Scottmoose wrote: ↑Wed Jan 01, 2020 10:04 pm
I hadn't seen that drawing, many thanks for the heads-up.
Scott, if its of interest here's a link to the Lii Audio page with their designs for these drive units, including both sheets for this pear-section design.
Given the layout & physical vent positioning, it's a mite awkward. However, you could plot Vmax and suspend about, oh, an inch thickness of material across the box to help damp it out (point of maximum velocity is where damping is most effective). Probably won't get all of it, but should help. But then, this is at best academic, purely from a quick look at the plans and noting broad features, not a critique (other than a gentle dig at a bit of purple prose ). For e.g., a vertical mode may not be audibly problematic; depends on the details, and ultimately individual preferences.
Scottmoose wrote: ↑Wed Jan 01, 2020 10:36 pm
Given the layout & physical vent positioning, it's a mite awkward.
Is it the parallel top and bottom panels that give rise to this potential issue? If so could it be mitigated by making them non-parallel, say by 'doming' the top panel?
In case it isn't obvious, I'm not a speaker theorist/designer!
Yes; you get a standing wave (eigenmode) developing between any parallel surfaces. Shorter distances = shorter wavelengths & require less material to damp out.
Possibly; it would have to be a fairly large alteration to produce a significant effect. As I say though, this is something of an academic point. Assuming it's there and there isn't something else not shown in the drawings that has been added to suppress it, the existance of a vertical mode is not automatically a bad thing and / or necessarily audible. I can note a technically questionable comment on the Decware site and also point out some inconsistencies in the published data, with a possible reason for that (values individually averaged from a batch), but I've not heard or tested the speakers, and who am I to disagree with those who have? They look like they have reasonable potential to me as an example of its type.
Bottom and top are 15mm birch plywood, sides are made from 3mm poplar plywood. The parts that were cut off making the bottom and top were later used for pressing/gluing the sides to the bottom/top.
I am seriously considering now to try such a speaker project in the near future. I have OTLs (PP & SE) for driving 8ohm speakers and don't have to experiment with resistors in series. Could be fun!
Wolfgang wrote: ↑Thu Jan 02, 2020 4:10 pm
I found a report...
Wolfgang although of a similar shape these speakers are designed to be resonant, as suggested by the name 'cello'. They are open at the bottom. There's more information about the concept on Troels Gravesen's website, about halfway down this page;
I believe that the Lii Audio pear-shaped speaker cabinet needs to be rigid, evidenced by its use of internal braces. The curved shape will help a lot too.
If I were to build a pair I would start by using the front baffle like the keel of a boat and set up the braces like the ribs of a boat. A secondary keel arrangement at the back will help to keep everything square. For sheathing the structure for the first layer I would use flexible MDF, like this;
This material presents an opportunity to add a lot of rigidity to the panels by exploiting the slots that facilitate its flexibility; having fixed the flexi-mdf in place I would then 'caulk' the slots with carbon-fibre tow, wet it out with slow cure epoxy laminating resin, add a lamination of wetted-out carbon fibre/glass/kevlar and then add another layer of flexi-mdf or maybe flexi-ply, subsequently adding additional layers as required. I would hope to use a vacuum bag approach to clamp everything up tightly, without voids.
Thanks Ray for clarification. I think I need to try this.Will first run some tests with bending/glueing flexi-ply/flexible-mdf in order to find out how easy (or not) it is.
In case anyone is thinking that it sounds like a bad idea to put a speaker cabinet into a vacuum bag and risk deforming, or even crushing, it - you don't. If you make the first layer flexi-mdf oversize by two or three inches you can just vacuum down the follow on layers against the first layer like this;
Atmospheric pressure takes care of the rest without stressing the cabinet.
In fact, if you do some pre-planning (essential anyway) and work quickly with a slow cure resin you could probably do all the layers, including the final veneer, in one hit.
I now have a set of CAD drawings for the pear-shaped Lii Audio speakers.
They need some work before they could be used for any CNC routing work (they're an old version of dwg file and they don't seem to have used polylines) but I do have the relevant shapes and dimensions now.
I now have a set of CAD drawings for the pear-shaped Lii Audio speakers.
They need some work before they could be used for any CNC routing work (they're an old version of dwg file and they don't seem to have used polylines) but I do have the relevant shapes and dimensions now.
Does this mean you got the "original" shapes, not just recalculated? Because from the numbers on the drawing only the radius cannot be exactly calculated. I remember that the guy who posted the building plans for the C-horns which I have built used Photoshop with the actual pictures of the C-horns (sides made of acrylic or plexiglass which made it possible to see the horn dimensions and shape) and could exactly recreate the dimensions (1:1 template for routing). That was the only way to get the exact shape of the curved horn inside the speakers.
The other question is of course how would one attach the sides (nothing really in order to clamp something on for using flexible mdf etc.) to the "ribs" if one wouldn't use already finished pre-shaped sides with the help of the "vaccum bagging" method?
Wolfgang wrote: ↑Fri Jan 03, 2020 4:05 pm
Does this mean you got the "original" shapes, not just recalculated?
That is what I had hoped for but unfortunately there seems to be an issue with the CAD files I received that distorts the pear-shaped section - I think its a result of the way the drawing has been produced (using splines!) coupled with a software compatibility issue.
However, I'm pretty sure I can regenerate the sections and I'll aim to do so over the weekend - If I can do that I will have all the information I would need to build a pair.
Wolfgang wrote: ↑Fri Jan 03, 2020 4:05 pm
The other question is of course how would one attach the sides (nothing really in order to clamp something on for using flexible mdf etc.) to the "ribs" if one wouldn't use already finished pre-shaped sides with the help of the "vaccum bagging" method?
I wouldn't pre-shape the sides. If I were to build this I would glue the first layer of flexible MDF to the ribs and then vacuum bag the additional laminations onto that - remember that the atmosphere is applying pressure equally to both sides of the panel. You would need to do the first layer on both sides at the same time to keep everything true but then it will be possible to work on one side at a time because there would be no stressing of the structure that might distort it and, anyway, applying the first layer of flexible MDF would make significantly strength the structure because of its shape - a bit like the shell of an egg. To glue the first layer to the ribs I would devise some clamps as you would only need to apply pressure to the rib areas.
Another option would be build the cabinet in two halves that you then glue together.
By referencing points on the jpg image of the pear-shaped section (from the Lii Audio website) to a grid of reference points in my CAD software I've been able to scale the jpg and establish that its basic shape is an elipse so I've created the correct the shape at 1:1 scale for most of the section. I just need to work out the curves for the rear port protrusion.