Isosceles ( A speaker project)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Back a little over a year ago I decided to design and build a two-way speaker system. The primary goal of the project was to use my basic understanding of wood working, electronics and sound to the reach an acceptable end result. To that end I think it was a great success. The system sounds good and I learned a lot. But after all the learning and experience I knew I could do better.

The basic flaws with the current design are two fold. One is a serious diffraction around 2K and the other issue is box resonance related. The diffraction issue is because the tweeter is located in the worst possible place on the baffle dead center to three sides. SPL measurements show it clearly as it is audible. The later issue is a little more subtle in that the mid-bass has a muddiness on most music material that ruins the overall illusion. It’s not in your face there but it becomes apparent during listening sessions. My believe is that this is related to panel resonance and/or standing waves in the box. I have given some thought on how to control the later without a new cab design but I thought it better to create a design that minimizes the issue to begin with and while I am at it correct the diffraction issue as well.

There are a couple of upfront challenges I foresee with this project. First will be the panel cuts as I have only the most basic power tools. But I have a few ideas brewing on how I can cut the angles with enough precision to achieve a professional finish. The second challenge will be getting comfortable with Soundeasy V14 so I am looking forward to the evolution of Pinkmouse’s “This is a journey into Sound(easy)” thread.

I was thinking of sharing my journey into this project if there is an interest.

So I’ll start with a name, description and a picture and see what happens…

Isosceles
A two way closed box design using the C30-6-23 and C173-6-93 Accuton drivers.
 

Attachments

  • isosceles.png
    isosceles.png
    84.6 KB · Views: 703
Hi,

I will point out such a cabinet design is extremely volume inefficient.
You'll probably end up with as much wood as you have airspace.
Not only that, triangular sections do not kill resonances, treating
the angled faces as a mirror you are back to a box and the usual
resonant modes of an orthogonal box somewhat modified.

FWIW with adequate damping the internal standing wave modes
of a box are a non-issue. Panel resonances (and consequently
panel transparency at certain frequencies) are the more likely
cause of your percieved problems.

:)/sreten.
 
Yes, I too have debated building a speaker with that cabinet design. I don't see any reason why it wouldn't work. However, for some reason I never went all the way with the design. I have done some trapezoidal enclosures before, and if I were to do a design such as the one you posted, it would be more like this.

http://undefinition.googlepages.com/Cygnusideas.jpg

Don't be discouraged from doing angled cuts and cabinets even though you don't have much of an arsenal of power tools. So far, I've cut most of my angles with just a cheap Ryobi battery-powered circular saw.

Is this your second design? Accuton drivers are pretty much the pinnacle of DIY drivers, but I don't know how easy they are to work with. They're also big, big bucks. Personally, I don't know if I trust my XO skills to them yet. I still feel safer wading in the waters of Vifa, Audax, and Seas. But it's your design, so don't let me or anyone else dissuade you.

-Paul
 
Hi sreten,

sreten said:
Hi I will point out such a cabinet design is extremely volume inefficient. You'll probably end up with as much wood as you have airspace.

I haven’t work out all the volume requirements for the bracing but the current cab dimensions on paper are 33 X 9 X14. Those dimensions provide about 33 liters of volume. About twice the amount of the final Vb target. Plenty left for bracing, drivers and cross over components. I agree with you in that as the Vb requirements increase the design becomes less practical.


Not only that, triangular sections do not kill resonances, treating
the angled faces as a mirror you are back to a box and the usual
resonant modes of an orthogonal box somewhat modified.

But they do reduce them no? I think I understand what you are saying in that the geometry of the design can also be thought of as half a rectangle.

I had this idea to measure the modes in the box. Simplistically I could think of having a mic inside and measure SPL at various points. I would like to see what the effects enclosure shape and damping material have on the box modes. (If you or anyone else has thoughts about this I would like to hear them.)



FWIW with adequate damping the internal standing wave modes
of a box are a non-issue.

I think adequate is the key word there. What is adequate for a given enclosure design type? Some say they have experienced that speaker materials have influence on how much sound from escapes from the box. The other factor I have observed is a reduction in SPL as the stuffing density increases. Or maybe I am just thinking about it too deeply.


Panel resonances (and consequently panel transparency at certain frequencies) are the more likely cause of your percieved problems.
You may be right. All the more reason to design a new cab!
Thanks for your feedback.
 
Hi Paul,

Nice designs. I was considering going traezoidal and decided to just remove the 4th panel the panel altogether.


On the angle cutting I am not discouraged…I am challenged! I was actually thinking of making the angle cuts with a router. The smallest of which requires an 18 degree cut (back corner). I foresee very difficult challenges trying to cut that angle on a MDF panel with a circular saw. But I digress … still learning a lot. I would be interested in your technique with the saw. I certainly haven’t thought of every possibility.

Yes. This is my second design using the same drivers. If you asked me how easy they are to work with six months ago I would have said extremely difficult (At least for the units that I have). Presently I would say they are not very easy. It’s really all about is how much time, effort and money you are willing (or able) to devote.

