900hz Smith Horn for 1.5 Drivers

[John W]

I just finished a pair of smith horns for a pair of 2345 drivers, and did some measuring this weekend.

I started with the design drawn up by yggdrasil in the 1200hz thread. http://audioheritage.org/vbulletin/s...ead.php?t=5454, then scaled it up for a 1.5 in driver, which comes out to 900hz. A lot of the design aspects are covered in the original 1200hz thread.
I’ve attached the spreadsheet with the modified numbers. If you have a hard time deciphering the spreadsheet, basically:

• the horns are an arc with a 12.75 inch radius
• fins are 7.29 inches long and 1.18 inches high
• the throat is 1.5 inches diameter

These horns are mahogany and the material is 1in thick.

[John W]

a couple more shots

[John W]

Here are a few in-room response plots from the horns. The first is the horn only, no filter, with the mic about 1 meter away pointed directly at the mouth of the horn. You can see the response gently rises to about 15k then drops off.
The second is horn only at a little higher volume level, and the mic about 2 meters away. For some reason this flattens the slight rise.
Finally here is a plot from about 2 meters crossed over to 15” E145 woofers at 800Hz using a Behringer DCX2396 digital crossover flattening out a couple of the peaks, and boosting the VLF response. This is a 2 way setup, the tweeter is not hooked up.

They sound very good. Especially at moderate volumes (mostly because of the room)

[4313B]

Looking good!

[paragon]

Very good work ! Good response. Excellent !!
Love it !

Eckhard

[spkrman57]

Shows what good thinking will come up with!

Ron

[Zilch]

Beautiful work, John.

[And nice measurements, too!! ]

[John W]

Thanks.
Back to Mr. Timbers notch filter, which I've been trying to figure out.
I can see the CLR filter to flatten out impedence peaks, and some attenuation. What else is it designed to do? Would most of it bennefit a passive crossover?

[jack_bouska]

Here are a few in-room response plots from the horns. The first is the horn only, no filter, with the mic about 1 meter away pointed directly at the mouth of the horn. You can see the response gently rises to about 15k then drops off.
The second is horn only at a little higher volume level, and the mic about 2 meters away. For some reason this flattens the slight rise.

The difference in spectral tilt could be caused by one or more of:
1) nearfield vs. farfield differences, the smith horn mouth is a diffraction slot in the vertical dimension, and you may need to make mesurments (from mouth to mic) at about 5x or 10x the largest dimension of the horn.
2) early reflections from the floor, or the speaker underneath can lead to dips in the response (mostly a sub 1khz issue)
3) the mic position in the room, and the measurment distance from the devices will vary the direct vs. reverberant (diffuse) sound ratio, and can alter the spectral tilt, depending on how live or dead your room is.

Try to make a few (3+) measurments from in and around your seating position, and visually average the various curves to get a better idea of performance. Concentrate on dips, peaks, and trends that are common among the curve families, as these will be associated with the device, rather than the room.

(nice craftsmanship, by the way)

Jack

[John W]

Thanks Jack,
Nice to hear from you again. I normally do the final adjustments with the mic in the listening position, but was wondering why they changed as you got closer. I figured it was somethin like you described.

[Joe Alesi]

Conratulations John,

They look gorgeous- Starts me thinking about growing it to a 2" throat for a 2440/2441. Any Thoughts guys?

[John W]

Conratulations John,

They look gorgeous- Starts me thinking about growing it to a 2" throat for a 2440/2441. Any Thoughts guys?

The 2397 design is already available for 2" drivers. This design is a slight shrinking from the already full grown original.

[4313B]

The 2397 design is already available for 2" drivers. This design is a slight shrinking from the already full grown original.

Yes, they were everywhere back then. If it wasn't a Smith it was a Serpentine.

[Joe Alesi]

Thanks guys--- opps , Back to school. I suppose what I should have said is grow it to a 500 Hz cut off.

[4313B]

500 Hz is a terrible crossover frequency.