PDA

View Full Version : need some horn advice



Jakob
08-06-2005, 03:57 AM
Ok, now I want my new 2441 drivers going. I do not have any horns at this moment, but I know what sound I want. Low compression without the "horn sound" and with a fair degree of dispersion.They should also be usable down to perhaps 650 hz. I've gotten tips about the 2382A, but when I collected the 2441 drivers I saw a pair of 2395 horns and lenses and it was love at first sight. Yes they are big, and i don't know if I am allowed to have them in the house by the government but love is blind, right? How ever, love is not deaf, and I don't have a clue what they sound like.

All help appreciated and if You have a tip on other suitable horns please let me know!

Regards: Jakob

norealtalent
08-06-2005, 06:13 AM
2395 is an awesome choice. 2309/2310 would be great too. I prefer the 2397 Smith Horns myself. They are very small comparatively and have an incredible soundstage. They are also more easily made elegant due to their smaller size and can be reproduced in the exotic wood of your choice.http://audioheritage.org/vbulletin/images/smilies/bouncy.gif Dave

Akira
08-07-2005, 09:09 AM
2395 is an awesome choice. 2309/2310 would be great too. I prefer the 2397 Smith Horns myself. They are very small comparatively and have an incredible soundstage. They are also more easily made elegant due to their smaller size and can be reproduced in the exotic wood of your choice.http://audioheritage.org/vbulletin/images/smilies/bouncy.gif Dave
i agree, can't beat wood.
i have used the large lenses for years, but in a commercial application. they have a huge image, short throw relative to velocity, a 'sofened' texture and a bit of HF loss. used with a 2482 phenolic they sound like a speaker. back in the disco club days (hey part time DJ going to school :o: ) they filled the room everywhere, i used to go to work early and crank it up....just drowning in effortless, engulfing sound.
also, the 2344 biradials sound great--natural without the exaggerated horn like quality of the older radials. smooth with an image that is as big as the speaker it is mated to in the 4430 monitor. this is the EXACT SOUND that you discribe: " Low compression without the "horn sound" and with a fair degree of dispersion."

my ultimate choice? 2344 made out of wood...i think someone already built a pair. maybe we could again see pics of those horns or hear the opinion of the builder?

p.s. i can't remember what i did with my only pair of lenses. i only remember that, up in canada 15 - 20 years ago you couldn't give the stuff away...i probably dumped them for next to nothing. different market up here. we have always laughed that the U.S. market is so behind the times...the quebec market is again years ahead of us (toronto)...they are always at the cutting edge.

Harryup
08-09-2005, 02:11 PM
Well I would make another decision.
I have the 2450 togheter with Tractrix horn made from Free Form plastics.
I do have in my storage room 2382, 2370(?) and 2397. I also have 2402, 2404, 2405. I have used passive crossover and JBL DSC280. Now I use BSS366T and divide at 675Hz. I will never go back to an additional treble unit nor my JBL horns. I play any kind of music without horn colorations. I have had visits from die hard HIGH-End devoted people that just hates horns. And they are stunned. So my advice if you like to try something entirely different go for Tractrix.

Bengt

Steve Schell
08-09-2005, 11:55 PM
Hi Jacob,

I will have to put in a vote for the horn type that I have found to be the most versatile and uncolored- the conical. A conical is a straight sided horn, and can be circular, square, rectangular, or some other shape. They offer the most uniform dispersion vs. frequency of any horn type, and do not sound like horns at all. The pattern is basically determined by the walls of the horn; I have found 50 degree horizontal by 35 degrees vertical to be very useful for hi fi listening. They lack low end loading due to the rapidly expanding throat, but this would not be a problem with a 650Hz. crossover point.

If you can, supersize them. My current horns are 30" deep with 24" by 32" mouths. You can go somewhat smaller with your crossover point; I run mine down to 300Hz. I have found that there is no substitute for large midrange horns to create "big goose bumps" vocal reproduction.

One nice thing about conicals is that they are easy to build. You could build an experimental pair in a few hours for evaluation, then build nicer ones if you like their sound.

Jakob
08-10-2005, 05:58 AM
Thank you for your replies!

Bengt: I've searched the net but on "Free Form plastics" homepage I couldn't find any info reg. tractrix horns. Are they old? How ever, I should be able to get these kind of horns somewhere else. What size are yours?

