Thoughts/ideas on a 21" ported horn sub

Kevin McDonough

Freshman
Jan 3, 2016
38
2
8
43
hey all.

Would like to get people's opinions on a possible design for a 21" ported horn if you would be so kind....

A little history to the design, it all started when I read a thread about the Othorn. I vaguely knew it existed before, but had never really paid it much attention. However one bored afternoon I read through the thread and was much more taken by it, and by the reactions people were having about it. It seemed like it was a MUCH better design than I'd initially thought, much more worthy of attention than I'd realised, so I started messing about with the design a little, trying some different drivers and things in it (and imagining I might have time/money to build a big pile someday :ROFLMAO:)

However I've always been partial to ported horn designs and so had a quick mess about in Hornresp, thinking that I would see what I could do in a ported horn that was the same sized box and compare it to the tapped horn version. Using the B&C sw152 and a carbon coned 21" eighteensound driver (the ones I had been trying in the tapped horn versions I'd been messing with) I was getting decent results, as good as the Othorn, but nothing that blew me away or seemed like a big improvement as such.

However I read a a random article about B&C's newest driver the 21DS115, something popped up on Facebook or whatever, and on a whim decided to try it in the ported horn design as well, just to see what would happen......

....and thought that cant be right! That's significantly better than the Othorn, right across its frequency range!
LOL


However having checked my sums and looked at everything, double checking all my stuff it seemed that yeah, this seemed to be a bit of a magic combination.

Playing about with it and refining it a bit, I was able to shrink it down to a box that was a chunk smaller than the Othorn, shaving 40 or 50 liters off of the overall volume of the box, not having quite such a compression ratio for the driver saving a little pressure on the cone, while still having what looked like a much better response!..... (my design in black vs Othorn in grey).

208860


After playing about a while and trying to work out how to fold it into an actual box, I came up with something like this......

208862
with the idea that the horn path would curve round to the right after coming through a Martin style letterbox throat, the driver baffle would have holes cut into it to open it up int the two central areas, and the port would be a long port off to the right as shown. This then turned into wood would look like.....

208863

I've had the design sitting unused for a year or so, I'm hoping maybe over the summer if I have time to make some prototypes and mess about with it. However I was reading some testing info posted on the "speaker freakers" facebook group about different shapes of ported horns, and measurement showed that both the fact that my port was rectangular and that it had a bend in it would add distortion and port compression over an equivelent round port. So i reworked the design a little and added in 4 straight 4" round ports rather than the big shelf port......

208864

The port done as 4 round pipes doesn't have the same cross sectional area, but it does then have the bonus of slightly improving the response as the 4 round ports take up a little less cab volume than the shelf, given the same overall cab size. Comparable cabs like the Void Trinity X and Nexus XL seem to have similar or less port area and seem to work ok, as do things like Martin's ported horn MLX and WLX subs, so I'm hoping it's still enough to work ok.

Just interested in peoples thoughts or ideas about the design in general, and specifically the choice of port method.


Kev
 

Attachments

  • 21portedhorntop.png
    21portedhorntop.png
    43.5 KB · Views: 14
Last edited:
haha, clearly I've stunned people in to silence with my amazing design!


Would be really interested in people's thoughts on the design, most on here have more experience in designing speakers than me, so do people think the round ports and the slight bonus that results in internal volume would be fine, or do people think the extra area of the shelf port would be required?

Ports are a combined area of 316cm2 by 35cm for the round ports, or 564cm2 by 75cm long for the shelf.


Kev
 
Would be really interested in people's thoughts on the design, most on here have more experience in designing speakers than me, so do people think the round ports and the slight bonus that results in internal volume would be fine, or do people think the extra area of the shelf port would be required?

Ports are a combined area of 316cm2 by 35cm for the round ports, or 564cm2 by 75cm long for the shelf.


Kev
Kev,

I'd prefer an offset horn driver arrangement rather than the small slot you chose for more even cone loading.
If you do use a slot, make sure it can accept the driver's 30mm one-way Xlim/Xmech excursion without the surround striking it.
I'd also like to see any valid measurements showing rectangular ports "add distortion and port compression over an equivalent round port".

