It is not for changing my own setup, I am happy with that. It is for learning but 2.5m makes it nonsensical. Thanks
BTW, here's a comparison of two pretty different waveguides. One smaller (28 cm), moderately wide-radiating, the other larger (46 cm), moderately narrow (and narrowing). Notice the SP curves though (i.e. the total radiated powers) - except the absolute values, given by the different beamwidths, their slopes are virtually identical. This means that the "timbre" of the reverberant field in the same room will also be the same, even if the polars of the larger one fall more rapidly off axis. I found this a bit unintuitive at first but this is how it is (the far off-axis polars simply contribute less on the bigger horn, as they are of lower level).
Last edited:
Hi mabat,
Thank you for the information re-fit of the interface and all the comparisons.
Kindest regards,
M
No worries, I have collected so many compression drivers 🙂 , that when you post the other one, I can try to print it also, for comparison. As you noted, different people have different preferences. Then I sell the rest of the compression drivers.
Thank you for the information re-fit of the interface and all the comparisons.
Kindest regards,
M
When I was an assembler at Meyer Sound, I could see a resemblance between the Leopards I was assembling and Bill Waslo’s fade-to-array trick he first used on his Small Syns (explained here) and again on his 3D-printed co-entrant waveguides where he eventually moved the woofer array slots to the sides of the cabinet (here).
Here’s the Meyer Sound leopard arrangement:
I’m wondering if a pair woofers behind one of these Ath horns could be employed to do something similar without many deleterious effects.
Apologies if this is threadjacking.
Here’s the Meyer Sound leopard arrangement:
I’m wondering if a pair woofers behind one of these Ath horns could be employed to do something similar without many deleterious effects.
Apologies if this is threadjacking.
I've tried a few variations of that with several different waveguides but was not overwhelmingly happy with how it sounded. It's also more or less what's being done with the Meyer X80 as well. You can get the woofers pretty close but still need to use a low XO (700hz in my case) and there is some lobing that still occurs. Might work better with a coaxial compression driver but that doesn't hit the price mark I'm aiming for with this box.
Not sure where to post this so Mods please feel free to separate this into its own thread or even delete it if these questions have already been asked.
After reading/searching this thread, am I correct to assume that the 3d printed horns presented here from Marcel's software (e.g ATH-E520G2) are best suited for prototype/ measurement/ development purposes in the pursuit of horn design and not final versions suitable for critical listening in a finished speaker system?
Would a good approach to obtain a final hifi version of one of these horns be to cast a mould from the 3d print (e.g ATH-E520G2) and then make a final horn from perhaps a suitable material (e.g reinforced fiberglass or even paper mache ?) Am I wrong? Can you get a final version of the horn from a 3d print or will the quality be better from a mould or from a converted file to a CNC printer (this CNC option sounds difficult to do).
The reason I ask is that from an outsiders point of view (me) it looks like you can just send the ATH code that has been generated to an outside 3d printed service provider and get a fantastic hifi horn or failing that buy a 3d printer and optimize the variables ( material type, % infill, temperature ect....) till you get it right.
Yes I do realize I've grossly oversimplified what needs to be done to get a final version of one of these horns and perhaps this might be the subject for it's own thread? The horns and measurements presented in this thread are very impressive and don't do justice of the hard work that has been presented.
Thanks
FM
After reading/searching this thread, am I correct to assume that the 3d printed horns presented here from Marcel's software (e.g ATH-E520G2) are best suited for prototype/ measurement/ development purposes in the pursuit of horn design and not final versions suitable for critical listening in a finished speaker system?
Would a good approach to obtain a final hifi version of one of these horns be to cast a mould from the 3d print (e.g ATH-E520G2) and then make a final horn from perhaps a suitable material (e.g reinforced fiberglass or even paper mache ?) Am I wrong? Can you get a final version of the horn from a 3d print or will the quality be better from a mould or from a converted file to a CNC printer (this CNC option sounds difficult to do).
The reason I ask is that from an outsiders point of view (me) it looks like you can just send the ATH code that has been generated to an outside 3d printed service provider and get a fantastic hifi horn or failing that buy a 3d printer and optimize the variables ( material type, % infill, temperature ect....) till you get it right.
Yes I do realize I've grossly oversimplified what needs to be done to get a final version of one of these horns and perhaps this might be the subject for it's own thread? The horns and measurements presented in this thread are very impressive and don't do justice of the hard work that has been presented.
Thanks
FM
The Leopard's slot loaded woofers require extensive processing to correct what you may consider "deleterious effects".