Steve S: I'd love to see some pictures of your giants! Is there a way to calculate the best size depending on what You want as lowest usable frequeny? And also, can You calculate what efficiency You get?This would be helpful when designing networks.
Regards: Jakob

Harryup
08-10-2005, 06:28 AM
Jakob,
the horn are possible to order from Germany
http://www.stereo-lab.de/tractrix.html
I use these ones
0303/03 2.0 " Hoerner, cf > 400 hz 425,00 € 1 Paar

I can really recommend them.

regards
Bengt

Ian Mackenzie
08-10-2005, 06:38 AM
Hi Jacob,

I will have to put in a vote for the horn type that I have found to be the most versatile and uncolored- the conical. A conical is a straight sided horn, and can be circular, square, rectangular, or some other shape. They offer the most uniform dispersion vs. frequency of any horn type, and do not sound like horns at all. The pattern is basically determined by the walls of the horn; I have found 50 degree horizontal by 35 degrees vertical to be very useful for hi fi listening. They lack low end loading due to the rapidly expanding throat, but this would not be a problem with a 650Hz. crossover point.

If you can, supersize them. My current horns are 30" deep with 24" by 32" mouths. You can go somewhat smaller with your crossover point; I run mine down to 300Hz. I have found that there is no substitute for large midrange horns to create "big goose bumps" vocal reproduction.

One nice thing about conicals is that they are easy to build. You could build an experimental pair in a few hours for evaluation, then build nicer ones if you like their sound.

Steve,

That could make a great diy project, can we indulge you, pay you...beg?

Ian

Harryup
08-10-2005, 06:57 AM
Hi Jacob,
They offer the most uniform dispersion vs. frequency of any horn type, and do not sound like horns at all.

Hi Steve,
I think I must make a strong reservation regarding "most" and " of any type".
Please study the background of Paul Voight's patent regarding the importance of the geometrics of the Tractrix curve. And the benefits just regarding dispersion. From DIY point of view the conical horn is extremely easy to build. We have made quite few comparison of the Tractrix horns with Altec and JBL original horns and som home made exponential, conical and even home made wodden Tractrix but the Stereo-Lab ones has sofar been superior with a rather large margin. Sofar we have just tested 2" on both Altec and JBL drivers but we are waiting for some 1" and some 7" to try out as well.

Regards
Bengt

Harryup
08-10-2005, 07:01 AM
Steve,
which drivers are you using down to 300Hz. A friend of mine has the Avantgarde Trio horns so I can easily agree upon that using a straight horn for the low midrange is very very tempting.


Regards
Bengt

aust-ted
08-13-2005, 04:02 AM
Hi Steve,
I think I must make a strong reservation regarding "most" and " of any type".
Please study the background of Paul Voight's patent regarding the importance of the geometrics of the Tractrix curve. And the benefits just regarding dispersion. From DIY point of view the conical horn is extremely easy to build. We have made quite few comparison of the Tractrix horns with Altec and JBL original horns and som home made exponential, conical and even home made wodden Tractrix but the Stereo-Lab ones has sofar been superior with a rather large margin. Sofar we have just tested 2" on both Altec and JBL drivers but we are waiting for some 1" and some 7" to try out as well.

Regards
Bengt

Bengt
Very interesting conclusions. Would be very interested in what ways do you consider the "Stereo-Lab ones has sofar been superior with a rather large margin"

There was an interesting discussion in the Audio Asylum High Efficiency speaker forum early this year which addressed differences between horn types. It was titled "Revisiting Horn Soundstaging & Imaging"was if you want to refer to it. Anyway the discussion did include comments on conical (waveguides) and tractrix. My understanding was that conicals had better imaging off axis but that the tractrix offered very good imaging at the sweetspot. If this correct, would it depend on how you want to use and listen to the horn before deciding which is best?

Now, I am yet to try either as I am currently using JBL 2380A CD horns and I am no expert on this but am keen to construct one or the others when time permits. As discussed previously, while conical horns seem easy to construct I gather it is important to get the throat design right. Earl Geddes recommends an oblate spheriod shape for that which is a squashed or stretched spherical shape.

Regards
Ted

Steve Schell
08-15-2005, 05:44 PM
Hi guys,

Great discussion. Bengt, you are correct that there is no "best"; all flare types have their strengths and weaknesses. Likewise, I try not to become too dogmatic and adopt one point of view to the exclusion of others (fail sometimes), as learning screeches to a halt at that point.

Conical horns do offer quite constant directivity vs. frequency, putting the same frequency balance into the reverberant field as into the on axis radiation. This is how most sounds propagate in nature, and is very natural sounding. Downside is that they must be larger for a given low frequency cutoff, due to the lack of loading from the rapidly expanding throat. Also, highs will be weaker on axis due to the wide dispersion characteristic. These aspects lead to a somewhat humped response curve on axis, which sometimes requires EQ to correct.

