I have designed a 135hz tractrix horn in hornresp. this is going to be a rectangular so when I am exporting the horn data I am choosing “conical” width flare.
Is it better to have the conical part on the horizontal or the vertical axis? This horn is going to be on the ground.
It seems as though you can only select CON for the width flare not the height, but I could always turn the horn on its side.
Is it better to have the conical part on the horizontal or the vertical axis? This horn is going to be on the ground.
It seems as though you can only select CON for the width flare not the height, but I could always turn the horn on its side.
Is the beam width the same on both axis?
How high will you run it? (1k?)
What is the directivity of the unit that takes over (at 1k?) from this horn?
//
How high will you run it? (1k?)
What is the directivity of the unit that takes over (at 1k?) from this horn?
//
If your rectangular horn sits on the floor (2Pi space), and you want the horn top and bottom flat. Then you can manually set the height S1 = S2, (in export data window) and Hornresp will re-calculate the width profile to maintain flare rate. This also applies if you want to switch the horn's orientation (adjust height or width and Hornresp re-calcs to maintain correct area). Depending on your design you may need Exp profile otherwise you could see a dip in the preview profile.
When you make a loading type horn, the directivity can seem somewhat arbitrary, or at the least something you need to be careful to select and hence to compromise. What is the point of you choosing tractrix? It's difficult to say which is better.
On the other hand, the conical side may be more effective at controlling lower frequencies. In your original configuration, the verticals would let go sooner.
On the other hand, the conical side may be more effective at controlling lower frequencies. In your original configuration, the verticals would let go sooner.
I think you shall optimise so that the horisontal directivity is as equal as possible at X-over.
//
//
It would be, the horn rounds out well and loses low frequency control at the same time. You have to make that choice.
Correct.
As you said in your first post, if the height rather than the width is to have the conical profile simply assume that the horn is rotated around its axis by 90 degrees.
Edgars midbass horn is well regarded, that has the straight parts top and bottom, with the sides being curved.
I've tried my versions both ways and didn't notice much (if any) difference by listening. Maybe measurements may show something.
Rob.
I've tried my versions both ways and didn't notice much (if any) difference by listening. Maybe measurements may show something.
Rob.
This appears to be one reason the Yuichi was given vanes. Not that it has to be a problem.. or the waveguiding aspects of the device can be considered separately.
If your horn sits on the floor, then the floor will act like a reflector for lower freq (3dB gain, 2Pi space). This "mirror" can be viewed as a phantom speaker image, making the horn appear 2x vertical size, so its vertical directivity improves. You need something like a BEM sim to see this effect. Starting at some higher freq the reflection starts causing an interference null (floor bounce). All of these effects depends on the height of the driver(s) and horn dimensions.
So I would put most of the expansion (or flare) in the horizontal.
This is the "pinching" that I referred to earlier, You can manually tweak the S1 or S2 height or width to minimize pinching. Could you post or PM the Hornresp design file?
@DonVK I was able to adjust the dimensions on the export screen to minimize the pinching down to just a couple of mm and still get the shape that I want. I think that is what you meant? It seems that the narrower I made the conical angle the less pinching there was on the exponential flare.
