standing waves occur when ,inside a box or a room , we have two or more parallel faces ,depending in which fluid they propagate ,in our case air , they cause a peak at certain frequencies . In a tall cabinet is easy to find them ,specially low frequencies . The bad thing is that they interact with the movement of the membrane,als
r they propagate trough the surfaces.An egg made of iron would be good...
r they propagate trough the surfaces.An egg made of iron would be good...Egg shaped speakers have already been comercialized. However, I can't recall the brand name of the speaker.
IMO, a "near cube" has advantages because the panels will be as small as possible for a given volume. More rigid, higher resonant frequencies. Still, I prefer an irregular box. There have been lots of references to "monkey coffins" in other threads and of the dozen or so speakers I've built (not a huge number I admit), the irregular boxes always had the cleaner sound, particularly the lower midrange. Conventional boxes, including the golden ratio boxes I've built, were never as good. I don't think standing waves and reflections are an issue with a stuffed box, as long as it isn't super shallow, only how the box structure vibrates and radiates. It's not that much harder to cut the panels to slope the front or back. Regardless, materials and bracing, not to mention drivers and design, probably trump shape as far as the final result.
CH
CH
I bookmarked this Post by GM ages ago. http://www.diyaudio.com/forums/multi-way/25799-golden-ratio.html?postid=300643#post300643 Worth a look if you want to try some different ratios, which GM says many of which are actually better than the golden ratio when it comes to speakers Though I do find the 1:1.4:2 one to seem a little unlikely (due to one of the dimensions being double the other)....
The way I dealt with the problem myself for my small (5L) enclosures was to use prime numbers for each of the W X H X D dimensions... no chance of any of them being a multiple of any other 151 X 173 X 211 = 5.5L which works out nicely for a target of 5L (using the 10% over rule of thumb).
Tony.
Though I do find the 1:1.4:2 one to seem a little unlikely (due to one of the dimensions being double the other).... The way I dealt with the problem myself for my small (5L) enclosures was to use prime numbers for each of the W X H X D dimensions... no chance of any of them being a multiple of any other
151 X 173 X 211 = 5.5L which works out nicely for a target of 5L (using the 10% over rule of thumb). Tony.
I don't see that.
It doesn't take a lot of box resonance to contribute to reducing the downward dynamic range of a loudspeaker. And the research Toole presents shows that the low Q event is more audible despite the peak level being lower.
dave
Waveform Mach 17 (were made in Canada)
An externally hosted image should be here but it was not working when we last tested it.
dave
I tried all kind of shapes and forms and without damping material they all had an impedance "blimp" in the fundamental tone range independent of shape. Some damping material removes that usually. I use only so much that the small wrinkles in the impedance curve disappear. It is best to look at phase because that is more sensitive. Another solution is an internal Helmholz resonator tuned to the problematic frequency.
Althou a cube contains more volume with less material, that does not necessarily lead to a better panel. One of the cardinal rules of bracing a panel is that the subpanels have a higher aspect ratio than the original. A panel tends to vibrate across its smallest dimension and being smaller is at a higher frequency.
A square panel has 2 reinforcing modes, and for its area the lowest resonant frequency.
A trapezodal panel (or subpanel) is even better (assumming that the "short" dimension is the non-parallel one).
dave
Attachments
Exactly why i suggested one be careful using root(2)
dave
after pulling out calculator says doh
didn't actually sink in when you said it Dave Tony.
whats wrong with golden ratio??
i have used GR for all of my loudspeakers, i see no problem and also no reason why anyone would want to deviate from it (except in the prime number case mentioned earlier, which may be just as good, though argueably better IMHO)
at the end of the day, this isnt about speaker box dynamics, its about aestetics.
even in my latest ML-T-TL the large open area is governed by GR with respect to the height as much as was possible. ie open area is GR and length conveniently is a multiple of GR...3 or 4 times it i think from memory. this still allows for QW resonance on which the design relies but should minimise all other spurious resonances. also driver distances on baffle are GR. as are braces internal to box geometry.
@Ged:
I know that you are a knowledgeable guy...but SURELY even a NEAR cuboid box will lump all the dimensional resonances unacceptably close together, thus near-as-damn-it giving one larger spread resonance of greater amplitude than the 3 separate ones most GR boxes would exhibit. much as i don't doubt what you say about CD horns and waveguides, and the like, i have to say that i couldn't agree less about a near cuboid box design-you only have to look at those horrid Bose things from a few years back.......
The the OP, i would use golden ratio, the height doesn't need to be the height dimension. Put the speaker in any FACE of the formed box that you like. The only affect will be the frequency of the sound reflected by the rear wall I think, and with sufficient damping material on the rear wall its effect will be minimised.
i have used GR for all of my loudspeakers, i see no problem and also no reason why anyone would want to deviate from it (except in the prime number case mentioned earlier, which may be just as good, though argueably better IMHO)
at the end of the day, this isnt about speaker box dynamics, its about aestetics.
even in my latest ML-T-TL the large open area is governed by GR with respect to the height as much as was possible. ie open area is GR and length conveniently is a multiple of GR...3 or 4 times it i think from memory. this still allows for QW resonance on which the design relies but should minimise all other spurious resonances. also driver distances on baffle are GR. as are braces internal to box geometry.
