Discussion on what materials to build speakers out of

[Peter Daniel]

This is how it looks from the other side (first order filters)

[454Casull]

Originally posted by planet10



I'd have to look it up to confirm, but i think that adding mass lowers resonance by the square.

dave

In that case resonant frequency would be linearly proportional to thickness.

[Peter Daniel]

The complete speaker also contains two woofers (12 and 15" Focal Audioms) in folded baffle setup. The boxes are made of dual layer 3/4" MDF, plastic laminate on the outside. Both sections are attached together with 0.75 x 1.5" aluminum bars. Bottom stand is also filled with sand. The whole thing is modular and easy to transport.

[Peter Daniel]

Some of you may remeber those speakers from years back, but this is the latest incarnation and better than ever.

[MJL21193]

Beautiful work Peter.

[MJL21193]

Originally posted by Scottmoose

MDF cabinets attempt a similar thing, but being far less massive need to be much thicker, obviously. Much thicker than most people realise -although it's dense, MDF isn't very stiff at all. For a typical FR driver application, you really need about 1 1/4in thick panels minimum to push panel resonance down into a place where it doesn't do any damage. Either that, or line the regular 3/4in material with at least 1/4in of concrete board, as Bob Brines does with his MLTLs. That works, but you then run into other problems such as energy storage. If you can't go to something with the density and stiffness-weight ratio of steel, better to push them up, out of the way, IMO.

OK, since everyone says there is such a hugh difference in materials, here's what I will do:
I will build three boxes, of three different materials - one of MDF, one from baltic birch plywood, and the third from cheap spruce sheathing. I will construct each box from the same thickness material (1/2"). I will build each the same way - with glue and 1.5" airgun nails. They will be identical in every way. I will make one hole for a midbass driver that has a normal Xmas~5mm.
I will then test each box for panel resonance, using nearfield frequency response with the mic 1/2" from the largest panel.
There will be no damping material, the boxes will be empty sealed boxes.

Has someone already done this and published the results?

Originally posted by planet10


In a box like the Fonken the largest unbraced panel is just over 4" -- sometimes they will get out to as much as 6-8". The bracing doesn't need to be a matrix you just have to orient it correctly (ie the ubiquitous shelf brace is one of the least effective.

Air space damping, no panel damping. That just takes you in the wrong direction. Lower panel resonance, greater energy storage, & lower panel resonance Q.

The box doesn't suck it out, it buries it in the low level noise prodcued by the cabinet.

Doesn't an undamped box add it's own colouration?

[Ed LaFontaine]

454Casull posted:

In that case resonant frequency would be linearly proportional to thickness.

I think there's an inverse relationship in there....

For my own piece of mind (I'm not going to build 3 prototypes) I will add 3/4" dowels across the interior of the case. They'll be spaced with some influence allowed to ratio's of 1.618...

[planet10]

Originally posted by MJL21193
Doesn't an undamped box add it's own colouration?

The boxes are damped, the panels only the inherent damping in the plywood.

All panels add colouration. The idea is to minimize them.

dave

[planet10]

Originally posted by Ed Lafontaine
For my own piece of mind (I'm not going to build 3 prototypes) I will add 3/4" dowels across the interior of the case. They'll be spaced with some influence allowed to ratio's of 1.618...

Ed,

What box are you building? Holey brace(s) are a lot more effective than dowels.

dave

[Ed LaFontaine]

Hi Dave,

The box will closely resemble this:



The box dimensions will be 39" h x 7.5" wide x 14" d. It is an effort to get as much out of an RS180 pair as they have to offer. I think the dowels should be connecting the side panels.