DIY 500Hz Smith Horn? Please help me design one!

[mpaschetto]

Hi Guys!

I've searched through the archives, and read successful accounts of scaling down the Smith horns for a higher cutoff frequency. However, I've yet to see any information about scaling UP a Smith horn for a lower cutoff frequency.

Has anyone tried this yet?

Is it simply a matter of following the formulas in the 'Distributed-Source Horn' article? If so, I'll plow through the math to get a good starting point, but I'd need a recommendation for the throat area (So). I'd like to use a mouth thickness (t) of 2" for better integration with my 2" compression drivers.

One more question:

Exactly what is the throat area (So) Smith refers to? Is it the starting area of one of the individual cells?

Thanks in advance,

Mike

[hapy._.face]

Is it simply a matter of following the formulas in the 'Distributed-Source Horn' article?

Beats me- Ken or Widgett might chime in here. Or you could PM them; I'm sure they have an answer for you.

[JuniorJBL]

Dr. Widget

Knows quite a bit about this "area";)

[Fred Ireson]

I'd like to know too how to lower cutoff - width's no problem

also - are there any followup to Johnny's 5Khz Smith-horn? don't understand the pointy conical thing in his drawing

Old Article


http://www.audioasylum.com/forums/HUG/messages/98507.html

[yggdrasil]

I'd stick with rescaling the JBL 2397.


When entering the numbers into the spreadsheet it gives you a radius of 62cm/24,5 inches. The horn will then be ~ 110 cm/43 inches wide.

[Fred Ireson]

thanks Johnny - 43 inches - haha - might do something funny with power response & EQ needs too (?)

do 1" format Smith load OK to 800Hz?

heres one with a funky Peavey 22A
http://img216.imageshack.us/img216/3853/shgjpg9or.jpg

Best,
Freddy

[yggdrasil]

do 1" format Smith load OK to 800Hz?

If you are referring to the 1200Hz Smith's I built - NO.

You will have to recalculate the horns to the actual height you will be building. E.g. 1"/2,54cm ~ 90cm / 36" wide.

[yggdrasil]

Actually I have played with the idea of building a giant Smith horn to replace the 12" midrange driver.


Even with 3" driver it gives a horn that is:

f0 = 300Hz: 225cm/88" wide
f0 = 400Hz: 125cm/49" wide
f0 = 500Hz: 80cm/31" wide

Given this - the Smiths get too large for a small home use below the 800Hz, like the 2397 specifications.

According to Widget they are not usable/flat down to 800Hz

[Flemming Skov]

Hi Guys!

I've searched through the archives, and read successful accounts of scaling down the Smith horns for a higher cutoff frequency. However, I've yet to see any information about scaling UP a Smith horn for a lower cutoff frequency.

Has anyone tried this yet?

Mike

Hi Mike,

Any progress in you search for a smith horn construction details? I have a pair of 2380 horn to start with but have relized that vocal sound is a bit "closed" or "in a box" like, so to say (Ido not EQ them).

For me it seem that a smith horn is the way forward so I am also looking for construction details regarding a smith horn for 500 Hz as the low and used with the 2445 driver to xover at 7kHz to my 2404. I have build a set of corner horn. Any interested see the link:
http://forums.klipsch.com/forums/thread/741550.aspx

Regards Flemming

[ralphs99]

I guess my first reaction is to ask why?
The Smith horn is an attempt at getting high frequencies to disperse better. The wavelengths at midrange frequencies are long enough that beaming is no longer a problem.
In fact containing midrange frequencies is more of a problem than spreading them out. I'm not sure how well a Smith horn can control midrange directivity.
Cheers,
Ralph

[Steve Schell]

Hi guys,

I thought I had posted the original article by Dr. Smith, "A Distributed-Source Horn" on the forum in the past. I couldn't find it in a search however, so here it is. Note that Dr. Smith mentions that his design was to be used above 750Hz., even though it has an exponential flare constant of 375Hz. This is very conservative design of the old school, no doubt intended to avoid the response ripples of a finite exponential horn down near cutoff. My guess is that his design could be reworked for a 2" throat and a slower flare rate and still have a reasonably compact horn. Smith horns were built to cover different frequency ranges in the 1950s; I have seen pictures of at least one system using three different sized Smith horns plus a woofer section.

People are often confused by what looks like a rounded triangular section in the throat in the drawing. It is only there to show the angles of the "torpedo" dividers, not to indicate a solid structure.

Sorry for the crummy photocopy- not my work. I have the magazine, so if there is interest I can make better scans from the original.

[Zilch]

JBL dimensions here:

http://www.jblpro.com/pub/obsolete/2397.pdf

See below for one approach to the recommended "lips." ;)

[Flemming Skov]

Steve,

In the text is noted 2 formula's for each calculation. The last formula using inches as the text says. Do you know if the first formula is expressed in the S - I unit system (metric) ? I will try to calculated one for xover at 500 Hz. The JBL2397, if looking in the data sheet, is measured at 500 Hz so it seem that this horn is capable of producing sound at that low frequency. Is it recommanded to use the JBL2397's overall measures in designing a new horn? Does anybody know?

