What the 'professionals' think makes a great speaker ~~~

[DSRANCE]

It has already been shown that the metal and paper cones have fundamental resonance problems. The metal cone high-frequency resonance peak is a real problem, Derek Hughes rejected the idea too, which he mentioned in an earlier post on this thread. Alan has also discussed the short-comings of the dome midrange unit you mention, elsewhere on the HUG.

If these materials could give the required results, I would suggest Harbeth would use them now rather than investing in the RADIAL research project, because it would be much cheaper and less time consuming than developing their own cone material.

Greg, I think that the proof of the pudding is there for all to hear, if they get the chance, as it's the very fact of the smoothness and sweetness, with no loss of treble "details," that separate current Harbeths from the "HiFi" competition.

I know I mentioned that I thought modern drive units to be better than most commercial models of a few decades ago, but designers of inexpensive speakers today have a huge problem in designing something that has to look good yet still have a "boppy" bass and a sparkly treble. I fully appreciate that this is nothing to do with the likes of Harbeth, but many commentators I respect have said that the smaller "meter-bridge" pro models do tend to boom and screech a little and I suspect that what is in this thread may have more than a little to do with it, not just under-damping the bass to make it "port-bound" and setting the tweeter too high in level.

Do ALL conventional drive units have resonance issues, or is it more a fact of consistency in manufacture?

[Gan CK]

I guess if one is not or has not been exposed to Harbeth's RADIAL, then i reckon conventional drive units aren't so bad after all. Since my exposure to Harbeth, even the old HL-mk4 with the TPX cone was able to let me immediately discern cone colorations from other spks. With RADIAL, i feel that this is even more pronounced today despite the fact that material science has supposedly improved considerably. Look at all the boutique spks that we have today. Alot of these spks use so called 'ultra high tech drivers' made of Ceramic, Aluminium, Magnesium, Diamond, Beryllium & whatever sandwich material but ironically, these spks often derail the listener away from the reality of music.

With Radial, i feel that Harbeth has really set the benchmark in what a hifi loudspeaker can achieve with regards to tonal accuracy/purity & musicality, whilst at the same time not sacrificing transparency, definition & musical involvement.

[A.S.]

... I reckon conventional drive units aren't so bad after all. Since my exposure to Harbeth, even the old HL-mk4 with the TPX cone was able to let me immediately discern cone colorations from other spks...

TPX plastic is by no means a conventional cone material. Why Audax couldn't promote it - or more accurately, why the speak industry showed total disinterest in changing from the status quo (polypropylene, paper and later kevlar) is, I believe, the greatest oversight the audio industry has ever made. TPX has (or had, when we looked at it's technicalities some twenty years ago) the lowest mass of any polymer, and hearing familiar music on a pair of Mk4s that I borrowed from Harbeth's founder Dudley Harwood changed my life. The shock of partially lifting the coloration veil had such a profound effect on me (I can easily recall the experience in about 1986) that I quit a very well paid job in the semiconductor industry to pursue the subject of speaker cone development via owning this company.

The conclusion must be that apathy is very much alive and well (in audio manufacturing and consumerism) and why bother tackling real engineering issues such as cone development, when you can tart up and flog the conventional as the new.

[Gan CK]

Yes Alan, the MK4 also had a very deep impact on me when i heard it the first time. I had a couple of references at that time from KEF 105, Gale 402, Rogers Studio 1 & probably Spendor SP-1 but the MK 4 at that time was to me a revelation in terms of tonal neutrality & musicality.

However, i discovered another revelation when i switched from the old HL-5 (with TPX) to SHL-5. That's the Radial revelation!

[EricW]

The conclusion must be that apathy is very much alive and well (in audio manufacturing and consumerism) and why bother tackling real engineering issues such as cone development, when you can tart up and flog the conventional as the new.

And yet presumably most manufacturers realize that cone material is not irrelevant (to say the least): I see all kind of reference to cone materials in marketing literature - composites, metals of all types, doped paper, mica-filled polypropylene, kevlar, doped kevlar, and on and on. It can't all be marketing, surely? Presumably manufacturers are chasing some kind of performance advantage, but apparently without - as you say - doing the fundamental work of developing the best possible material. Very curious indeed.

