Yes, sorry for the TLA (three-letter-acronym)
This article has a series of measures on enclosures with round overs & champhers from none to at least 4". Was very informative.
dave
I'm pretty sure you are talking about
"Acoustic Diffraction: Does It Matter?" James Moriyasu ( Feb 2003 )
Also
In the Master handbook of Acoustics ( F. Alton Everest ) has an excellent chapter on Diffraction that includes a section on Diffraction by loudspeaker cabinet edges
It depicts analysis of Vanderkooy's calculations on baffle diffraction
This has some informative material
True Audio TechTopics: Diffraction Loss
And references this very informative Speaker Builder Magazine article
by Kral, Robert C.
"Diffraction - the True Story"
1/80, pp.28-33
Syd
"Acoustic Diffraction: Does It Matter?" James Moriyasu ( Feb 2003 )
Also
In the Master handbook of Acoustics ( F. Alton Everest ) has an excellent chapter on Diffraction that includes a section on Diffraction by loudspeaker cabinet edges
It depicts analysis of Vanderkooy's calculations on baffle diffraction
This has some informative material
True Audio TechTopics: Diffraction Loss
And references this very informative Speaker Builder Magazine article
by Kral, Robert C.
"Diffraction - the True Story"
1/80, pp.28-33
Syd
Last edited:
What are your opinions on using blankets on the front
baffle like felt e.g. ?
Felt or foam especially near the speakers can have much
influence on how a speaker sounds IMO.
There was a time when i used it extensively, but i somehow
got rid of it. Not by its looks only, but also sonically. Difficult to say.
There are some manufacturers which use absorbtive materials
as an integral part of their designs - Duntech e.g.
Can it substitute roundovers ?
baffle like felt e.g. ?
Felt or foam especially near the speakers can have much
influence on how a speaker sounds IMO.
There was a time when i used it extensively, but i somehow
got rid of it. Not by its looks only, but also sonically. Difficult to say.
There are some manufacturers which use absorbtive materials
as an integral part of their designs - Duntech e.g.
Can it substitute roundovers ?
i used the largest router bit i could, 32mm (1 1/4") radius:
which seemed to help a bit with the top end, here is the EQ'd waveguide response, (no smoothing) at 7.5deg intervals off axis. (70 deg guide)
An externally hosted image should be here but it was not working when we last tested it.
which seemed to help a bit with the top end, here is the EQ'd waveguide response, (no smoothing) at 7.5deg intervals off axis. (70 deg guide)
An externally hosted image should be here but it was not working when we last tested it.
You can't seperate the two. The frequency response roughness is defined by the strength and delay of the secondary sources. The interference pattern is the frequency response roughness evaluated over a range of angles.
David
Except that you can have perfect axial response with multiple strong interference sources, which is often true in "On axis optimized" box speakers. A lot of speakers are designed with far too much weight on on-axis response linearity, at the cost of off-axis smoothness.
I experimented once with using a barrier reflection to cancel an edge reflection. Generally edge reflections (from diffraction) are out of phase with the impinging signal and barrier reflections are in phase, so if the timing is right and the relative strength is right you can get one to cancel the other. (At a point in space.)
Still, it isn't a practical method and I haven't seen anyone doing it, at least not commercially.
So in practice I don't believe you will see "perfect axial response with multiple strong interference sources". Due to typical symmetry in cabinets the tendency is to have the roughest response on axis and, due to scattering of reflections, to have smoother response off axis. Tweeters on a circular plate are a typical example: reflections from the circular edge will be a worst case (all the same distance/time away) but moving off axis will spread the reflection energy and reduce any axial ripple.
I agree with the sentiment that some compromise should be made between on and off axis smoothness. Getting the best average response over a likely listening window has always been my goal.
David
I'm not saying that the on-axis response is least affected by the ripple, rather that it's attempted to be corrected in the crossover by some designers. Which of course with this sort of thing ONLY works on the design axis.
In any case, no reason not to take every effort to squelch the diffractions in the first place, unless you don't have the tools to do it or are working towards a commercial pricepoint.