Regards
Flemming

[Steve Schell]

Flemming, I am not good enough at math to decipher the equations, but I doubt very much that any of them used metric units. The article is from 1951, a time when the metric system was used very little in the U.S.

One general principle that is used in exponential midrange horn design is to use a flare constant that is lower than the horn's intended use, often an octave lower or close to that. The Altec and RCA 300Hz. multicellular horns, for example, have an actual flare constant of about 180Hz. The reason is that a finite exponential horn (all of them in the real world) has an impedance mismatch at the mouth due to the abrupt termination, and reflected energy travels back down the throat to affect driver loading, both constructively and destructively depending on frequency. This results in both peaks and valleys in the response. These response ripples are concentrated in the octave above the cutoff frequency, and the shorter the horn is (the smaller the mouth) the worse their amplitude becomes. The popularity of the old multicells with audiophiles is due in part to their very smooth response, which is attributable to their conservative design and large mouth sizes (small ripples).

Zilch, I am fascinated by the picture you posted. Do you have these horns, or know where the picture originated? The reason I ask is that a friend of mine owns several pieces from Bart Locanthi's estate, including three horns that look almost exactly like the ones in your picture. Two of them are fitted with large format Pioneer compression drivers, and the third has a small format 1" driver. The drivers are branded Pioneer and some other name, Elite or similar. They closely resemble the TAD 4001 and 2001, but were made before Pioneer settled on the TAD brand name. Anyway, the horns my friend has are beautifully made but in a raw wood finish, as they were apparently prototypes built to Mr. Locanthi's specifications for testing.

[johnaec]

Zilch, I am fascinated by the picture you posted. Do you have these horns, or know where the picture originated?At auction on eBay, with lots of pics: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=7422379040&rd=1&sspagename=STRK%3AMEWA%3AIT&rd=1

John

[Mr. Widget]

The drivers are branded Pioneer and some other name, Elite or similar. They closely resemble the TAD 4001 and 2001, but were made before Pioneer settled on the TAD brand name.Originally the high end Pioneer products were all called Pioneer Exclusive. Later they were called TAD. The ET-703 tweeter is the only product to survive from the early days. The current ET-703s are still labeled both Pioneer Exclusive and TAD.


Widget

[Steve Schell]

Thanks johnaec, that explains it. Sure wish we knew who made those horns. I'll have to take some pictures of the Locanthi horns the next time my friend is in town.

Exclusive, that's it! Thanks Mr. Widget; I couldn't think of that name to save my life.

[johnaec]

The only thing that concerned me about those black "Smith" horns on eBay is that unlike original Smith horns, the ones on eBay had no vertical flare. Maybe it's insignificant - I don't know...

John

[Maron Horonzakz]

Its not significant.

[John W]

These horns are/were mine. I bought them with some other drivers from the Seattle Rep Theater, which is a large playhouse downtown.
I can probably answer a few questions or take a few more measurements if anyone is interested, but I will need to pack and ship them out as soon as payment clears.
As far as construction is concerned the fins are mdf, the top and bottom look to be baltic birch, and the sides and support pieces are mdf. The fronts are curved 1/4 plywood. I can't get in and look at the exact shape of the fins.
I received one note from the auction that said the horns were probably built by MSD Audio which closed its doors around 1994.

[jacopuzzo]

Hi guys,

I thought I had posted the original article by Dr. Smith, "A Distributed-Source Horn" on the forum in the past. I couldn't find it in a search however, so here it is. Note that Dr. Smith mentions that his design was to be used above 750Hz., even though it has an exponential flare constant of 375Hz. This is very conservative design of the old school, no doubt intended to avoid the response ripples of a finite exponential horn down near cutoff. My guess is that his design could be reworked for a 2" throat and a slower flare rate and still have a reasonably compact horn. Smith horns were built to cover different frequency ranges in the 1950s; I have seen pictures of at least one system using three different sized Smith horns plus a woofer section.

People are often confused by what looks like a rounded triangular section in the throat in the drawing. It is only there to show the angles of the "torpedo" dividers, not to indicate a solid structure.

Sorry for the crummy photocopy- not my work. I have the magazine, so if there is interest I can make better scans from the original.

Hello,
I'm trying to build a Smith horn and it would be very helpful if you could send me some better scans of the article.
Thanks in advance
:)

[hjames]

Hello,
I'm trying to build a Smith horn and it would be very helpful if you could send me some better scans of the article.
Thanks in advance
:)

Whoa - that post was from almost 2 years ago!
There are other scans of Smith Horns in the forums - best bet is to do a Search for them and see what you come up with - ... like this:

http://audioheritage.org/vbulletin/search.php?searchid=964154

or THIS!
http://audioheritage.org/vbulletin/showthread.php?t=19962&highlight=smith+horns

and then check THIS!
http://audioheritage.org/vbulletin/showthread.php?t=18356&highlight=smith+horns

Lotsa info here!