Can I ask what may be a hard question? Within the limits of (known) materials science, how much room is there for a "better" material than RADIAL? Put differently, if you were to conceptualize the ideal cone material (capable of existing in the universe as we know it, i.e. must have mass etc.), what is the gap between RADIAL and the ideal?

{Moderator's comment: Interesting question. What would you think was the primary requirement or two?}

[EricW]

What would you think was the primary requirement or two?}

You will reach the limits of my knowledge very quickly, but I would guess (not necessarily in order):

1. Relatively low mass
2. High rigidity
3. High internal damping

I would image other materials might give you two out of three (e.g. 1 and 2 in the case of metal), but getting the right combination of all three would be extremely tough. And maybe there are others I haven't thought of.

[A.S.]

The real killer problem is 1), low mass. As this tread is about efficiency, you can imagine that every gram of mass you can shave raises the sonic sensitivity by about 1dB or so (for a typical 5-8 inch unit). So, to get from say 85dB to a nice marketable 90dB, if you lose 5g or weight, you've achieved that. Conversely, add a gram or two and that tolerable 85dB drops to 83 or so. A noticeable reduction.

Furthermore, your 2) and 3) imply rubbery-like damped materials which are, by definition .... heavy. So you see we are in a real engineering trap here. We want it all, we want it now, but it must be a light as a feather! The only way around this is some lateral thinking.

[Labarum]

Furthermore, your 2) and 3) imply rubbery-like damped materials which are, by definition .... heavy. So you see we are in a real engineering trap here. We want it all, we want it now, but it must be a light as a feather! The only way around this is some lateral thinking.

In 1957 Peter Walker of Quad offered his lateral thinking in the Quad full range electrostatic loudspeaker - a thin plastic membrane (clingfilm) stretched tight in an electrostatic field. That original electrostatic will still give the best modern speakers a run for their money.

There is also the Magneplanar loudspeaker with a membrane stretched in a magnetic field maintained my a distribution of many small permanent magnets. The membrane has a conductor running through it which reacts with the magnetic field to displace the membrane. This is in effect a full range ribbon speaker.

Both these systems reduce the mass of the radiator, but there are other engineering trade-offs.

{Moderator's comment: Low mass, yes, but also low damping. See the problem?}

[A.S.]

I would much like to develop the discussion of the RADIAL technology but it really is far more complex than just peering at a few graphs. It's a sort of 3D engineering problem which (here) can only be viewed in 2D.

I'm still not finished PDFing the RADIAL project. I've tickled this documentation for over 18 months but the sheer volume of it and the fear of accidentally discarding some vital pages (which I've done several times and managed to Undelete from the server) means it needs full concentration, and certainly can't be subcontracted.

Over the next months I'll try and find a way of exposing more without - and I'm very aware of this - giving away anything that could advantage others in this field. It's always far easier (and cheaper) to take-on someone else's brainchild and develop it than do some genuine blue-sky research yourself!

Only last week I rediscovered the meeting minutes from the Govt. committee overseeing the RADIAL project (because it was part funded by a Govt. grant) at which I expressed my "disgust" at the slow technical progress. Fortunately, the senior member agreed with me. The path to RADIAL was not direct - there were many almost insurmountable technical hurdles.

[HUG-1]

And here is something from the archive. Announced in 1993, just as we were in the final run of the RADIAL cone research project.

[EricW]

...we are in a real engineering trap here. We want it all, we want it now, but it must be a light as a feather! The only way around this is some lateral thinking.

I have one thought - possibly lateral, possibly not - though it doesn't relate to cone material as such.

But when it comes to mass, I imagine the significance of mass is not absolute, but that what counts is the amount of mass that needs to be driven in relation to the amount of power available to drive it. So, for example, a top of the line Mercedes Benz will be relatively fast despite being relatively heavy, because of the amount of power available from the engine.