One of the first speakers i built, had a rather flat cabinet with
rounded edges. When placed in front of a wall it simulated flush mounting.
Drivers were placed rather asymetric and its homogeneous sound
was due to keep severe diffraction at least out of the midrange, i am
convinced.
But problem with that cabinet was large baffle and back area, which made
cabinet resonances audible ...
Last edited:
That's where I went next.
Since the problem is how to soften the transition from 2pi to 4pi loading on a small baffle; Placing my drivers in-wall dropped that transition point and greatly helped.
( Thanks to Google Book preview, portions of Electroacoustical Reference data
By John Eargle dealing with these topics of discussion are available. )
Syd
Other diffraction solutions
Hi al meta,
Which wave guide and driver are you using? How do you like them?
This is from my previous speakers. 4" PVC pipe cut in half length ways and taped to the cabinet edges. Looks crappy but improved the sound.
I'm trying out different horns and waveguides now on a driver. So I bought a couple of kids swimming foam noodles about 3" thick and cut them in half length ways, too. I tape them to the outer edge of the horn and use black Gorilla tape to stick them on. Cheap and effective for most diffraction.
Hi al meta,
Which wave guide and driver are you using? How do you like them?
This is from my previous speakers. 4" PVC pipe cut in half length ways and taped to the cabinet edges. Looks crappy but improved the sound.
I'm trying out different horns and waveguides now on a driver. So I bought a couple of kids swimming foam noodles about 3" thick and cut them in half length ways, too. I tape them to the outer edge of the horn and use black Gorilla tape to stick them on. Cheap and effective for most diffraction.
Last edited:
This is from memory, but the experiment was with a bookshelf speaker with the tweeter near a top edge. The reflection from the unradiused cabinet edge was putting the usual ripple into the HF response. I think I put a 3/4" to 1" tall barrier, like a small "wall" protruding off the baffle and a similar distance to the side of the tweeter.
The idea was that the negative reflection from the top cabinet edge and the positive reflection from the side dam would be same timing, same strength, and opposite phase.
I remember it working pretty well, at least for a stationary microphone position.
David
This is a good point to make. These baffle edges don't have to be solid MDF. You can tape on cardboard or styrofoam and manipulate it into bevels or radiuses and see the effect at mid and high frequency. I have often wondered if a particular edge was causing a response abberation. Tape on some cardboard (I like the thin stuff off the back of a pad of paper) and you can fold it forward as an extension of the baffle, then back at 90 degrees to recreate the edge. At mid and high frequencies these are sufficiently solid barriers to show the full effect.
To answer the original question, one of you enterprising chaps can put some 2" styrofoam on the edges of a cabinet and sand a little, take a measurement, sand a little...
Who has some time this weekend?
David
hi,
your extended radiused baffles look like a good size to help, have you managed to take any sweeps and see how much difference they make?
the waveguide is my own concoction based on geddes OS but with the roundover being continuous from guide into cabinet, it is a 12" waveguide with the roundover making it more like 15" OD and almost tracking a le'cleach mouth termination as far as practical in the cabinet, extending the roundover further.
as seen here:
how to make a circular waveguide: in pictures - Speakerplans.com Forums - Page 1
i used a bms 4538, and the combination is great, it's true that a waveguide sounds very nuetral without the typical 'horn sound', you wouldn't know there was a guide there if you couldn't see it, and this driver can get up to 20k very cleanly as seen on the graphs, with an on axis dip at around 4k, which is a small part of the soundfield when they are 70 deg -6dB dispersion.
i think we are always constrained by how practical it is to make large roundovers, with relation to their benefit, but it would be interesting to see how they correspond between radius and wavelength in the real world.
TBH i just went with what seemed right and was practical, which included making the cabinet the shape it is to minimise large flat areas and sharp corners, and it seemed that jmmlc's style mouth termination helps alot.
i'm very happy with the results, especially as it was my first go at fibreglssing and quite tricky with the router for the cabinet.
al
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.