Folding a Midbass Horn

[dirkwright]

I'm trying to figure out a method for folding a midbass (100-500 Hz) horn into a La Scala type of shape. I think this is called a "W" folding. I think I have an idea for how to do it, but not sure it would work since I've never built a folded horn. I got the data for a straight horn already, but it's too large for my house.

[AllenB]

Like this? AudioFanaticLaScala.html

[dirkwright]

Originally Posted by AllenB

Like this? AudioFanaticLaScala.html

Thanks Allen! I have those drawings. I guess I should have been more clear. The problem I am having is going from a given straight horn to a folded horn in the style of the La Scala. This particular bass unit is strange in that both the top and bottom panels of the horn are flat and parallel.

In the first figure is an exponential expansion for the horn I want to make. Imagine that this is half of a horn and we are looking down on one side of the La Scala in plan view. So, the problem I am having is where to cut this curve so that it fits perpendicular and fits into the 37 cm half width. I can't come up with a formula that only has 1 variable in it. I've got 2 equations and 3 unknowns... sigh...

[weltersys]

Originally Posted by dirkwright

Thanks Allen! I have those drawings. I guess I should have been more clear. The problem I am having is going from a given straight horn to a folded horn in the style of the La Scala. This particular bass unit is strange in that both the top and bottom panels of the horn are flat and parallel.

In the first figure is an exponential expansion for the horn I want to make. Imagine that this is half of a horn and we are looking down on one side of the La Scala in plan view. So, the problem I am having is where to cut this curve so that it fits perpendicular and fits into the 37 cm half width. I can't come up with a formula that only has 1 variable in it. I've got 2 equations and 3 unknowns... sigh...

Whether you choose "W", or spiral folds you want to make the horn keep approximately the same cross sectional area (height x width) as the straight horn.

Since the top and bottom are going to be parallel, this will require a horn expansion viewed from the top as expanding more rapidly than a horn that expands in both horizontal and vertical dimensions.

The more stations you choose to keep the exponential expansion the same as your model, the more side pieces of the horn are needed.
In my experience with straight horns, each using the same throat and mouth size and same length, going from a curved, to four flat sections and then to a two section horn made little audible difference.
Folding in itself is a compromise to sound quality, so trying to get the expansion perfect is a little like "putting lipstick on a pig". That said, I have heard plenty of "W" folds, AKA "re-entrant" speech horns that sound pretty un-pig like.

[dirkwright]

Originally Posted by weltersys

Whether you choose "W", or spiral folds you want to make the horn keep approximately the same cross sectional area (height x width) as the straight horn.

Since the top and bottom are going to be parallel, this will require a horn expansion viewed from the top as expanding more rapidly than a horn that expands in both horizontal and vertical dimensions.

Thanks. I agree of course. My problem is where do I cut the horn flare to fit into a "W" shape? It's a basic math problem, as far as I can see it. A "W" folding has only 2 of the 90 degree turns in the folding, which is pretty good in my opinion.

Otherwise, I can make this horn for 1/4 space and make it straight. The size in that case would be 82.5 cm long by 48.5 cm square, not including the rear chamber. Add 12 cm to the length for that...

[weltersys]

Originally Posted by dirkwright

Thanks. I agree of course. My problem is where do I cut the horn flare to fit into a "W" shape? It's a basic math problem, as far as I can see it. A "W" folding has only 2 of the 90 degree turns in the folding, which is pretty good in my opinion.

Otherwise, I can make this horn for 1/4 space and make it straight. The size in that case would be 82.5 cm long by 48.5 cm square, not including the rear chamber. Add 12 cm to the length for that...

Cut the horn flare where the first half expansion most closely matches the the area of the model ;^).

The math is the easy part, you can simply multiply height times width to see what the area should be at whatever point in the horn expansion.
The difficult part is making a fold that uses the least parts, shares the most walls, and still conforms to the above.

I do the math, lay out the expansion, then cut the picture up and flip the parts around until they fit into the desired shape. Add a few reflectors, loft it up full size and cut the parts.

Art

[dirkwright]

Originally Posted by weltersys

I do the math, lay out the expansion, then cut the picture up and flip the parts around until they fit into the desired shape. Add a few reflectors, loft it up full size and cut the parts.

Art

Jaha! you understand my problem. There must be a mathematical way to figure out precisely where to cut the expansion so that it all comes out right instead of experimentally cutting up paper until it all fits. I tried to show that in my attachments.

[dirkwright]

OK, I think I've got it. The first image is the horn expansion for one leg of the "W". The line on the left is the length (from throat to mouth) and the one along the bottom is 1/2 the area. I cut it at 45 cm along the length and it seems to fit OK, as shown in the second image. I've sketched in 1/2 of a 12PE32 as well. So, the horn winds up about 65 cm deep and 74 cm square. It should perform well if it's put up against a wall.