Tractrix horns sound very good and offer smooth sound in a shorter horn than an exponential, as the horn terminates at 90 degrees at the mouth (an exponential never does!). The tractrix will progressively beam high frequencies like the exponential due to the curved throat. This is not a problem if a tweeter is crossed in before the beaming becomes severe.

The drivers used on my conical mid/high frequency horns are the ones that my partner Rich and I have been developing. They are based on the 1930s RCA MI-1428B driver, which used a 4 1/2" center-suspended cone with 2" voice coil as a diaphragm. This driver format offers unique possibilities, as the loosely suspended cone resonates at 125Hz. and performs well down to 200Hz. We have also been building a scaled up version of the driver with a 6 1/2" cone. for use in bass horns. The first version of this driver used a 6:1 compression ratio and 3" diameter exit; currently we are working on some with a 4:1 compression ratio and 4" exit.

Ian, here is a drawing of some 2" throat conical horns I built for a friend. I generally use a throat plate made of two thicknesses of 3/4" plywood. The driver bolts to the rear one, then the assembly bolts to the other one which is permanently attached to the horn. The round to rectangular transition is accomplished through the 1 1/2" thickness.

aust-ted
08-16-2005, 05:51 AM
in part

"Downside is that they must be larger for a given low frequency cutoff, due to the lack of loading from the rapidly expanding throat. Also, highs will be weaker on axis due to the wide dispersion characteristic. These aspects lead to a somewhat humped response curve on axis, which sometimes requires EQ to correct. "

snip and

"The drivers used on my conical mid/high frequency horns are the ones that my partner Rich and I have been developing. They are based on the 1930s RCA MI-1428B driver, which used a 4 1/2" center-suspended cone with 2" voice coil as a diaphragm. This driver format offers unique possibilities, as the loosely suspended cone resonates at 125Hz. and performs well down to 200Hz. We have also been building a scaled up version of the driver with a 6 1/2" cone. for use in bass horns. The first version of this driver used a 6:1 compression ratio and 3" diameter exit; currently we are working on some with a 4:1 compression ratio and 4" exit. "

Ian, here is a drawing of some 2" throat conical horns I built for a friend. I generally use a throat plate made of two thicknesses of 3/4" plywood. The driver bolts to the rear one, then the assembly bolts to the other one which is permanently attached to the horn. The round to rectangular transition is accomplished through the 1 1/2" thickness.

Steve, thanks for the valuable insights here. I was thinking of constructing a wider dispersion conical horn but the horn characteristics you refer to above provide a sound basis for narrower dispersion angles. The uniform dispersion characteristics you mention plus the disadvantage of relative large size of a conical for a given cut-off freq make a design like the one you have kindly shown us a great idea provided one can live with the relatively long horn. Can you give us an idea of the low freq capability of this horn?

Your description of the RCA MI-1428B driver sound similar to the tweeter in my RCA LC 1Cs. Not surprising given the connection. How closely are they related?

Regards
Ted

Steve Schell
08-16-2005, 04:30 PM
Hi Ted,

My conical horn efforts to date have been cut and try. The dimensions of the horn in the drawing were determined by copying the length and mouth size of the RCA 12 cell horns I had used previously. When I built and installed the first horn in my system, the openness and clarity of the sound was so much better than the RCA in the other channel that I spent half the night building a second one.

The second design I built have a shorter 30" length and a 20" by 32" mouth. This reduced the vertical pattern from 38 degrees to 35 compared to the earlier horns, and increased the horizontal from 42 degrees to 50. I have found that the 35 by 50 pattern works well in a living room situation, allowing three side by side listeners to hear each horn well while sending minimum energy to the side walls, ceiling and floor. I did encounter a narrow peak at 7kHz. in the newer design, which I think is caused by a one wavelength delayed reflection off the top and bottom throat surfaces, which recombines on axis to create the peak. Placing small bits of felt in the throat top and bottom made the peak disappear.

Conical horn flares can be modeled with McBean's "Hornresp" program, which can be downloaded from: http://www.users.bigpond.com/dmcbean/
I'm not too swift at these simulation programs, but you may find this one useful. It would shed light on the size of a horn needed for a given cutoff, etc. I have found that longer is generally better when it comes to l.f. cutoff.

Other embellishments to a basic rectangular conical design might include: a bit of curvature in the initial throat section a la Geddes to encourage the formation of a spherical wavefront; a gently rounded mouth termination and/or break to a second, wider flare near the mouth to control reflections; filling in the corners to smooth the shape, perhaps going so far as to create an oval flare. There are many possibilities, and I'm building new horns to incorporate some of these ideas now.