Hybrid Trachorn

[soundbound]

Your horns look top notch. Beautiful work. I like the looks of that Cornscala too. I'd really like to own a set of those Cornscalas. They must sound good.

[idahodewitt]

Where does one find the pattern to cut the horn's flare to the proper proportions? This looks like a project I'd be into tackling, but I'd want to make sure I'm doing it right.

L.D.

[Guest David H]

http://www.volvotreter.de/dl-section.htm

www.alkeng.com has a few links and great info, just click the Klipsch speakers link.

[capo72]

GotHover,

I'm curious about the process you used to layout the flare. Did you plot the points by hand or use some other method? If you followed the link I posted eariler in this thread you can see how I plotted the points, from the volvotreter spreadsheet, when I built mine a little over a year ago. I have since gained access to Autocad software and i'm able to plot the curve from the spreadsheet via a script file. It now takes about 5 minutes to do what previously took me about 2 days to do. The Autocad plot should be much more accurate, but I don't like the way it looks.

When I plotted by hand the flare only gets bigger as you go away from the throat toward the mouth. But with the Autocad plot the flare actually gets smaller than the throat diameter before it begins to get larger as it flares out to the mouth. I have tried different mouth ratios and throat sizes and the Autocad plot always shows that initial decrease in cross section area before getting larger and flaring out. If I undrestand the Tractrix science right, then this would be increasing the compression on the driver there by changing the Fc.

I may try building one that way and testing for actual response to see if there is any difference between to 2 plotting styles. I would like to know your, or anyone with experience in this area, thoughts.

Jeremy

[tigerwoodKhorns]

Jeremy,

If it is supposed to always flare and in AutoCAD it is not flaring my guess is that the equation is causing the initial compression. Check the points that are generated. You shoudl be able to follow the trend to see where the compression is comming from.

[Guest David H]

GotHover,

When I plotted by hand the flare only gets bigger as you go away from the throat toward the mouth. But with the Autocad plot the flare actually gets smaller than the throat diameter before it begins to get larger as it flares out to the mouth. I have tried different mouth ratios and throat sizes and the Autocad plot always shows that initial decrease in cross section area before getting larger and flaring out. If I undrestand the Tractrix science right, then this would be increasing the compression on the driver there by changing the Fc.

Jeremy

Jeremy my horn also decreases size just slightly before expansion. Cormpression was also my assumption.

[Marvel]

The tractrix horns sold by Al Klappenberger also get slightly narrower before the flare expands. He talked about it in one of the threads, because he had tought of having them molded. That would have brought up other issues with the mold parts.

Bruce

[Guest David H]

I have the oportunity to purchase a cnc router to make horns baffles etc, if there is any interest for having more horns built. Here is a pic of the router.

[WMcD]

Of course I can't speak for Al K.

However, some years ago I was designing and building exponential midranges. They were based on the technique used by Bruce Edgar. That is, with planar tops and bottoms as we are discussing here.

The issue is that the top and bottom walls diverge (are offset) from the center axis at an established linear rate. Then you have to take the exponental area expansion and back figure the distance (offset) from the center axis to the curved side walls.

The problem is that near the throat, the top and bottom walls are diverging so quickly (relatively) in the vertical that the side walls have to narrow in the horizontal to honor the exponential area goal near the throat.

It is interesting that if you want an exponential area, and can curve all four walls, then the offsets of the walls from the center are themselves exponential. This results in a square throat growing to a rectangular mouth (same with a square mouth with different numbers). The power coefficients for side walls, when added, sum to the exponential area coefficient . . . this is because, like we learned in high school, if we multiply exponential functions (to get area), the equivalent is adding the exponents. [Granted, only of interest to the math inclined, but I was happy to see that I was on the right track.]

= = = = = =

I get a little uncomfortable with the tractix. This is simply because the tractix is not, in a pure sense, and area calcualtion. Rather, it is based on the side wall location and symetric (round and bastardized to square) cross section. Nonetheless, people calculate the resulting area. Then we proceed as above with the exponential.

The same issue probably comes up. The linear top and bottom walls expand quickly and therefore, near the throat, the horizontal must be narrowed.

= = = =

In my fooling around with numbers, I realized that this is a common problem. Generally you have the "bell" of the horn with a desired geometry but getting it to work at the throat and mate to a round driver exit is a bear. The K-5 patent uses parallel top and bottom walls and straight divergent side walls with vanes to honor the exponential, and thereby get to a retangular geometry which is suitable for the bell. Note the K-5 uses straight sidewalls, somewhat like a radial sectoral, whereas the bell we are talking about uses straight top and bottom.