@Ged:
I know that you are a knowledgeable guy...but SURELY even a NEAR cuboid box will lump all the dimensional resonances unacceptably close together, thus near-as-damn-it giving one larger spread resonance of greater amplitude than the 3 separate ones most GR boxes would exhibit. much as i don't doubt what you say about CD horns and waveguides, and the like, i have to say that i couldn't agree less about a near cuboid box design-you only have to look at those horrid Bose things from a few years back.......
The the OP, i would use golden ratio, the height doesn't need to be the height dimension. Put the speaker in any FACE of the formed box that you like. The only affect will be the frequency of the sound reflected by the rear wall I think, and with sufficient damping material on the rear wall its effect will be minimised.
Last edited:
The Fibonacci sequence might be more practical/simple. It's not un-GR; the ratio between consecutive numbers approximates the GR higher N goes. (so, to get closer to the GR just use centimeters as your measurement standard).
I'm thinking there should be a 3D equivalent to the Fibonacci, that works in 3D space just as in 2D below...
And for the record, I don't think anyone said GR was a bad thing.
- keantoken
I'm thinking there should be a 3D equivalent to the Fibonacci, that works in 3D space just as in 2D below...
An externally hosted image should be here but it was not working when we last tested it.
And for the record, I don't think anyone said GR was a bad thing.
- keantoken
lol neither do i.......but other 'experts' seem to think that using GR is 'basic' or something for beginners, and others use prime numbers or like you mentioned, fibonacci sequence. This is fine, but it can make these 'beginners' feel looked down upon by the doctors of maths that are here-although im sure in the most part this is NOT the intention.i have experienced this myself-and i, like many, am not stupid, but also am not a doctor of maths. i for eg am competant enough to do the calculus required to calculate responses ON paper, but much more than this really tests my attention span, not that i couldnt do it if i really felt the need. I get more pleasure by actually biulding them and testing, rather than the calcs and models of predicted behavior. Neither way is wrong.
suffice it to say; that GR (IMHO) is more than sufficient for the majority of speaker builders (doctors or ordinary, run of the mill people), and the only reason to deviate from them would be aestetics or if building a non parallel box design(which i would still use SOME GR in).
have to say though, i do like the blog and design within, keantoken. My HNC electronics thesis project was an amplifier...actually the P3a on elliots site, with a few paralleled output Tx's added for good measure
Last edited:
>>> Thats why as far as the internal box effects are concerned shape doesn't have much of an effect.
I agree with this. Imho, if the box is sealed and of reasonable size (not way too deep or way too wide, etc.) and it's stuffed to dampen standing waves, i have not heard much or any difference. I have heard ported boxes sound better when they leave more room behind the driver to 'breathe'... at least this is what i have experienced.
Godzilla
I agree with this. Imho, if the box is sealed and of reasonable size (not way too deep or way too wide, etc.) and it's stuffed to dampen standing waves, i have not heard much or any difference. I have heard ported boxes sound better when they leave more room behind the driver to 'breathe'... at least this is what i have experienced.
Godzilla
In the pursuit of science, sometimes tact and integrity gets set on the back burner. Sometimes it's lost altogether. Sometimes 70 years (average lifespan) isn't enough time, and it's all about time! There is a lot to learn! Generally, well-meaning people will make mistakes simply because they don't understand things. This can be forgiven, and at least as long as we're focused on learning I expect society will improve.
You've seen my blog? Funny, I didn't think anyone had seen it.
- keantoken
You've seen my blog? Funny, I didn't think anyone had seen it.
- keantoken
The point is that a small deviation from "cubic" yields a fairly large spread in the three coincident resonances from a cube. My thesis was on non-rectangular rooms so how resonances are spread by "shape" is something that I am very familiar with. The usual "golden rules" like the GR turned out to not be very useful - not that they were bad mind you, they just didn't yield much if any improvment. In shape it turns out there are a few shapes to avoid, mostly dimensions that are integral numbers of each other, but if you avoid those, then everything else turns out to all be about the same - including a "near cube". But the "near cube" ends up with the lowest surface area FOR A RECTANGULAR STRUCTURE (I can't believe that I have to make that qualification!).
Isn't "rectangular structure" still a bit vague/misleading? Perhaps you mean "for a rectangular prism, composed of flat panels, joined at their edges?"
Isn't "rectangular structure" still a bit vague/misleading? Perhaps you mean "for a rectangular prism, composed of flat panels, joined at their edges?"
I don't think its vague, but at any rate I certainly did not mean a "prism". A "Prism" comes to a point.
..twas a joke. A playful exchange about how far you would have to go to remove all doubt. But a prism is not a pyramid...
Prism (geometry) - Wikipedia, the free encyclopedia
Regards,
Ron
Prism (geometry) - Wikipedia, the free encyclopedia
Regards,
Ron
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.