vibrations will pass between them, so it would seem to be that an Aluminium headshell would pass on all vibration to an Aluminium armtube that it was bolted to. However ANY gap and 100% of the vibration will be reflected back to the headshell. Of course at a microscopic level there will be many such gaps as no headshell is a perfect match and so most vibrations are again reflected. Filling the gaps with a glue for example will help, but because the glue will have a very different acoustic impedance there will still be a lot of reflection as the vibrations cross into the glue and then onto the armtube. However if we follow this argument it's obvious that for the point of view of passing vibrations from the headshell, it is important that the headshell and armtube should be made of the same material - an Aluminium headshell on a Titanium tube, or carbon fibre tube will reflect more than Aluminium/Aluminium. Of course Septum's armtube simply avoids all these problems. It's also interesting to consider the effect of cartridge bodies - in theory a cartridge with an aluminium body will work best on an aluminium headshell, a wooden cartridge best on a wooden shell and so on. Lastly, Lead is a special case as because it is so soft, when clamped to another metal it will fill voids itself without any bridging "filler". For more info on this topic go see this excellent page

*3 - Before I get letters - I know that sound doesn't travel like a laser beam, but in wave fronts, but I can't draw them and the effect is the same:-)

*4 - This is very controversial. We all know that because of RoHS lead is now banned in electronic components. Turntables are a borderline case and until someone is taken to court their legality is questionable. Currently the majority of top turntables have to use lead - it's a unique material, hugely superior to any other metal for energy absorption, for example the speed of sound in lead is under a 1/3 of that in any other metal. Add its density and it's irreplaceable, just remember that you can make a very reasonable bell from any solid metal except lead. Personally I feel that we may now be seeing the best turntables we will ever see because when engineered without lead they will be hamstrung. The same applies to valve amps which can no longer be sold with NOS (New Old Stock, like Western Electric 300b's) valves.

Copyright 2007 Geoff Husband - www.tnt-audio.com

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