On the other hand, we can look at Don Keele's description of his constant directivity horns. There we see that he had a similar problem at the throat. He had to narrow down the vertical axis to obtain an exponential/hyperbolic area until reaching the bell of the horn.

And if we look at the tractrix horn in the Forte II and the like, there is a matching section near the throat of round cross section which feeds the pyramidal bell.

Best,

Wm McD

[greg928gts]

I have the oportunity to purchase a cnc router to make horns baffles etc, if there is any interest for having more horns built. Here is a pic of the router.

Will that machine cut the increasing side angles of the top and bottom pieces of a wooden horn?

Greg

[Marvel]

The problem is that near the throat, the top and bottom walls are diverging so quickly (relatively) in the vertical that the side walls have to narrow in the horizontal to honor the exponential area goal near the throat.

Never had it explained,Gill, but that makes so much sense. Thanks!

Bruce

[Guest David H]

Greg the Debece is a 3axis 3D router so it will cut the increasing angles, but will require several passes and will be stepped.

[Guest David H]

Of course I can't speak for Al K.

However, some years ago I was designing and building exponential midranges. They were based on the technique used by Bruce Edgar. That is, with planar tops and bottoms as we are discussing here.

The issue is that the top and bottom walls diverge (are offset) from the center axis at an established linear rate. Then you have to take the exponental area expansion and back figure the distance (offset) from the center axis to the curved side walls.

The problem is that near the throat, the top and bottom walls are diverging so quickly (relatively) in the vertical that the side walls have to narrow in the horizontal to honor the exponential area goal near the throat.

It is interesting that if you want an exponential area, and can curve all four walls, then the offsets of the walls from the center are themselves exponential. This results in a square throat growing to a rectangular mouth (same with a square mouth with different numbers). The power coefficients for side walls, when added, sum to the exponential area coefficient . . . this is because, like we learned in high school, if we multiply exponential functions (to get area), the equivalent is adding the exponents. [Granted, only of interest to the math inclined, but I was happy to see that I was on the right track.]

= = = = = =

I get a little uncomfortable with the tractix. This is simply because the tractix is not, in a pure sense, and area calcualtion. Rather, it is based on the side wall location and symetric (round and bastardized to square) cross section. Nonetheless, people calculate the resulting area. Then we proceed as above with the exponential.

The same issue probably comes up. The linear top and bottom walls expand quickly and therefore, near the throat, the horizontal must be narrowed.

= = = =

In my fooling around with numbers, I realized that this is a common problem. Generally you have the "bell" of the horn with a desired geometry but getting it to work at the throat and mate to a round driver exit is a bear. The K-5 patent uses parallel top and bottom walls and straight divergent side walls with vanes to honor the exponential, and thereby get to a retangular geometry which is suitable for the bell. Note the K-5 uses straight sidewalls, somewhat like a radial sectoral, whereas the bell we are talking about uses straight top and bottom.

On the other hand, we can look at Don Keele's description of his constant directivity horns. There we see that he had a similar problem at the throat. He had to narrow down the vertical axis to obtain an exponential/hyperbolic area until reaching the bell of the horn.

And if we look at the tractrix horn in the Forte II and the like, there is a matching section near the throat of round cross section which feeds the pyramidal bell.

Best,

Wm McD

It's true that rectangular horns are a compromise, but overall the trachorn sounds good.

I think Bob Crites might add...If the math is bad and the sound is good, you might be an audiophile..... Or a redneck.

[DrWho]

The area expansion rate is what determines the acoustical loading. The shape of the side walls determines the polar response.

So if you're "pinching" in one dimension and expanding in the other, but the rate of area expansion is always positive, then there is no compression happening. The power transfer will be the same, however, the focusing of that power can be different.

Life gets complicated when you consider that the "focusing" is going to be different at different frequencies.

[whatever55]

Has anyone designed one to be used on a Jubilee bass bin.... as good as the 402 I am hearing people want to try a very expensive driver to minimize the EQ needed.... if I read correctly.

So would one of these Trachorns be the answer.

I have a pair from Al on my Belles and it was the best money I ever spent.

[Guest David H]

Has anyone designed one to be used on a Jubilee bass bin.... as good as the 402 I am hearing people want to try a very expensive driver to minimize the EQ needed.... if I read correctly.