There is nothing wrong with the safer waters. I can be a little too adventurous sometimes … like to go out over my head and see if I can swim.
 
Allways astonished by the global acoustical knowledge of sreten:bigeyes: :cool:
Also allways astonished by the relatively poor means of the diy community to cancel/reduce standing waves in our enclosures without using damping material thus reducing efficiency thus reducing dynamic range. Indeed we see anechoic shaped cubics but too small to have any effect or unparalleled walls allowing standing waves to develop in the height of the enclosure (ie your isosceles ttruman). Nevertheless it's so simple to build inner dimensions smaller than the wave lenght at least in the upper-bass/low-mid ranges which are the most critical.

An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
 
crazyhub said:
or unparalleled walls allowing standing waves to develop in the height of the enclosure (ie your isosceles ttruman).

ouch! Geez crazyhub I havn't even posted the first draft of the internal drawings. :irked:.
But I somehow get the feeling I may be getting in over my head already.

In any event my idea to correct that was to place a pryamid at the top and bottom of the cab. I hope that is to your liking. :)

BTW your speakers are very impressive. I am very interested in how to eliminate standing waves in a box but I fully admit I am not well versed in this subject. However I am willing to learn and that's why I am here.
 
ttruman said:

The smallest of which requires an 18 degree cut (back corner). I foresee very difficult challenges trying to cut that angle on a MDF panel with a circular saw.
Piece of cake. Try it on a piece of scrap. Even chintsy circular saws have a tilt adjustment. By the way, don't worry if it's 17 or 19 degrees or whatever. The pieces will fit together without much trouble (especially with a brad nailer or screw n' glue)

The cinderellas have a 6.5 degree bevel, I think.
http://undefinition.googlepages.com/cinderellaprogress

I just finished these. I think it was a 12 degree bevel. I forget, but it cutting was very easy.
http://undefinition.googlepages.com/VaagenRough1.jpg
However, I used clamps to put this one together. And that was a headache!
http://paulcarmody.googlepages.com/ClampingNightmare.jpg

Any more questions about poor man's non-square woodworking, drop me an email. As Andy G would say, "Cutting crooked is easy!"
 
Please: how guys do you automatically insert "Originally posted by..." in a quote? Never be able to discover the trick :confused:
In any event my idea to correct that was to place a pryamid at the top and bottom of the cab. I hope that is to your liking. ;)
mhhh...not really...but maybe pyramids of the biggest sorts:) something like 10" high otherwise waves will have almost equal sizes thus producing a standing wave.
;)
 
Thanks for the trick ttruman!
ShinOBIWAN said:
Great idea for the cabs Hubert but then I saw what drivers you had in them. Ugh! Filthy Accuton. :D ;) :clown:
God, my acoustical diodes network seems to work perfectly and even better than intended because it does even kill all the distortions related to the stiff cones of my very bad C220NT6!!!:bigeyes: :xeye:
Lucky Shin with your perfect AT drivers: no need of complicate cabs or filters :)
________
Masohub.;)
 
crazyhub said:
Allways astonished by the global acoustical knowledge of sreten:bigeyes: :cool:
Also allways astonished by the relatively poor means of the diy community to cancel/reduce standing waves in our enclosures without using damping material thus reducing efficiency thus reducing dynamic range. Indeed we see anechoic shaped cubics but too small to have any effect or unparalleled walls allowing standing waves to develop in the height of the enclosure (ie your isosceles ttruman). Nevertheless it's so simple to build inner dimensions smaller than the wave lenght at least in the upper-bass/low-mid ranges which are the most critical.

An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Very interesting. Got any measurements of the efficacy of this sort of internal layout? I'm interested as I have been exploring similar ideas.
 
Brett said:
Very interesting. Got any measurements of the efficacy of this sort of internal layout? I'm interested as I have been exploring similar ideas.
No measurements, even no comparison with/without diodes. Just respecting the propagation laws; my network acts as preventing the standing waves to expend lower than a given frequency (size of the cells), as breaking the waves and guiding them to the farest ends of the enclosure (top and botom), as preventing the waves to return to the driver cone (at least with a reduced level), as preventing own resonances (vents effects thus each cell possess two wide opened sides and almost all cells have different volumes). A first and quick view could letting one think that this network works like a tranmission or resonant line but...absolutely and definitely not.
What I can honestly say is that with a piece of only 10"x10" of very thin sort of felt behind the woofer, even running fullrange, absolutely no resonance is audible if I stick my ears close to the driver or close to the vent. Frankly, never had I expected such an efficient effect before building this network.
My other speaker uses a SEAS CA26RE4X, sealed, heavy and dead walls, several braces, gold number sized (modulus from Tony Gee); the C220NT6 in my special enclosure kills the bass and low-mid ranges of the modulus by far even if vented with a high fb (theorically largely inferior than the sealed -Q 0.56- SEAS); Thus: quality of the driver, of the enclosure properties...or both?...
;)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.