What drivers, excursion levels and voltage are you comparing in post #1?
Reducing port volume (from 42,300 to 11,060) by 74% seems extremely small for a driver with the displacement of the B&C 21DS115.
How does the port velocity compare between the two at 15mm excursion?

Art
 
Last edited:
I'll add to Art's comment about port area by saying that the BR 21" design I'm currently working on has 1260 cm^2 of port area, and that's probably just about enough.
 
In my mind, this design combines the common problem of both horn and ported enclosures - energy storage. The step response will be extremely poor, as is every enclosure type compared with sealed. A design such as the Othorn is still ringing significantly at 50ms, whereas a typical sealed loading is all done by about 10ms or less. Extremely poor transient response may not be significant to your sound requirements...
A port area of 316cm2 is impossibly small for a driver of this displacement and will produce significant distortion and output compression. A very well-respected designer I used to work with always started with a port area equal to the effective cone area and compromised backwards from this figure.
Having said all this, it's still well worth a couple of sheets of board and a weekend's work!
 
Thank you all for your responses :D

Art, the paper I refer to about the different shapes of ports was one by Scott/DIYRM that was put on the 'speaker freakers' group on Facebook. I think this link should work.....

https://drive.google.com/file/d/1SU...JqXFMWy2NNZchaQ9VuNpPZ88MaV8OS7qhzQpgCteVB_Tk

Though I didn't have time to sit down and go over it in the most amount of detail ever, what I took from it was that a round port with a fairly specific flange/rounding at the end gives the best possible results, but short of that a basic round port is the best option, and everything else outside of that starts to increase problems, for a given size.

Now I realise that my own design idea doesn't really follow that, as I'm comparing much smaller overall area of round ports to a bigger rectangular ports. But the round ports provide me with the potential advantage of a much bigger internal volume and better response overall, as long as they can be made to work within the design.

I agree with all of the rest of the comments that I would always err on the side of using much bigger ports than this, but then I wonder how commercial designs do it. All of Martin Audio's ported horn designs for example, seem to have relatively small ports for a given size.

Their newest big horn sub for use with MLA/Wavefront, the MLX, only has 2 rectangular ports for each of the 18" drivers....

208875
And their previous generation ported horn subs similarly only 2 fairly small round ports, very similar looking to the 4" pipes I plan to use, for each 18" driver....
208876
208877



Similarly with Void Acoustics for example, they have various subs that use round ports, a double 18" example with 4 round ports of a similar size to mine, and a very similar 21" that, while a little more cube like than mine, has a very similar overall volume and only uses 3 round ports of a similar size for the design....

208878
208879
And even a DIY option that people seem to regard as sounding great, by Marc O, only has a pretty small/thin slot port who's area seems much smaller than expected for a high power 18"

208880


All of these designs seem to get good reviews and by all accounts sound good, despite having ports smaller than what I would certainly expect was ideal.

Plugging the numbers into BassBox Pro, my usual go to for reflex designs, to check the port velocities etc, it says that in both of my designs, at the rated 1700 RMS power the drivers only reach about 10mm of xmax until way below their pass band, and they need to be pushed to nearly 4000w before they start to approach their xmax of 15mm.

Obviously the rectangular port fairs a bit better with the much bigger area, but even the smaller set of 4 round ports (which incidentally BBP says should be longer than what Hornresp says to get the correct tuning, but still just about short enough that they can be a straight port and not have to bend within the cab) only JUST begins to reach what BBP considers the danger zone for port velocity when pushed to that 3000-4000w.

Also Art, as per your suggestion I tried modifying the design to show an offset horn. the response isn't quite as good as when a normal horn is used, but it's still possible to get a decent response with the cab being a similar overall volume, and the rear chamber and porting etc all being fairly similar.


(I've not had a chance to grab screen grabs of Hornresp and BBP while making this post, but will add some in later today.)
 
Last edited:
Adding some screen grabs and data, first the hornresp input screens. I'm way less experienced at horn design than all you guys, so of course feel free to point out any obvious mistakes I've made, always happy to learn how to be better.

Normal horn with bigger shelf port.....