The Leopard high frequency driver uses four exponentially expanding waveguides coupled to a diffraction grate, allowing a lower crossover frequency than a typical Ath style horn of the same depth.
The center to center side slot spacing of the Leopard would cause off axis cancellation if the crossover frequency was raised.
The 3d printed horns are suitable for critical listening in a finished speaker system.
The details required for the aesthetic finish desired can be accomplished in a number of ways, and as TNT implies, some have been covered in this thread.
There are hundreds of sources available that go into "how to finish":
https://www.fictiv.com/articles/how-to-finish-3d-printed-parts
https://all3dp.com/2/pla-smoothing-a-beginner-s-guide/
I guess I had a similar question. Is there a certain amount of infill and external perimeter lines one should printing with to allow the horn to be inert enough for final use?
Certainly 😉.
The amount of infill and external framework required depends on span dimensions, shape, SPL requirements and material used.
Some experiments using A600G2 MEH and 13cm low sensitivity midbass.
I think that the printed horns are usable for critical listening, I wouldn't offer the kits if I thought otherwise. It may just require quite an amount of manual work, and that can be tedious, depending on how much you like it 🙂
My approach would be to print the petal parts as light as possible, i.e. with only a very sparse infill, like maybe 5% (depends on the infill type). The mounting flange and the throat area can of course be heavier, even solid. If you print the parts yourself, you have complete control over these things. Then I would apply epoxy coatings.
The filament used affects mainly how fragile or sturdy, and stable the final product is, more than anything else, IMHO. For home/indoor use (which is the only target for all of my horns) I don't think this is critical. Only beware that raw PLA doesn't endure direct sunlight, I know that very well.
Last edited:
I'm curious about the acoustic properties of a nearly hollow plastic petal. Specifically, does it resonate or behave differently compared to aluminum waveguides or solid ABS with 100% infill?
As an experiment, I 3D printed one A520G petal from PLA with 0% infill (hollow) and filled it with acrylic mass. Afterward, when tapped, it produced a completely dead "thud" sound. There was a good deal on acrylic mass at a nearby hardware store, so this approach would not be too costly, one tube of 400gr was 1,99€. However, the total weight of the petals might become an issue—each A520G will then weight around 9 kg. I'm not sure how to handle the assembly or stand design to support a front-heavy waveguide if I continue down this path.
Has anyone else experimented with different infill percentages or materials for petals or waveguides, and how did those changes affect the sound characteristics?
As an experiment, I 3D printed one A520G petal from PLA with 0% infill (hollow) and filled it with acrylic mass. Afterward, when tapped, it produced a completely dead "thud" sound. There was a good deal on acrylic mass at a nearby hardware store, so this approach would not be too costly, one tube of 400gr was 1,99€. However, the total weight of the petals might become an issue—each A520G will then weight around 9 kg. I'm not sure how to handle the assembly or stand design to support a front-heavy waveguide if I continue down this path.
Has anyone else experimented with different infill percentages or materials for petals or waveguides, and how did those changes affect the sound characteristics?
Better just print with hard TPU >50D, it's cheap 2 euro kg in pellets (regranulate), making filament cost 4-5 eur/kg. You can make as thick as you want.
For smaller waveguides (<300mm mouth) I make the thinnest part of the walls 3mm thick, adding extra material around the throat and flanges where pieces get joined together if necessary. I do 40% infill with 5 wall layers, so the thin wall sections are effectively solid.
For this particular waveguide, I tested before and after adding dynamat car sound dampening material. I wasn't able to see any measurable difference in distortion or frequency response "smoothness".
In my case it hasn't been any serious systematic analysis, but what I've found is that a thicker petal can be nearly (not completely) hollow and it seems to work just fine, it resonates certainly much less than an aluminum horn, which I would never use. The petals of my latest kits are roughly 1" thick and after adding even a sparse net of internal braces (e.g. a cubic infill) this seems to work just fine. And I'm sure there are better materials than PLA. That's just what I use the most, because it simply works, it's cheap and easy to work with.
I've never tried solid petals - way too long printing time and it would be heavy, which is not what I want. With very light (free standing) waveguides there is no problem mounting them simply as cantilevers - that's one of the reasons for me. If it was heavy, a different mounting would be probably needed, adding a mouth support, etc.
@galucha - very nice.
Last edited:
Thanks. One of the nice things about this division is that almost all surfaces of the joints are planar, i.e. it's very easy to sand them.