Would not the same be true of a speaker drive unit?

Right track or dead end?

[Gan CK]

Despite the widespread appearance of so many 'exotic' materials on loudspeakers over the past few decades, i feel that only Harbeth's Radial is a genuine technological breakthrough & a major milestone in pushing the boundaries of loudspeaker design. Its really very very difficult to pin down the sonic characteristic of Radial because it really doesn't seem to have one. Maybe its still not perfect but at this moment, i feel that its head & shoulders above all others.

[Paul G Smith]

I have one thought - possibly lateral, possibly not - though it doesn't relate to cone material as such.

But when it comes to mass, I imagine the significance of mass is not absolute, but that what counts is the amount of mass that needs to be driven in relation to the amount of power available to drive it. So, for example, a top of the line Mercedes Benz will be relatively fast despite being relatively heavy, because of the amount of power available from the engine.

Would not the same be true of a speaker drive unit?

Right track or dead end?

What about the brakes and shock absorbers and the tyre wear A powerful engine needs equally powerful brakes.

Imagine then, a speaker cone which had an extremely low mass, approaching the mass of the air itself, needing therefore very little power to accelerate, but equally importantly requiring very little braking power to stop it. The air surrounding it would absorb the kinetic energy of it's movement and so it would therefore be more likely to follow the instructions given it by the driving force, taking very little time to accelerate and decelerate.

Colouration would be near zero and we would all buy one!


This was the theory behind Quad's electrostatics and the Fane Ionophone but these solutions have their own problems and limitations.

[honmanm]

But when it comes to mass, I imagine the significance of mass is not absolute, but that what counts is the amount of mass that needs to be driven in relation to the amount of power available to drive it. Would not the same be true of a speaker drive unit?

Probably not the car analogy because at its heart a speaker drive unit is a transducer or servo motor - converting one form of power to another. Maybe tyres would be the automotive analogy? An amplifier that is able to source or sink infinite current should be able to hold a driver in precisely the right position at all times (that is a bit of an oversimplification for other reasons).

Real world amplifiers will fall short of that ideal, and real-world speakers too. That's where the fun begins - finding those combinations where the components do not interact to the detriment of the sound quality.

On a different note, there has been some mention of panel speakers - Maggies and Quads - on this thread. I suspect that the Quad membrane is light enough that the air itself provides sufficient damping - though I do wonder how the surface wave in the membrane is damped at the boundaries where it is bonded to the frame... I'd love to understand the physics.

In the case of the Maggies although the membrane is light and stiff the "midbass" wires are comparatively heavy and if you tap the back of a Magneplanar speaker you'll probably hear the panel resonate. But if you look at a "naked" Maggie there is a carefully positioned rivet that pins the membrane to the magnet assembly, towards the middle of the membrane but clearly not centralised. At a guess it breaks the resonance of the panel into a number of lesser & less objectionable modes (analogous to different sized panels on a box speaker?). One of the black arts at the Magneplanar factory is the tensioning of the panels, I've heard...

I'm pretty sure the designers of dipole-radiating speakers have a relatively easy job of controlling resonances in the structure of their products, maybe this outweighs the imperfections of the panel itself.

[A.S.]

...I'm pretty sure the designers of dipole-radiating speakers have a relatively easy job of controlling resonances in the structure of their products, maybe this outweighs the imperfections of the panel itself.

By which you mean, I think, that as the supporting structure of frame of a panel speaker is unlike a box (the box traps sound inside, the frame is completely open) the contribution of the box to the overall sound can be bypassed.

Interestingly, a few years ago a highly respected speaker manufacturer with excellent technical resources invented a box cabinet damping system which substantially reduced the contribution of the cabinet to the overall sound. Test measurements confirmed this remarkable achievement. But to my ear, the overall sound was consequently rather bass shy and devoid of any warmth - uninvolving is a word. Carefully tuned, the cabinet can make a very useful contribution to enriching the sound and pulling the listener into the music as a Stradivarius does and a cheap mass produced violin rigidly glued doesn't. Don't underestimate the importance of a well tuned box!

