Questions Regarding Horn Design and Dimensions

[TN Allen]

The latest idea for producing a horn involves milling block sections made of 2" stacked and glued building insulation foam board. There would be 4 blocks, these would be about 16" X 30" X 8+". These would contain mirror image sides of a spiral horn, split along the center plane of the spiral. 2 blocks would be glued edge to edge to form a 32" X 30" X 8" block containing one mirrored horn, 2 others the opposite mirrored horn. These two 32" X 30" blocks would then be assembled to contain the complete horn, other than the 3D and petal sections.

The assembled horn would connect via a short 3D printed 90 degree bend on one end to an Orthophonic tonearm that I have removed from the Granada. The other end of the spiral would have an approximately 17" diameter, which would connect to a 3mm Baltic Birch petal mouth increasing the mouth termination diameter to about 30". I have attached a 3D model of the assembled sections. with the exception of the largest section, the illustrated sections represent the core of the horn. The core is subtracted from the blocks, which are then milled. There are also photos attached of a quick experimental trial run milling the foam.

The foam is so soft that milling it to produce a smooth surface is a problem. Light sanding improves the surface but easily removes too much. In this case I ground/milled one sample with a spherical burr, the other with a mounted wheel dressed as a sphere. The dressed wheel produced a marginally better surface. Neither is as good as I'd like, but may need to suffice. Rustoleum improves the surface somewhat, but probably isn't worth the trouble. The 3D CAD horn model core is based upon the 14' horn dimensions in Modern Gramophones and Electrical reproducers. If this were to work out well enough, I'd add the larger mouth completing the original MG&EP 68" diameter design. There is also the possibility of replacing the Orthophonic tonearm with a longer tonearm consistent with the original horn dimensions.

[Inigo]

A very good idea indeed! Much easier to produce! And the material is inert enough, I think... If not, it could be made on wood blocks, or this foam could also serve as a mould for making a horn in solid clay or something, that in turn could be used again as counter mould or mother for repeating the block with another mouldable material many times...
How did you mill such a complex shape? Did you use computer-driven tools? By hand?

[recordmaker]

These foams are normally coated with an epoxy or other gel coat and this can be sanded and polished for use as the mould and treated with a release wax before making the fiberglass layup that will become the horn.
There are also special machining grade boards that can be used with a polish and a seal coating to form a mold.

[TN Allen]

It is a very nice Haas CNC machine. The design was done in KeyCreator. One of the most remarkable changes in maching has been the potential to produce complex shapes very accurately and efficiently that are not based upon geometric primatives. Before the availability of CAD and numerucally controlled machining, the shape in the photo was much more difficult to produce. I wonder what gramophone designers and manufacturers might have done with this technology.

This foam is ordinarily used to insulate buildings, it very soft, and was never intended for milling. Were cost irrelevant, I would use signmakers foam, which mills beautifully and is easily sanded and coated, however, the foam in the photo costs ~$40/4'x8'x2" sheet, signmakers foam cost ~$500/4'x8'x2" sheet several years ago. For the present, the insulation foam seems like a good choice for a prototype.

Wood would be good as well, but even Basswood, which would be a good choice would cost $3/board foot. The foam panels contain 64 board feet. If the prototype compares favorably to the Credenza I have, then perhaps the cost of nicer material will be worth considering.

[physicist]

You might want to consider using rigid polyurethane foam in place of the polystyrene insulating foam.
It's stronger that the polystyrene and machines well.

This is the website of a UK supplier:-

https://www.allscotltd.co.uk/polyurethane-foam-sheet

I'm guessing something similar should be available over the pond.
It looks to me as if pricing is not dramatically different to the polystyrene foam and it
is available in thicker blocks.

We used this for many years to cnc mill anatomical models in the days before the availability of 3d additive printing.

anatomical model

[TN Allen]

Thank you, I'll look into that.

[Inigo]

Of course the foam you've carved serves the purpose for checking the operation of the horn, as long as it contains the desired volume of air, without being altered or deformed, and doesn't introduce parasite vibrations. If successful, then other options to redo it in a more solid or stable material can be investigated...

[TN Allen]

The image is of my latest thinking regarding milling a horn using the Wilson & Webb radii on page 100 in Modern Gramophones and Electrical Reproducers. The top layer is 60% transparent to show the lower layers of foam. The throat is 1.280" diameter to make up to an Orthophonic tonearm I have, and the mouth radius is 17.5". This horn is all but the last 2' of the W&W horn sound path dimensions. It appears that all but the bell shaped mouth can be milled from a single 4' X 8' X 2" sheet of insulation foam. I haven't been able to find Polyurethane foam at a reasonable price, though I'm still looking. I think I can dress the Kutzall burr to cut a cleaner surface, which may make the insulation foam milled surface acceptable. The key is to avoid sanding the milled surface as it is so easily damaged, even with gentle sanding. The mouth may be best done in the insulation foam as well, as this is a trial horn, and I don't want to spend money on materials for a prototype that may not succeed. 8 segments can make up the mouth section, which will require an additional ½ sheet of 2" foam. I'll need to keep my grandsons clear of the mouth, they'd have a great time poking fingernails into the foam.

The 2 contorted throat sections can be 3D printed. I may need to add a short 1.280" diameter flanged sleeve depending upon the vertical location of the platter. From what I've read though in W&W that may not adversely affect the sound.

This has been quite an interesting design project. The laptop the school supplies me limped along through some of it, but balked at all the computation required to produce and subtract some of the solids. This apparently all but did the laptop in. The IT dept. took it in to try to speed up the calculations, and replaced the hard drive with a new solid state HD. I was told the battery was "swollen", and likely to explode and create quite a fire. I was also told Lithium Ion battery fires are difficult to extinguish. The man who did "surgery" as he put it, said he knows from previous experience.

[Inigo]

Nice! Let's see how it results! I wish you luck with it!

[TN Allen]

A little progress.
The first passes have a .5" change in depth, which was just to try the geometry and setup quickly. I'll redo the toolpaths with a .01- .05" dropdown (change in Z) for the finish passes.

The geometry will require 4 setups because the Haas milling area is 30" X 16" (X&Y), and the horn geometry is 42" X 24", were the blocks 1" more in Y, or a few more in X, there would not be sufficient travel insude the shroud. Fortunately the Haas has 21" of travel in Z, depending upon the tool length. Even with the extention, there is adequate travel in Z to accomodate the cutter, which requires a maximum depth of 7.75".

There is much more to do, the printed sections and turned mouth, and assembly. With teaching and cold weather in Maine, progress will be slow.