New member, starting sub design.

Jump to Latest
S

sellurstuff

Member
Joined 2022
#1
Hi there,

I'm interested in understanding some of the fundamentals of loudspeaker design. I thought the best way to understand would be to start on a design and see what problems I encounter and what areas I need to do a bit more research in. I have a little bit of familiarity with the mathematics of acoustics as well as T&S parameters and other fundamentals of speaker design. I'm particularly interesting in horn design, having been reading about the paradise garage and the 'Levan horn' that was used as a sub in that club. I'd like to design and simulate a club subwoofer. I was wondering if there were any recommendations for steps in the design process or walkthrough type guides, as well as software for simulating the design?

Thanks
 
#3
planet10 said:
A bass horn wil be large, and is not trivial to design. Perhaps something less ambitious to start?

What kit do you have for support.

dave
Click to expand...
Do you mean kit for the actual build? Or design?

I have access to a wood workshop for the building stage when it comes to that. However I'm more interested in design and the mathematics at the moment. I wanted to try work on a few different designs and simulations before actually settling on something to build. Large isn't an issue really, I have a big enough space to work with a large design and I'm aiming to end up with something that could function for party type set up. But yeah, the physics is my real interest at the moment, and horn design seems the most interesting to me acoustically. Can I ask what the main difficulties are when approaching this type of design?

Thanks
 
#4
Design is tricky, build will likely be tricky.

A good primer on horn design is Dinsdale’s Woreless World series. http://p10hifi.net/planet10/TLS/downloads/Dinsdale-Horns-3parts.pdf

But one really needs to actually model the horn, the math is trickier than he shows. MJK is likely the best modeler but MIA. HornResp is popular and constantly evolving, Alabak is probably capable as well.

Note that the mouth size circumference should be the wavelegth of the lowest frequencies you want to reproduce. One can use the floor, the floor/wall, or a corner to reduce the most size by ½. ¼. ⅛ approximately. In most rooms, diven the relative size of the horn and the room one almost always ends up with wall/floor or corner loading. The K-Horn is an example of a speaker that requires a good corner to fully load the horn.

dave
 
#5
Yeah, so I'll have a lot of help with the build side of things further down the line but the design is something I'm taking on myself. I was aware a bass horn is at the more difficult end of the design spectrum but thought I'd venture down that path and see what difficulties I approached! Thanks for the link working my way through that paper now, really nicely written.

So the mouth size circumference should be the wavelength of lowest frequency. The research I've come across seems to suggest the horn length should also match the lowest frequency. Hence where folding helps. So is it the case that circumference and length should match? I'll be working in 1/4 space. Does that division effect length as well as circumference?

I've heard the MJK modeler mentioned on other forums. Is that totally defunct now? Is it worthwile reaching out to the designer to ask for the software?

Cheers
 
#6
The mouth, throat and length all play a role.

I leave the details to a partner that really knows his way around horns. He is a master at MJK and has a large suite of design tools.I just draw the boxes after he designs them.

I did do some paper & pencil > real working examples (PA stuff), just using Dinsdale as a ref. They turned out OK, but i know we could do better today.

dave
 
#8
I strongly suggest your first speaker to be a low Q sealed sub. You do want success on the first try?
That will get you to understand how to measure the driver, how to prototype the box, and then how to measure prototype in situ before building a nice box.

Bass horns are difficult and very big. You can play with tools like Hornresp.
http://www.hornresp.net/
As a hint, Klipsch used the room walls as the mouth of his horns.

Note: A low Q sealed sub blends into a room much better than a ported sub.
 
  • Like
Reactions: MrKlinky
#10
Yeah I appreciate the responses and understand the practical advise.

As I said I'm interested in the acoustics and mathematics of horn based speakers which is leading me towards at least attempting a design and simulation. It may be that I don't actually go forward with the build if I'm not convinced of my work.

But just in terms of simulation. Is Hornresp the best modelling tool available? Are these elusive MJK tools not possible to get a hold of?

Thanks
 
#13
Davor D said:
"Low Frequency Horn Design Using T/S Driver Parameters" by D.B.Keele
Click to expand...

That paper also influenced my early designs. Last i saw of them was them going off to live in the wall of a club in Seattle.

mcphee-bass-MkII-4up.jpg


dave
 
#14
For a horn subwoofer, a scoop is a well studied and tested model, used by most reggae soundsystems. They are relative big for not going that low, but when used in stacks of 4 they couple and go lower than the individual scoop would go. The main reason that they are still used is the specific sound due to the out of phase horn and the fact that they can go very loud. Prime examples that are well documented are the Hog scoop and the Mogale scoop, both original designed for the Precision Devices PD1850 subwoofer driver (but usable with a lot more). Picture is a typical reggae/dub soundsystem with 6 hog scoops as subs.

But these days tapped horns, transflex and similar alignments go louder and lower easely, and the shift is also happening in reggae soundsystem land where more and more soundsystems replace their scoops with transflex and tapped horns.
 

Attachments

  • todo_reducido[1].jpg
    todo_reducido[1].jpg
    25.8 KB · Views: 79
#15
sellurstuff said:
But just in terms of simulation. Is Hornresp the best modelling tool available?

Thanks
Click to expand...
Sellurstuff,

Hornresp simulations hold up well to actual outdoor measurements of builds that conform to the simulation.
Akabak allows more horn sections that can allow for the cross sectional changes that occur through folds, though those differences don't amount to much, and probably are not worth the added complexity, especially when considering the influence of the room.

If you can build a good box, Hornresp will get you within a few dB of actual response- put the box in an acoustically small room, and the response will vary by as much, or even more than +/-20dB depending on the measurement and box placement.
Because a large bass horn also has directivity in the upper sub range, room response variations will be larger than using a relatively small front loaded box, or multiples spread around the room.
Hornresp is completely adequate for design and predictions of simple folded horns like the four "Levan horn + extension" and additional Klipsch horns used in the Paradise Garage (and many nightclubs), but the difference in sound between a 5000 square foot club and an average domestic room should not be underestimated.

Art
 
Last edited:
  • Like
Reactions: BP1Fanatic and waxx
#16
I too would recommend getting to know Hornresp. The software can be used very in a dynamic and interactive way that will quickly will get you an understanding of how the various parameters interact. Also there is a large community of users, you should find plenty of support if needed .

The more difficult bit is trying to turn those outrageous sims into something that can be physically built.😎

As ever there are many options for bass speakers - with full sized FLH horns being the biggest and most efficient. The physical size of FLH horns are too much for most users. Hence the endless quest to find the best design compromise to suit various environments.
 
  • Like
Reactions: BP1Fanatic
You must log in or register to reply here.
Top Bottom