[Gan CK]

Yes the contribution of the enclosure must not be overlooked.

I've heard not so subtle differences between poorly made & well made acoustic guitars & violins. The better ones always sound richer in tone just like Harbeths always sound richer than many lesser but much more expensive spks. However, to a lot of ignorant audiophiles, they think that the BBC thin wall cabinet is simply a marketing ploy to cut corners & manufacturing cost. What a joke!

[A.S.]

...However, to a lot of ignorant audiophiles, they think that the BBC thin wall cabinet is simply a marketing ploy to cut corners & manufacturing cost. What a joke!

Surely you are pulling our legs?!

I can tell you from bitter experience that not one cabinet maker in ten would be remotely interested in tackling the complexity of a 'BBC thin wall cabinet'. I know this because production here has a constant brief to cover the eventuality that our cabinet suppliers suddenly cease and leave us high and dry. That leaves me with the task of explaining to a prospective supplier what the thin-wall concept is, what the likely skill-set and precision tooling the cabinet maker will need and an indication of the hidden difficulties and complexities. Most walk away at that point as an impossible specification with ridiculous tolerances at an unprofitable price. The occasional one will go so far as to make a sample of two and then, fully appreciating the magnitude of the task and our QC requirements, intentionally price themselves out of contention.

The issue is that the Harbeth front and/or back panels are removable, and are made at a different time on a different machine to the main top/side/bottom/side wrap; they are not naturally tongue-and-grooved into the cabinet as it trundles along the production line as all 'modern' cabinets are. Hence, assembled without the front/back as an open shell, great skill is required by humans to make sure the cabinet dries truly square etc. etc. etc.. There are so many internal and constructional factors that have to be carefully controlled piece by piece, let alone the parts you see - the veneer. Bringing on a new cabinet supplier and having them up to speed at the right QC level, all issues tied down, needs a minimum of about two years. That's been our experiece.

I admire and respect anyone able to take on the challenges of the 'BBC thin-wall cabinet' and to deliver in bulk to a great technical and cosmetic standard at a price we and you can afford. It is nothing short of a woodworking nightmare. But is does sound great.

[Gan CK]

No i am not kidding Alan. I have received comments via emails & personal messages from people with regards to the BBC or rather Harbeth thin wall cabinet. In today's context where many solid, fanciful looking & supposedly inert loudspeaker cabinets exist, many simply fail to see the merits of the thin wall cabinet. The worst was when a guy said it was a marketing ploy. Thank god we the Harbeth users who frequent the HUG here know better. Some time ago I posed this question to you on why a Harbeth thin wall cabinet can sound so unboxy & uncoloured in the critical midband. And i recall you used a cracked bell as an illustration on how the BBC lossy cabinet works & that has been deeply rooted in my mind since then. So much so that i have also used the same illustration to enlighten others. I've always been fascinated with the BBC thin wall lossy box & how it works to deliver a thoroughly musical, warm & involving sound, whilst not forgetting the role of the excellent Harbeth Radial driver & the complex crossover to seamlessly integrate the drivers.

[Labarum]

A lot of furniture made in the austerity years following World War 2 used "thin wall" construction. I have some of it and am amazed at the carpentry skill displayed in this furniture. A thin plywood skin was used with hidden bracing - all to construct strong and durable products using the minimum amount of wood which was in very short supply. Once particle boards became available the style of construction changed and the skill used in its construction diminished.

I guess the Harbeth thin wall cabinets use some of the techniques then common in any furniture factory.

[Gan CK]

I came across a French loudspeaker that also use thin wall cabinets but quite different from those that Harbeth uses. Their cabinets are larger, front & back panels are not removable & i suppose the internal walls are not damped with bitumen. A tap on the cabinet told me so. And it uses high efficiency full range drivers made from paper.

The resulting sound? Yes it was quite dynamic with tight bass but the midband is extremely peaky & coloured. Unbearable for me honestly. A far cry from Harbeth's tonally accurate & sweet midband.