Mylar would be better.
This looks like it, but it's the 3" brother:
Aftermarket alternative (1.75"):
Last edited:
Yes sir !!!
Your current horn project graphs look awesome
If I may humbly suggest...why stop at 500Hz?
Why not try to take "full-bodied" reproduction as low as we can? Make it even more "full-bodied"?
In my mind, this is the reason behind a unity/synergy/MEH.... take the idea of "full-bodied" reproduction as low as possible.
To date, that unity/syn idea has been compromised with the need to make mounting surfaces that work for larger drivers on horns that have less than the excellence your horns, OS horns etc., have been able to achieve for the top half of the audio spectrum.
Horns like yours and OS nail the top half of the spectrum imo...and you've got it reaching into the upper end of the bottom half of the spectrum now, at 500Hz.
I have horns that can get my CD's to reach to below 500Hz, (with good response, but they are surely not as refined as your postings).
My experience so far, is full spectrum integration ala unity/syns has beared more fruit than mainly trying to optimize the top half.
But If you guys can figure out how to integrate your exemplary designs with additional drivers below the CD, and integrate the lower half of the spectrum into a single horn....
i say go man go !!!!!! And go some mo' !!! I'll do my best to build such.
Yep, and the easy part is looking really good !
I figure lower frequency port size is mainly about what is the SPL design goal. Which of course also begs, what size drivers are being used.....
I try to have all driver sections including the CD, run out of steam at near the same SPL. So pretty high SPL designs (for the sake of maximizing normal listening level sound quality)
Need to get all that out, to explain my perspective....
Anyway, using a couple of 10"s in a build which can reach straight to CD at 500Hz, and was designed to cross to a sub at 100Hz....
each of the 10" gets a port with around 40 cm^2 area. About an 8:1 ratio, driver Sd to port area.
https://www.diyaudio.com/community/threads/syn-10.383607/#post-6963151
I think smaller drivers, if they can reach to 100Hz at desired max SPL, could probably use a higher Sdort ratio.....further minimizing port areas.
But port noise at SPL would need to be tested.
Minimizing port length (thru the horn walls) is supposed to help increase flow.
It's one reason I've always had a hard time picturing how to mate lower frequency cones onto anything other than a flat horn wall.
Seems like ports would have to be longer ducts (than just thru a flat horn wall) to feed into curved horns, and therefore need greater cross sectional area to accommodate equal flow. Seems to work against minimizing port areas and potential horn disruptions.
I figure lower frequency port size is mainly about what is the SPL design goal. Which of course also begs, what size drivers are being used.....
I try to have all driver sections including the CD, run out of steam at near the same SPL. So pretty high SPL designs (for the sake of maximizing normal listening level sound quality)
Need to get all that out, to explain my perspective....
Anyway, using a couple of 10"s in a build which can reach straight to CD at 500Hz, and was designed to cross to a sub at 100Hz....
each of the 10" gets a port with around 40 cm^2 area. About an 8:1 ratio, driver Sd to port area.
https://www.diyaudio.com/community/threads/syn-10.383607/#post-6963151
I think smaller drivers, if they can reach to 100Hz at desired max SPL, could probably use a higher Sd
ort ratio.....further minimizing port areas. But port noise at SPL would need to be tested.
Minimizing port length (thru the horn walls) is supposed to help increase flow.
It's one reason I've always had a hard time picturing how to mate lower frequency cones onto anything other than a flat horn wall.
Seems like ports would have to be longer ducts (than just thru a flat horn wall) to feed into curved horns, and therefore need greater cross sectional area to accommodate equal flow. Seems to work against minimizing port areas and potential horn disruptions.
By the easy part I meant the integration of the ports. At 500 Hz the 1/4WL is roughly 17 cm (6.8"), that should be comfortable enough to work with. With the hexagonal approach the walls are virtually flat after a few centimeters. I can easily imagine adding three 16-ohm drivers to an existing design, could be 8" I guess, maybe smaller.
Last edited:
I haven't started yet. Too much to be done before it's too cold outside.
- I want to finish the big concrete horn this autumn, at last, I have it almost ready to be casted.
You’re assuming right and you are exactly right on what the plan is for the future.
There is melamine foam in the whole back enclosure, but nothing between the cone and the opening on the sides. So the space marked in red is next to be experimented on.
Attachments
Gotcha.
From looking at the hex horn's picts, I thought there was more curvature than you are saying, that would still be hard to mount to.
The hexagonal shape allowing drivers is very interesting. I really hope someone gives this a try.
I use Really from Kvadrat for the resistive ports.
It's pretty flow resistive 70% fabric and 30% polyester compressed material and does a perfect job.
https://www.kvadrat.dk/en/really/solutions/textile-felt
A good exercise is to do a close mic measurement of what comes out of the ports.
You only want sub 1000 Hz frequencies, since the horns takes care of directing everything above that.
When high frequencies come out of the ports they mix up with what comes out of the horn and create the pattern you see.
It's quite nice already!
Just find something with more flow resistivity and put it there against the holes. Make sure air doesn't go around the material, but through it.
Is it possible to use ATH to simulate a fully "source defined" axisymmetric enclosure? I've drawn a spherical enclosure for a coax driver in CAD but exporting it to Akabak for each iteration feels slow. Simming a quadrant seems to work though I can't seem to input arcs in the rear correctly ("control points are not cocircular" no matter what I do) but axisymmetric would of course be nicer
It's possible to add a user-defined rear spline (cubic Bézier) to a waveguide (with or without a user-defined source, that doesn't matter):
https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7228432
If it's axisymmetric, I'd suggest to use an axisymmetric model as well, it will be much faster.
https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7228432
If it's axisymmetric, I'd suggest to use an axisymmetric model as well, it will be much faster.
Problem is, I also have the waveguide user-defined. I can match a small R-OSSE and a Bezier to the coaxial profile but it's kinda burdensome and I keep failing to make the Bezier work from the Desmos script anyway...
It seems I got axisymmetry to work though with the whole thing user-defined, and meshing problems I had are gone. The enclosure in ABEC however looks pretty rough to me and there doesn't seem to be a way to control the resolution. Not sure if that's screwing up the results or if I'm missing something else...
It seems I got axisymmetry to work though with the whole thing user-defined, and meshing problems I had are gone. The enclosure in ABEC however looks pretty rough to me and there doesn't seem to be a way to control the resolution. Not sure if that's screwing up the results or if I'm missing something else...
In my experience, resolution is normally at the end of the line. For example, 50 in the following line.
point P0 114.3 0 50
You can also control resolution in the mesh statements. (These might not be used in CircSym, I'm not sure.)
; -------------------------------------------------------
; Mesh Settings
; -------------------------------------------------------
Mesh.LengthSegments = 20
Mesh.AngularSegments = 20
Mesh.ThroatResolution = 64
Mesh.MouthResolution = 64
Mesh.SubdomainSlices =
Mesh.WallThickness = 2
Your simulations might be squigly due to your ABEC settings rather than resolution.
; -------------------------------------------------------
; ABEC Project Settings
; -------------------------------------------------------
ABEC.SimType = 2 ; 2= Free Space, 1= Infinite Baffle
ABEC.SimProfile = 0 ; 0 = CircSym
ABEC.MeshFrequency = 33000
ABEC.NumFrequencies = 100
ABEC.f1 = 500 ; [Hz]
ABEC.f2 = 20000 ; [Hz]
I don't always use high resolution for running simulations. I might only increase it to create the STL file for printing. If you post your config file someone could probably show you what to try.
Last edited: