Hello,
I did not simulate with axidriver so far, because I assume significant differences between the axisymmetric and the "linear" waveguide. The SPL distribution will differ because the axisymmetric WG opens quadratic, the one for a line source linear, if setting distance and area of the "wavefront" into relation. The lobing, shown in Jean Michel on LeCleac'h horns will show an additional impact.
Apart from that the simulations look good.
Regards, Timo
I did not simulate with axidriver so far, because I assume significant differences between the axisymmetric and the "linear" waveguide. The SPL distribution will differ because the axisymmetric WG opens quadratic, the one for a line source linear, if setting distance and area of the "wavefront" into relation. The lobing, shown in Jean Michel on LeCleac'h horns will show an additional impact.
Apart from that the simulations look good.
Regards, Timo
soongsc said:
Yes - an Air Motion Transformer driving a "dipole horn" (DiHo) - or a "Dipole Directivity Control Device" (DDCD) - if you will.
http://www.kinotechnik.edis.at/pages/diyaudio/DDCD/DDCD_dipole_horn.html
Michael
I think you will find the AMT type wave front quite complicated. Each fold would act like a diffraction source. I have not looked, but if AxiDriver can multiple point sources, it would probably be better than plane wave.mige0 said:
Yes - an Air Motion Transformer driving a "dipole horn" (DiHo) - or a "Dipole Directivity Control Device" (DDCD) - if you will.
http://www.kinotechnik.edis.at/pages/diyaudio/DDCD/DDCD_dipole_horn.html
Michael
BTW, I agree with tiki about simulation axisymmetric and implementing line source.
is there somewhere a new documented study on baffle edge diffraction?
here is Linkwitz study.
http://www.linkwitzlab.com/diffraction.htm
I still doubt that baffle edge diffraction is audible.
here is Linkwitz study.
http://www.linkwitzlab.com/diffraction.htm
I still doubt that baffle edge diffraction is audible.
MisterTwister said:
Dave Ralph's page has measurements with and without felt around the tweeter.
http://www.speakerdesign.net/home.html
I can't find it now but Dennis Murphy did measurements with and without a roundover on the baffle edges. They showed similar improvements with the roundover.
gedlee said:
In the very first paragraph Linkwitz notes that "At those frequencies almost all tweeters are highly directional and little sound reaches the cabinet edge to be diffracted."
But that is not the case with a constant directivity horn or a waveguide. There's a significant amount of high frequency energy radiated off-axis, by design.
I think you've misunderstood. He's referring to the tweeter lowpass. Even the most dispersive tweeter seldom has significant impact from diffraction above about 10K, CD drivers even less.Patrick Bateman said:
I have to disagree with the latter statement, a CD horn/waveguide reduces the energy reaching a baffle edge vs. a standard tweeter by design. The energy that would normally be directed along a baffle axis is directed forward, so the diffraction signature is dramatically reduced. Above it's cutoff, that is.
Dave
At the bottom end of a tweeters range it is irradiating the baffle - far MORE than a waveguide. If the X-over is at a few kHz then this is a major effect as this is the region where our ears are the most accute to diffraction effects. To me the whole discussion of "diffraction level" effects is completely beside the point. The important factor is the "diffraction audibility" first and then we can talk about what levels are significant. Most work that has discounted baffle diffraction has assumed that its only the effect on the far field frequency response that is an issue. But that is clearly incorrect. It's the audibility of the group delay that is created by the diffraction that matters and this is not at all evident in the frequency response curve. The whole discussion that I have seen on any of these links misses the point IMO.
In most basic tweeter based designs its a non issue at higher freq.s..
The reason is simple - the baffle is to large (and rectangular), and the tweeter's lower freq. response to attenuated to effect the baffle significantly.
This was SL's point in that link. (..it's relative non-effect in the time-domain can be seen/measured in any number of basic designs.)
*IF* you add-in a waveguide however, then you've just added-in a potentially significant source of diffraction: mouth exit to baffle transition.
The real problem isn't diffraction (for most designs at treble freq.s.) - its *reflection*.
A high amplitude reflection at a baffle's edge can create an apparent secondary source significantly off-axis. Particularly disturbing are those horizontally opposed to the tweeter. It can screw with lateral image localization, and to a lesser degree with image localization in respect to depth - when it's effect is somewhere in the 1.5-8 kHz region. It also rarely shows up (measurably) except in wide-angle polar plots. It is of course much less of an issue if you are listening on-axis. (..though we do hear with a degree of polar uniformity at these freq.s - so it does effect the perception on-axis, but to a much lesser degree depending on the amplitude of the reflection.)
The reason is simple - the baffle is to large (and rectangular), and the tweeter's lower freq. response to attenuated to effect the baffle significantly.
This was SL's point in that link. (..it's relative non-effect in the time-domain can be seen/measured in any number of basic designs.)
*IF* you add-in a waveguide however, then you've just added-in a potentially significant source of diffraction: mouth exit to baffle transition.
The real problem isn't diffraction (for most designs at treble freq.s.) - its *reflection*.
