Quite possibly my last speakers

[simon]

I suppose I'm looking at it with the perspective of a Structural Engineer. Vibrations in panels are a function of length (span) and panel thickness(depth), as well as material properties, and applied loads of course. I think spray foam would go some way towards a spaceframe.

Footbridges for instance are often made in a 2D truss style - the open nature of the truss allows a much increased second moment of area (broadly analageous to strength) without using a lot of material that a solid beam would require. This has the benefit of weight and economy, but with bridge design you can chase your tail - the more material you use the heavier the bridge and thus the more material you need to make it strong enough to carry the addtional weight. And the larger the section the stiffer it is and therefore will attract more load... Provided the global analysis of the truss is adequate then the shorter length of each truss member is beneficial as vibration will tend to be less. There are a lot of generalisations here though as the overall design is a balance of many considerations.

But using this as a 2D example and considering a 3D version, a spaceframe, the expanding foam would form a "honeycomb", a very lightweight spaceframe, hopefully with largely equal length "members" (if the foam expands evenly as you mentioned).

So my thinking, which may be completely erroneous, is that instead of using a sandwich of say 3 x 18mm sheet materials (which seems to be regarded as good for reducing vibrations (a much reduced span÷depth ratio) but very heavy) a couple of say 12mm skins, or even 6mm, with 24mm thick spacer blocks glued between the skins to keep them equidistant should be more rigid but very much lighter.

I'm guessing that lighter materials would store less energy, but maybe that's not the right way of looking at it.

Anyway I've digressed a long way from your speaker thread, apologies.

[Max N]

Rohacell sheets might be worth a try. Possibly easier (less messy) than expanding foam, and more consistent. But one of its attractions is low mass, which may not be important in a speaker?

[IslandPink]

I'm not sure, Simon. There is the problem that a panel material has to work across the full frequency range, if not 10 octaves, then maybe the panel is active over say 7 or 8 . It may be that a deeper composite structure is more rigid at 50 or 100Hz, but if it absorbs and re-radiates sound at 2kHz, it could colour ( or more likely de-colour) the sound in the low treble.
I am thinking of the initial attempt i made with a lower-mid horn using two layers of bendy MDF plus polyester-based filler. It really didn't sound anything like as good as the later one made from pine blocks.

Probably worth a small-scale test before committing to anything like that.

[simon]

No I'm not sure either Mark. I suspect there's a good reason it's not commonplace. Structurally I think there's some merit, but acoustics isn't exactly the same as structural strength. I'm curious enough to try a small baffle some time maybe.

Max - interesting stuff, I haven't come across it before. It'd be interesting to get hold of some for the "handling" test to better understand its properties.

[Max N]

I haven’t done any composite design for a long time, but from what I remember…
The typical use would be: skin - film adhesive - core (rohacell or aluminium honeycomb) - film adhesive - skin
Vacuum bag to consolidate and into the autoclave for curing.

The main thing is that the core must do two things
1. Resist any shear between the skins
2. Maintain uniform separation between the skins.

But all of the above is to maximise strength and stiffness of the structure, with little consideration of damping, so maybe not directly applicable

[chris661]

It may be that a deeper composite structure is more rigid at 50 or 100Hz, but if it absorbs and re-radiates sound at 2kHz, it could colour ( or more likely de-colour) the sound in the low treble.

While true, the mechanical energy at 2kHz is much smaller than at 100Hz. ie, panels are far less likely to be activated.

4x the excursion is required for every octave more of LF extension.

Chris

[simon]

Thinking about this some more (and I probably shouldn't be wasting my time!) I wonder if a sandwich of two outer skins plus a middle layer, all say 5.5mm ply (as that what it seems to be these days) with appropriate spacers, perhaps 5.5mm between layers, might be worth trying some time.

This would give an overall section depth of around 28mm and should be really quite rigid and strong but also relatively lightweight.

The size and location of the spacer blocks would be important I think, and the middle layer would have plenty of holes in so it more resembles a spaceframe. This would reduce weight a touch (maybe a good thing) and allow the expanding foam to flow between the two voids.

My gut feeling is a holey middle layer would be better than a solid one that would form two separate "compartments'. The middle layer could even be a different material if you fancied a bit of CLD.

[IslandPink]

While true, the mechanical energy at 2kHz is much smaller than at 100Hz. ie, panels are far less likely to be activated.

4x the excursion is required for every octave more of LF extension.

Chris

Yes but on the opposite side of the equation you have the Fletcher-Munson curve.
Also the smaller size of the unsupported areas of the thinner panels, between the stiffening pieces, are more likely to have resonances in the upper mids and low treble.
I'm not saying it's not worth trying.
My own experience of loudspeakers is that the fewest ( ie. none ! ) voids was best , although there's a lot of differences in the three materials I used ( chipboard, MDF and then perspex ) in various ways.
This is from experience with the Quasars and horns.
I remember having conversations with James about how panels with voids were death to tone in musical instruments - which is a somewhat different problem, but still has some relevance to mounting a loudspeaker.

[vinylnvalves]

Thinner panels supported panels will generally have a higher natural frequency- so not a surprise.

The old chestnut about panels acting like musical instruments, last thing we want is a speaker box adding “tone” to the signal. Yes the BBC used thin 9mm panels, those speakers arent that neutral. ( by modern standards)

Adding damping is probably the most effective way, Metallic structures bonded to wooden panels good for CLD’s. CLD’s need relative movement to work effectively. An aluminium panel bonded this green glue to chipboard or ply, or more aluminium better will result in the most damping without significant movement. Top speakers have gone away from wooden cabinets - look at Magico cabinets.

Cork is an auxetic material so a very good damper layer, probably the most readily accessible we can get our hands on.

Perspex is damped inherently by its structure, but isn’t that stiff, so works well as thick panels without anything else.

[simon]

this green glue

What's green glue?

[vinylnvalves]

https://www.greengluecompany.com/

It’s a viscous damper. I have been looking at the using the joist tape, for attaching aluminium ( or steel) onto the inside of some cabinets. There are lots of other adhesives which are good dampers, such as epoxy I need to check my notes, from over a decade ago, i did with Sheffield University damping UTC for what products have the best properties, in the audio frequency range. Thinner is generally more effective as the have more shear.

The best cabinet I have heard was at the WAM show where someone had lined the bass cabinet with code 4 lead sheet he had bonded on the inside.

[simon]

Interesting stuff, thanks. Curious it's called green glue but isn't actually a glue...

[IslandPink]

The old chestnut about panels acting like musical instruments, last thing we want is a speaker box adding “tone” to the signal.

That wasn't my point.
Have you found a way of preventing a loudspeaker from interacting with the panel that it's mounted to ?

[vinylnvalves]

Magico’s approach, aluminium plate costs make birch ply look cheap though. https://www.magicoaudio.com/enclosures

Sorry for assuming you were going down the rabbit hole of using panel vibrations to argument the lower frequency bands, which were popular in their day, and with some full range speakers today.

There is a trend to use a false speaker fronts, which allows steel/aluminium plate to be used as hidden to dissipate energy better.

[Scottmoose]

Magico have some of the better box construction at present. Avalon too, albeit with more conventional materials.

Fortunately, other than a couple of boxes from Ocellia, Musical Affairs & the Bosendorfors, I don't know of many current speakers (whether wideband or multiway) that are designed to use panel resonance, which isn't a bright idea if you don't want the speaker to introduce more colourations of its own devising than you can possibly avoid. I know a lot of designs, most of my own included, that use high rigidity structures with a high panel Fs, but they do that specifically to avoid / reduce audible panel & structural resonance under operating conditions.