Reconstructed Smith Horn Plans

[whizzer]

As a long-time audio freak and music lover and sometimes DIY loudspeaker systems builder, I've long been intrigued by horns, up till now, mainly bass horns, But the seemingly rather simple construction of of the Smith Horn as shown in Ruedigger's posting of Smith's original article prompted me to consider building a pair--I'll probably get around to it next summer when I'm "off contract" at the University for a month. I've used the Eminence ASD1001S driver and like it for several reasons--even though it's inexpensie, it's well constructed and has a smooth, non-peaky, hi-fi-like response, devoid of the severe peaks and dips in the 7-12 kHz range that impart that shrieky, fatigue-inducing harshness that mars the performance of many drivers in its price range--and some that cost a lot more, too. And it works very well with the small elliptical waveguide sold by Parts Express. Admittedly, its response dwindles rapidly above 18 kHz, but my hearing dwindles rapidly above 13 kHz, so that's a non-issue. Anyhow, I think it'll be a good way to get a feel for this design and let me know if I want to scale the design up for a 2-inch driver--the taper, for an 800 Hz crossover, would be the same, only the vertical dimension wouold change. The reproduction quality of the article was poor, and the drawing looked rough and not exact anyhow. I've analyzed the geometry of the design, and made a few non-critical modifications in the views attached, courtesy of Adobe Photoshop, v. 7.0.

[whizzer]

I reached the limit of attachments in the above post, so here are the rest. (For some reason, theyre out of order, but since they're labeled, it probably doesn't matter).

[Ruediger]

Great!!!!Fantastic!!!!Thank You!!!

Ruediger

[whizzer]

I'm sorry I misspelled your name. Naturally, you don't use inches in Europe, but if you open these things in Photoshop, you can set the rulers to metric. The native resolution is 120 dpi and the scale at that resolution is 1:1.

[Mr. Widget]

I had never really bothered to decipher the magazine article that you so carefully interpreted. For example I hadn't noticed that plug before. I wonder how these "Smith" horns sound compared to the JBL 2397 "Smith" horns?

I hope someone builds a pair and does a comparison. If anyone in the SF Bay Area is interested, I'd be happy to lend my pair of JBLs for the comparison.


Widget

[whizzer]

I had never really bothered to decipher the magazine article that you so carefully interpreted. For example I hadn't noticed that plug before. I wonder how these "Smith" horns sound compared to the JBL 2397 "Smith" horns?

I hope someone builds a pair and does a comparison. If anyone in the SF Bay Area is interested, I'd be happy to lend my pair of JBLs for the comparison.


Widget

One difference I noticed right off is that a JBL version I have seen has a "throat," apparently a short conical horn, between the driver and the horn proper. Another version I've seen, and I don't know if it was the 2397 or or some other model, has the vanes not reaching to the front periphery of the radial front, and the edges of the top and bottom curve upward and downward, respectively, apparently either exponentially or hyperbolically, like a "normal" radial horn, relying, evidently, on expansion rather than diffraction for vertical dispersion. Whether these physical differences make an audible difference or not would require, just as you suggest, listening and testing.

[pos]

For example I hadn't noticed that plug before.

I don't think that plug is supposed to be there.
I remember reading it appeared in original article drawing just as a guidance for the vanes.

[Mr. Widget]

I don't think that plug is supposed to be there.

That would make sense... I can't imagine it working out well on axis.


Widget

[whizzer]

I don't think that plug is supposed to be there.
I remember reading it appeared in original article drawing just as a guidance for the vanes.

It wasn't mentioned in the article at all, and the labeling on the drawing was impossible to make out, so I went with what I could see. It will certainly make construction simpler if it's not supposed to be there. I thought it looked weird, but it was, apparently, there. Perhaps a caption indicating "guidance" for the positioning of the vanes might have been less confusing. Undoubtedly, Smith did not foresee a hobbyist of the future electronically poring over a poorly reproduced and skewed reproduction of his sketch in order to make a plan.

[Steve Schell]

AFAIK there is absolutely no plug in the throat on any Smith Horn design. This has been a point of confusion in the past as the drawing in Smith's article can easily be misinterpreted. Installing a plug will only serve to keep most high frequency energy out of the center channels. I will try to find the article and perhaps do a hi res scan of this area in the drawing.

[Steve Schell]

Whizzer, Ruediger's scan of Smith's article in this thread contains (barely) enough resolution to determine that there is no plug. Just save page 2 and blow up the throat area in your picture viewer:

http://audioheritage.org/vbulletin/s...ght=smith+horn

As you can see, what appears to be a plug is actually just an indication of the center axes of the "torpedo" cell dividers and the twenty degrees included angle between each pair of them. The original drawing is confusing, no doubt. Aside from this frequently made error, you've done some very nice work here.

[spkrman57]

Great info!

Ron

[Maron Horonzakz]

Disregard the plug,,, I have a Smith horn built in 1950,,useing the plans from the AUDIO mag.. With the correct septa shapes,,,This horn is larger,,31" wide. mounted to a 375 JBL driver... The polor response is better with minimal lobeing.