So would one of these Trachorns be the answer.

I have a pair from Al on my Belles and it was the best money I ever spent.

Dont know how these would work on a Jub I have used them with KHorn LaScala and DIY CornScala.

[PrestonTom]

Has anyone designed one to be used on a Jubilee bass bin.... as good as the 402 I am hearing people want to try a very expensive driver to minimize the EQ needed.... if I read correctly.

So would one of these Trachorns be the answer.

I have a pair from Al on my Belles and it was the best money I ever spent.

I am glad that you are happy with yours on the Belles.

However, it would not be a replacement for the K402 ( I am referring to Al's version of the trachorn, which apparently you have on your Belles).

Although the K402 is a tractrix, it is modified so that its polar response is CD (the spectrum off-axis is approximately the same as on-axis). This is where the CD equalization comes in (this is not an unusual requirement and it does not denote a poor horn design). The 402 is also a 2inch throat, while the trachorns are smaller. I doubt you could use the trachorn in 2-way set up, but this is just my guess.

I don't think anyone has published the polar patterns or the distortion measures of the trachorn (with a favorite driver), so there are still a number of unknowns about its measured performance. The k-402 horn has the above advantages (CD, low distortion, 2inch throat, and a relatively, narrow & controlled dispersion) and does not necessarily need to be used with "very expensive" drivers. One can use the Klipsch driver, a B&C driver, or a P.Audio driver or an expensive TAD driver. There are a number of choices.

These are two different horns with two different sets of strengths and weaknesses.

Good Luck,

-Tom

[DrWho]

as good as the 402 I am hearing people want to try a very expensive driver to minimize the EQ needed.... if I read correctly.

EQ is needed in both cases...either by collapsing polars or by electronic compensation (either active or passive).

I might suggest looking into how the polar response ties in with room acoustics and how it affects what we hear. In fact, that is one of the problems I have with these simple tractrix calculators. There is no optimization of coupling the horn to the driver and no way to adjust for a desired polar response. It doesn't take too much listening to realize just how much of an influence these items can have on the perceived performance.

[mikebse2a3]

Has anyone designed one to be used on a Jubilee bass bin.... as good as the 402 I am hearing people want to try a very expensive driver to minimize the EQ needed.... if I read correctly.

So would one of these Trachorns be the answer.

I have a pair from Al on my Belles and it was the best money I ever spent.

Hey whatever55

I'm one of the Jub owners with the TAD TD4002 installed on the K402. I can assure you in my case I didn't do this because of EQ issues but instead it was because I believe the performance of the K402 is exceptional and any advanced driver designs should perform extremely well on it. In other words IMHO the K402 offers an absolutely excellent platform for such investigation of different HF compression driver's designs and this is why I'm trying the TAD TD4002. The TAD drivers have design features such as a 100mm Beryllium diaphragm and surround and other design features which have been reported to offer very high performance. Like everything there is a cost/performance factor that sets in and each person has to decide that value for themselves. The $$$$ extra it cost to go with the TAD/K402 combination is something that falls under the cost/performance debate. For myself with what I value in music reproduction the TAD TD4002 was worth it and the K402 is IMHO for many reasons the best horn available to use with it based on a desire for a 2-way system, with a flat on and off axis frequency response over a very wide bandwidth, and Polar Control angles that offer very good room integration and very importantly are maintained over a very wide bandwidth.

mike tn

[cjgeraci]

Has anyone designed one to be used on a Jubilee bass bin.... as good as the 402 I am hearing people want to try a very expensive driver to minimize the EQ needed.... if I read correctly.

So would one of these Trachorns be the answer.

I have a pair from Al on my Belles and it was the best money I ever spent.

About a year ago, I came to a crossroads. Try Al's Trachorn with other drivers or go to a two inch throat such as with the 402s.

I chose the former.

Some doubt whether Al's Trachorn could be used in a two-way application.

Actually, answer is "yes."

I've used a pair of Al's Trachorns in a two-way configuration for about a year now - with Altec 902s. No tweeters needed.

The Trachorns do not beam at upper frequencies, despite what some might think. The horns sound quite a bit better than my other two-way horns in the house, 511s.

Since Jub bass bins naturally do 500 hz easily, Al's Trachorns with 902s would be an easy top end for you - without much additional expenditure.

Whether or not one would prefer Trachorn/902s over various K402/driver combos (or vice versa), I will leave others to decide with their own ears.

Good luck in your quest.

Carl.