A high amplitude reflection at a baffle's edge can create an apparent secondary source significantly off-axis. Particularly disturbing are those horizontally opposed to the tweeter. It can screw with lateral image localization, and to a lesser degree with image localization in respect to depth - when it's effect is somewhere in the 1.5-8 kHz region. It also rarely shows up (measurably) except in wide-angle polar plots. It is of course much less of an issue if you are listening on-axis. (..though we do hear with a degree of polar uniformity at these freq.s - so it does effect the perception on-axis, but to a much lesser degree depending on the amplitude of the reflection.)
The real problem is diffraction. What you are calling "reflection" seems to be nothing more than a semantic version of "diffraction". The effect at a boundary edge in all the literature I've seen that creates the effective "secondary" source is part and parcel of diffraction. Distinguishing it as a reflection confuses the issue.
I'll point out a few references that I quickly found, two in the AES anthology "Louspeakers: Volume 3":
1. "Application of the Geometric Theory of Diffraction (GTD) to Diffraction at the Edges of Loudspeaker Baffles", R.M. Bews and M. J. Halksford (ppg. 107-115)
2. "Computation of Diffraction for Loudspeaker Enclosures", Juha Backman (ppg. 189-198)
A third reference I could find is "Testing Loudspeakers" by Joe D'Appolito, page 59, the section entitled "Edge Diffraction".
I don't know if the paper by H. F. Olsen in his pioneering work used the term diffraction, I haven't seen the entire paper, but from all of my reading it's all in the category of diffraction. To distinguish it separately as a reflection, making it essentially a separate phenomenon from what some might call diffraction, does not conform to any references I've seen. All software that models the response due to baffle edges uses, I believe, something along the lines of the GTD. It's not a separate phenomenon (i.e. not a reflection). It's all a continuum.
I suppose that one could argue that any incongruity in the expansion of the wavefront is a "reflection" or "diffraction", but I see one distinction only. What I see as the distinguishing characteristic at the baffle edge (or nearby driver that is an effective cavity) is that it allows a sudden drop in pressure when the wavefront first reaches it, causing an inversion of the signal at that point. It is effectively a time delayed "source" that is 180 degrees out-of-phase with the main wave at that point. It's an inversion, to use D'Appolito's terminology.
It is also not restricted to the off-axis. The maximal impact varies significantly with baffle shape, driver position on the baffle and the driver's directionality.The off-axis is often impacted less than the on-axis, especially for symmetrically positioned drivers.
In the end it is still semantics, but there's enough confusion in audio without calling the same phenomenon by two names. To me and in the literature with which I'm familiar, the baffle edge influence is, as far as I have seen, uniformly referred to as diffraction.
Dave
The lower spectrum of any driver will have wider dispersion. Thus diffraction waves will be caused in this region of the driver when mounted on a baffle. In the higher frequencies, since most of the energy is in the foward direction, it will not create strong diffraction waves at the baffle edge. This seems to be the same with horns/guides in the open.gedlee said:
Diffraction and reflection are inseperable effects since whenever diffraction occurs there will also be a reflection. The reflection part propagates back to the source and the diffraction part away from the source. However, in most of the literature these effects are simply lumped into a single term - diffraction - since "back to" and "away from" can often be ambiguous. Hence it is common just to refer the entire concept as diffraction.
Somewhat off topic as it relates to Coax drivers with CD's. Coax HF drivers all seem to exibit quite a ripple right after the XO freq that doesn't appear as drastically in the fixed WG applications being discussed here. Does anyone think that adding foam under the dustcap at the transitional area might help to smooth the response or does the use of foam only apply to HOMs? I know the question is quite broad and may assume too much but any info would be helpful.
I was specifically referencing *reflection* (..in that sniped quote of mine). (..which is why I formatted my reply with 2 paragraphs - one for diffraction and one for reflection.)
(Physically) consider a tweeter "proud" of the baffle and it's tendency to radiate beyond 90 degrees (+/-) lower in freq.. Specifically NOT considering a bounded wave, but rather higher freq.s that still have this extreme angle of incidence.
Also, consider the "omni" nature of diffraction from a typical cone surround - which is also usually "proud" of the baffle. Think about the freq.s for this point of diffraction. Now consider the radiation of those higher freq.s in relation to the baffle and it's edges.
Consider the addition of a "grill cloth frame".
Finally, consider the reflected wave itself. Does it react exactly like diffraction? Does it interact with the primary wave the same?
(Physically) consider a tweeter "proud" of the baffle and it's tendency to radiate beyond 90 degrees (+/-) lower in freq.. Specifically NOT considering a bounded wave, but rather higher freq.s that still have this extreme angle of incidence.
Also, consider the "omni" nature of diffraction from a typical cone surround - which is also usually "proud" of the baffle. Think about the freq.s for this point of diffraction. Now consider the radiation of those higher freq.s in relation to the baffle and it's edges.
Consider the addition of a "grill cloth frame".
Finally, consider the reflected wave itself. Does it react exactly like diffraction? Does it interact with the primary wave the same?