You have to realize that in order for the concept of the large radiating surface on a relatively small baffle to work, the baffle really can't be too big. The ideal would be a piston in free air, but without the back-front cancellation.
Adding a radius would make the baffle as a whole larger. The larger you make the circular baffle, the more the system resembles a single point source on a circular baffle, which is pretty much a worst case scenario. Therefore you have to ask yourself, what can be gained by adding the radius? The radius would have to be pretty large to be effective below 2k1 (where I put the crossover in my system). For that reason I think a radius might do more harm than good in this particular case.
So what exactly is worst case on the sphere I suggested?
But you don't have to try of coarse
. A ~ 10 cm sphere would be enough for your tweeter, how hard is that?
I know it works, I already tried it:
Car Audio | DiyMobileAudio.com | Car Stereo Forum - View Single Post - VW Passat B5 (2000) SQL Build
Those are tweeter pods for the Vifa XT25 I designed based on earlier models I created for my own car. Hanatsu had those printed at Shapeways and measured them. (see above link)
The diameter of that one is 7 cm as it would be crossed over a bit higher.
But you don't have to try of coarse
. A ~ 10 cm sphere would be enough for your tweeter, how hard is that? I know it works, I already tried it:
Car Audio | DiyMobileAudio.com | Car Stereo Forum - View Single Post - VW Passat B5 (2000) SQL Build
Those are tweeter pods for the Vifa XT25 I designed based on earlier models I created for my own car. Hanatsu had those printed at Shapeways and measured them. (see above link)
The diameter of that one is 7 cm as it would be crossed over a bit higher.
Attachments
Last edited:
I see a lot of talk about diffraction on pipes like that, but it's not an issue for mids or bass- The driver's too directional by the relevant frequencies to matter, and it's typically WAY out of the passband anyway. I'm all for perfectionism, but having no meaningful baffle bypasses the diffraction issue in large part. It's 4pi radiation for the entire pistonic range of the driver, and whatever actually "sees" the lip is breakup mode HF stuff. Tweeters, you'd definitely want to be careful.
Where do you have this information from? Any real source, or just a "good guess"?
Olson's curves do lack the resolution we get so easily today with our computers. But they are from 1950 (the AES paper from 1969 is a reprint from Audio Engineering).
-Bjørn
Seems that in this case there is good support, but that is not true in general. For example, Keele mentions in one of his papers how the data for horns that Olson shows is incorrect. And I have found several errors in his books as well. So its not as if I mean to imply that everything that he did was wrong, only that I have seen enough errors to not take it for granted that it is right.
Something in the Olson data is odd, just look at the curves for the cylinder - clearly straight line were drawn. You never see anything so clean in real life. They may be based on measurements but there was definitely some hand manipulation going on as well.
Last edited:
I did use the word: effectively, no?
BTW, I don't think your statement (*) is correct. I believe that diffraction (in our context) essentially creates a similar effect to a helmholtz resonance.
Last edited:
It is important to recognize the type of change diffraction is causing. Smooth and flat response is a very important goal, the most important goal. But if the resonance disappears on the listening window curve, I wouldn't worry about it too much. By getting successive views through reflections, which may have a slightly different comb pattern, the ear/brain may be able to ignore this type of coloration. Things like ERB might reduce its audibility further. Again, all I'm saying is that let's focus on the things that we know make a difference rather than those that appear attractive and fixable, but may not be audible.
We use gating all the time to remove reflections and see the direct response. Some will argue that the small delay from diffraction cannot be treated the same way. But do we really know? If you could gate those reflections out as well, would we still be discussing this issue?
That's not really correct IMO. Diffraction is delayed in time and from an auditory perception perspective is an entirely different thing than a resonance that is synchronous in time. The later can be masked, but the former gets unmasked at higher SPLs - that's a very very different thing.
Now a delayed resonance is more like diffraction, but most resonances are not really delayed - cone resonances for example are not delayed. A basket resonance can be delayed by some amount but usually not large. A cabinet resonance could be delayed by a significant amount and would appear very much like diffraction.
Clearly its not simply that "diffraction is similar to a resonance".
If you wanted to consider the ear in this context, then I would say that its gating time is somewhere in the 2-6 ms. range (pretty wide because it is certainly frequency dependent). This is why diffraction is such an issue. It is right in the gating time period causing confusion in our hearing as to whether it is in the direct signal or is it a reflection. The same issue happens with measurements in small rooms. Right about the gating time it is difficult to tell if an aberration is from the speaker or from the room.
Why is there a gating time to our hearing at all? Think about it - it takes time for the wave on the cochlea to traverse its length so clearly all frequencies are not detected simultaneously as they are in an FFT. LFs are delayed much more than HFs, which is also why our critical bands are so skewed with frequency being very wide at LFs and very narrow at HFs.
Do you have a paper for your diffraction study?
How was diffraction simulated?
How were stimuli listened to?
Were different playback levels used, and normalized for playback on reference system?
No disagreement from me there, that's why I proposed the tweeter modification mentioned above.
In my car I have a 2 way plus sub. The tweeter has a sphere shape, the woofer looks like a woofer on a short cylinder
Posted link shows old and new pods are both highly compromised uses of tweeter.
New pod is lumpier than old pod.
I'm thinking more along the lines of front/baffle and rear/panel edge's diffraction creating a resonant condition - specifically a build-up and release of group of near freq.s. Effectively and "open port" on the sides of loudspeakers, modulated by the driver on the front baffle.
Yes? No? Maybe?
Pods introduce raised lip as reflective ring; standoff becomes baffle step. Response shows diffraction in 2-3kHz; likely right in crossover region.
I'm thinking more along the lines of front/baffle and rear/panel edge's diffraction creating a resonant condition - specifically a build-up and release of group of near freq.s. Effectively and "open port" on the sides of loudspeakers, modulated by the driver on the front baffle.
Yes? No? Maybe?
Maybe, maybe? I am not sure because I can't quite figure out what you said.
The diffraction off of the back panel edge will always be very small because of distance and the fact that it is in the shadow zone of the drivers. The main sources of diffraction are those things mounted on the front baffle and that baffles edges. You can basically ignore anything beyond that.
Ok, so No. (..I thought the rear cabinet edge's diffraction was more significant and in conjunction with the front baffle's edge diffraction creating a low pressure region on the side of the cabinet when the front/source was in high pressure, and vice versus.)
-good to learn something!
Sorry Keyser, my bad.
Last edited: