I'm trying to understand myself and this is how I visualize it.... (be gentle)
A sound wave that is 50 feet across will not reflect off surfaces like a tennis ball. Think of the initial sound pressure wave as a balloon expanding. It will follow any path it can find to release the pressure. This allows for bizarre things like Helmholtz resonator bass traps. A few tiny holes in a panel can suck the energy out of sound wave the length of the wall its mounted on.
The sound wave is coming out of the horn before the edges of the wave have made it off the back of the driver's cone.
Last edited:
This is not the image I trying to paint but it seems to be what most are thinking of.
I'm saying that regardless of the wavelength, it reflects as soon as it (any portion) contacts a barrier.
It can't reflect until the entire wave has hit the wall. To do so would lengthen or shorten the wave. If a 30hz wave (37ft) hits a 12 inch surface it doesn't reflect a 1khz wave, it wraps around the surface.
This is why we have baffle step issues (2pi vs 4pi) a sound wave longer than the face of a speaker wraps around the speaker before radiating out.
There may be reflections of some sort, but they won't be of significance primarily due to the wavelengths. It should be much like any closed box in that sense. If reflections at these frequencies were an issue in a T/L, they'd be even more so in closed boxes. Yet adequately stuffed closed boxes do not exhibit resonances associated with frequencies below those for which the the distance is less than a quarter wavelength. Even unstuffed ones generally show little of this at the low end, for these frequencies it's in the area of box pressurization. For a T/L there won't be pressurization, but due to wavelengths involved, by extension I don't see how it could be an issue in the area of T/L tuning frequencies.
Dave
I think it reflects a 30hz wave back off of that 12 inch square. How can the entire wave hit the wall? Things happen over time - a 30hz wave is not a data packet deflecting off an obstacle in one piece.
I postulate that the reflections are significant and are what excites the column of air to resonance. If the space is too small or the wrong shape the air will not resonate, regardless of the reflections.
My idea on how this works explains how the sound wave manages to get around a 180* bend (or 2) in a TL. It is backed up by the fact that a low frequency sound wave will reflect from a hard surface at close proximity.
I think if you measured near the surface of a 12 inch column would notice changes in SPL as the wave passed, but I'm not sure you could get a reflection.
low frequencies do some weird stuff.
Sho'nuff. This is why I build protorypes. TL's seem to be a lot more complex in execution than simple math models suggest. My experience. Some worked well, some not. I would never suggesst someone "delete" a paper because I was in a disagreeing camp on design philosophy. Actually, I very much enjoy reading older papers where I know advances have been made. Sometimes you actually find insights not considered in the "newer" paper. Heck, we still teach the Bohr atom model because it is easy to understand, even though we now think the position of an electron is a statical probability within a region. Does that mean I should burn my Basic Electronics book? I think not.
John,
How about this off the cuff analogy. Because the wavelength is long compared to any of the box dimensions, it is more like the snake you use to clear out that napkin caught up in the plumbing works... it just follows the pipe. As wavemegths start approaching the dimesions of the duct (much higher than what we are interested in) it starts behaving like you describe, but by then it is a "who cares". It is at that point, hopefully, killed by the damping. It can have the same kind of effects as in any other box, but is too high in frequency to have any affect on the TL funktion.
dave
How about this off the cuff analogy. Because the wavelength is long compared to any of the box dimensions, it is more like the snake you use to clear out that napkin caught up in the plumbing works... it just follows the pipe. As wavemegths start approaching the dimesions of the duct (much higher than what we are interested in) it starts behaving like you describe, but by then it is a "who cares". It is at that point, hopefully, killed by the damping. It can have the same kind of effects as in any other box, but is too high in frequency to have any affect on the TL funktion.
dave
OK, OK, I'll take a shot at the middle question.
Having just got back from the pub, let's hope I can still think (and draw) straight. 😀
Question:
Answer:
Sound of all frequencies will bounce of a hard, solid surface like a concrete wall. It doesn't matter how close the speaker is to the wall.
In the pictures below, we can see how this works. The red arrows show some of the directions the sound travels after it leaves the speaker.
In the left picture (with no wall), we see that the listener hears only one of everything - the sound traveling directly from the speaker to his ears. This is good.
Notice how confused and unhappy the listener in the second picture is though! This is because he is hearing two of everything - firstly the sound that comes to him directly from the speaker, and then a moment later, the sound that bounced off the wall first before reaching him.
Apparently people with good ears and Hi-Fis find this unpleasant, which is why folks like Earl Geddes work hard to invent loudspeakers that don't bounce too much sound off the walls.
OTOH, teenage knuckledraggers prefer to maximize this effect by shoveling their "subs" up against a wall, or preferably right into a corner (so it can be near two walls) because they want MORE BASS and don't care what anything else sounds like, so long as the furniture's jumping.
This works like so:
Assuming the speaker is about a foot from the back wall, the path traveled by the sound that bounces off the back wall on it's way to the listener is about two feet longer than the path traveled by the direct sound. This means that the reflected sound arrives at the listener about 2mS after the direct sound.
At low bass frequencies, the delay is a small fraction of a wavelength so the direct and reflected waves add nicely to give a result at the listener's ears of about double the amplitude of the direct sound alone.
As frequency increases, though, the delay results in greater and greater phase shift between the direct and reflected sound, with the "addition" becoming less and less effective, until at about 250hz the two are completely out of phase and cancel each other out.
At still higher frequencies, bla, bla, geez I'm sick of typing now. Sorry folks, that's as coherent as it get's tonight (after 2AM here).
Cheers - Godfrey
Attachments
I think water flowing through the pipe is a pretty good analogy. It doesn't care a whit if it bends around a corner and it's certainly not going to back up and flow the other way because of a 'reflection' at the bend. Now water flow is DC (0 Hz) but, at bass frequencies, the airflow is close enough to DC-like in a typical TL that we can look at it much the same way. The cycles last long enough compared to the pipe's dimensions that there are long periods of time where the air is flowing only one way or the other.
Last edited:
They do, but at ever decreasing effect as the frequencies go down, but even DC will see some effect of a bend. As the frequency goes up the bend effect gets more and more pronounced. Studies have shown bend to have a significant effect on horns.
Analytically bends can be exceeding difficult which is why they are nevr included in models.
At some frequencies with some bends the sound can be virtually entirely reflected from the bend. It doesn't go arround it at all.
Analytically bends can be exceeding difficult which is why they are nevr included in models.
At some frequencies with some bends the sound can be virtually entirely reflected from the bend. It doesn't go arround it at all.
More accurately: "how does the low frequency sound wave go around an 180* bend if it does not reflect off the sidewalls of the pipe"
If one is careful with the input of the geometry, the bends can be accounted for somewhat in Martin King's modeling software.
In a horn such as Sachiko the expansion at the bends act as low pass filters.
dave
Dang, I feel slighted. I didn't rate getting called either a putz or a turkey. Maybe he'll correct that in future AA postings. 😀
What you describe is fluid motion, not wave motion. A wave will propagate, expanding straight ahead until it meets an obstacle or barrier. The effect the barrier has is usually reflection. As the wave expands from the source, it immediately meets the barrier of the sidewalls of the pipe. It reflects and it is this reflection that maintains enough energy within the pipe to excite the column of air to resonance.
"couple of clueless, arrogant pseudo-scientists (Earl Geddes and Martin King)"
I am honored to be mentioned in the same insult as Earl, I am a little disppointed that Earl reached "putz" status and I fell short. Something to strive for in the future.
He has been warned by "the Bored" so I think the length of his stay at AA will be under review by the higher powers.
Last edited:
- Status
- Not open for further replies.