208881


Same horn with the 4 smaller round ports....
208882

The response from these two......(round ports in black)
208883

Cone displacement and port velocities from Bass Box Pro, firstly at 1700w (red obviously being the smaller round ports)
208884

And at 3500w, where BBP just begins to consider port velocity a problem....
208885



And as a comparison, Art's suggestion of an offset horn that has about similar volume and similar response.....
208886

and the two types of horns compared (offset horn in black this time....)
208887
 
Last edited:
Yes a round port will perform better than a rectangular port of the same length – intuitively the round port has less surface area for the given port volume, so less loses.

BUT … often it can be easier to use a much larger and longer rectangular port to give you the same tuning frequency, this may give you better performance than the smaller round port. Often port design can be integrated into the box design to strengthen the box minimizing losses due to the box flexing etc. ... There is no simple answer.
 
Thanks Peter, I know the page well and refer to it often. It's what initially led me to testing sealed subs which I now find it difficult to move away from, as all other loadings simply sound to me so poor and 'smeared' in comparison. I just have to live with their dire efficiency and hunger for amplifier power!
 
Is the design in the OP not essentially the same as Riccis SKhorn design?

Well, in terms of the fact that it is a Ported Horn, then yeah it's a similar design.

But most of the actual design is different, it has a different horn and port area, different internal volume, different fold design etc.


Following Art's suggestion I started working on an offset horn version, playing with the design, and it's actually came out well. It's a much simpler build, less pressure on the cone, I think that it's actually the one I'll probably run with and try a prototype over the summer.

208891

208892



Still needs a lot of work: might play with the ports more, needs lots of bracing added etc, but definitely worth a little sawdust I think.

Kev
 
Last edited:
  • Like
Reactions: Carl Klinkenborg
In Hornresp, double click on Le to change to the Lossy Le model (Le should turn red). That should be closer to reality. These B&C DS varieties have fairly high Le's so it's going to make a difference.
 
In Hornresp, double click on Le to change to the Lossy Le model (Le should turn red). That should be closer to reality. These B&C DS varieties have fairly high Le's so it's going to make a difference.

hmmmm

yeah that's not ideal lol, it changes the response quite significantly, and not in a great way lol :oops:

208897



are there any links for reading on what changes this Lossy LE function makes, why it's applied etc? would be good to understand it a bit more.

Kev
 
I’m Far from knowledgeable but if you turn the driver to face away from the ports and rework the horn path along 3 walls of the box would reduce the rear volume while increasing the horn length. I have no clue wether this is advantageous, just an observation.you might even get the ports bove and below the driver
 
  • Like
Reactions: Roland Clarke
Check out the SKHorn by Ricci on data-bass. Hes also working on a single driver version. Both are exceptional designs and have unbelievaable output.

Thanks. The single driver Skram design is done. Plans are in the forum. I don't have my cabs completed yet so I can formally test it yet though. Both cabs are getting great feedback.

Also for anyone who drops by DB occasionally there is a new V2 version. If you got to the forum and then click the DB logo it will take you to the new version. There are some new tests that are not on the old DB which is no longer being updated.
 
hmmmm

yeah that's not ideal lol, it changes the response quite significantly, and not in a great way lol :oops:

View attachment 208897



are there any links for reading on what changes this Lossy LE function makes, why it's applied etc? would be good to understand it a bit more.

Kev

There's a bunch of discussion on inductance going back many years on various forums. It's spread all over but the majority is at DIYAudio, AVS or DB. David kindly incorporated complex semi inductance modeling into HR last year. D. Bolserst has developed a free excel sheet that can derive the semi inductance specs from .txt files of impedance measurements via a curve fitting process. It's downloadable from DB or DIYAudio. All of the drivers that I have measured have the semi inductance parameters used by HR posted in the driver page. Click the Le field in HR and it will pull up a screen for their entry. The Le label will turn green when active. It is best to use the specs from the exact driver but they can be used generically with the factory specs. Try using the 21DS specs I have measured at DB for your HR sims using that driver. It is a LOT closer than using nothing or the Lossy Le function (Le label RED). I'm hoping that in a few years these specs will be adopted and reported by the major MFG's, but the industry is slow to change.

Long story short...Yes it matters if you care that what you are simulating will perform like the real speaker.