Well i wondered since i contructed my tractrix horn wrong, what the difference would be between a correct one and as well a plain old open baffle of the same size.
This so happened to perfectly supplement my current reading up on horn design, almost text book perfect examples actually; The wide flare is needed to achieve LF performance. The end plates in the radial horn, while limiting dispersion also do a good job of preventing the air from escaping the pressure wave. You could of course go for a compound horn, with all the pitfalls that brings to the table. Interesting to see that the horn indeed do add linearization here - as so often claimed for Tractrix. Further horn/driver impedance matching could probably smooth the response even more - compression driver loading the diaphragm. Thanks a lot :-) incredible timing!
thats good to know, my thoughts where that the wide cheeks also gave a direct vision of the tweeter itself, so no obstructing like the traxtrix. screwing up the top end at 30 degree compared to the screwed up horn :)
A tractrix horn (and parabolic horns) will behave very much like a conical horn as it reaches the mouth and the curve becomes similar to a line. The reason you see the incorrectly oriented horn behave like it does is because it behave more like a conical horn at the throat. The big dip at 22KHz looks like baffle step, and you can see how the tractrix flare handles it much better. A conical horn will perform better with a circular or parabolic curve at both the mouth and throat. You can also see clearly the major weakness of the tractrix design, it tends to have poor off axis response compared to conical or parabolic horns.
No...this was very insightful for viewers and 6-7 db is huge on the low end. Most designs I've seen using horns are for individuals looking to crossover at 1000 hz with a larger woofer 10-12 inch. Now you can see why! :>)
@@joppepeelen I don't know why one viewer stated that your design was wrong but I have seen some larger throat horns like your first one that you said was a mistake. The best one is the one that works best with whatever woofer that you choose to mate it with!??!?!
@@gzubeck3 well a tractrix has a certain slope./flare and mine did not look like one although i did use a tractrix designer, thats why he said the flare was rotated 90 degrees , and hes right about that. this works to in the end , but if i wanted to make a tractrix the second version should be it. not the old one :) i just interpreted the drawing wrong :) since the program i used only poops out a curve. and i used that curve 90c rotated.
Interesting how the tractrix horn had a smoother mid range frequency response. This might be a total waste of time, but I would be curious if an ESL type driver would work in this horn.
as far as i know it should not. many times before people tried an i beleive the awnser was nope :) but i could be wrong , its just what i remember (witch indeed might not say that much haha) i believe it damps it. so the benefit is kind of gone :(
None of the expected pcb layout cad programs. So i just use sketchup. It works the fastest and the best for me. Butit cant adjust trace width onn the fly .... it means start all over
The original horn/waveguide reached lower because it was wider. Be interesting to see the results with a three dimensional horn vs your two dimensional horn.
Interesting results. It would be good to see single line averaged from 60/45/0. Your original horn design looks the best to me. You called it the Guy Horn?
This so happened to perfectly supplement my current reading up on horn design, almost text book perfect examples actually; The wide flare is needed to achieve LF performance. The end plates in the radial horn, while limiting dispersion also do a good job of preventing the air from escaping the pressure wave. You could of course go for a compound horn, with all the pitfalls that brings to the table. Interesting to see that the horn indeed do add linearization here - as so often claimed for Tractrix. Further horn/driver impedance matching could probably smooth the response even more - compression driver loading the diaphragm. Thanks a lot :-) incredible timing!
thats good to know, my thoughts where that the wide cheeks also gave a direct vision of the tweeter itself, so no obstructing like the traxtrix. screwing up the top end at 30 degree compared to the screwed up horn :)
Красота. Какой размер приводных винтов на вашем станке?
A tractrix horn (and parabolic horns) will behave very much like a conical horn as it reaches the mouth and the curve becomes similar to a line.
The reason you see the incorrectly oriented horn behave like it does is because it behave more like a conical horn at the throat.
The big dip at 22KHz looks like baffle step, and you can see how the tractrix flare handles it much better. A conical horn will perform better with a circular or parabolic curve at both the mouth and throat.
You can also see clearly the major weakness of the tractrix design, it tends to have poor off axis response compared to conical or parabolic horns.
Thank for the explanation ! im a horn noob :) so everything is welcome haha
No...this was very insightful for viewers and 6-7 db is huge on the low end. Most designs I've seen using horns are for individuals looking to crossover at 1000 hz with a larger woofer 10-12 inch. Now you can see why! :>)
Tbanks ! witch one you prefer ?
@@joppepeelen I don't know why one viewer stated that your design was wrong but I have seen some larger throat horns like your first one that you said was a mistake. The best one is the one that works best with whatever woofer that you choose to mate it with!??!?!
@@gzubeck3 well a tractrix has a certain slope./flare and mine did not look like one although i did use a tractrix designer, thats why he said the flare was rotated 90 degrees , and hes right about that. this works to in the end , but if i wanted to make a tractrix the second version should be it. not the old one :) i just interpreted the drawing wrong :) since the program i used only poops out a curve. and i used that curve 90c rotated.
This is very very interesting.
Interesting how the tractrix horn had a smoother mid range frequency response.
This might be a total waste of time, but I would be curious if an ESL type driver would work in this horn.
as far as i know it should not. many times before people tried an i beleive the awnser was nope :) but i could be wrong , its just what i remember (witch indeed might not say that much haha) i believe it damps it. so the benefit is kind of gone :(
Very interesting what software did you find works best for modeling the profile?
None of the expected pcb layout cad programs. So i just use sketchup. It works the fastest and the best for me. Butit cant adjust trace width onn the fly .... it means start all over
@@joppepeelen very cool thanks!
I think Tractrix has less low end, because it is small.
You could scale it up so it would be same width as "screwed up" horn.
well its not that much either. but dispersion ... is kind of a thing the screwed up version did much better :)
@@joppepeelen Horns usually reduce dispersion. It can be a good thing (less reflections).
I don't know how bigger tractrix would affect dispersion.
@@lightbit553 if the curve would be less steep i think it becomes better just like the screwed up horn :)
The original horn/waveguide reached lower because it was wider. Be interesting to see the results with a three dimensional horn vs your two dimensional horn.
Nicely done sir! Xx
Interesting results. It would be good to see single line averaged from 60/45/0. Your original horn design looks the best to me. You called it the Guy Horn?
called it my screwed up horn :)