baffle step elimination (would this work) (1 Viewer)

[Ron D Core]

If I were to elevate the speakers about .25" off the actual baffle with MDF rings, and glued a .25" thick piece of sound absorbing foam to it (so its flush with the speakers), would it help reduce the baffle step and diffraction? Maybe I should draw it out. I saw something like this at a local hi-end audio store. The mid-range and tweeter both had a piece of foam mounted on the baffle around them. The foam peice was about .25" higher than the drivers. The woofers had nearly flat cones and they even had foam on them. It seemed odd, but the speakers sounded great.

 

[Pete Mazz]

TW does this alot with tweeters. He uses thick felt. It'll smooth out the HF response. I have the felt, but haven't tried it yet.

Pete

 

[Greg Monfort]

It helps with early reflections (a good thing), but has negligible impact on baffle step diffraction. This is a function of baffle width so either the baffle needs to be wide enough, is dealt with electronically, or a second driver is used and rolled off to suit.

GM

 

[Ron D Core]

Will a narrower baffle help with anything? The Focal drivers have straight edges, making it easy to bring the edge of the baffle closer to the speaker.

 

[Patrick Sun]

Common wisdom is that a narrow baffle helps with imaging (but there's more to it than just a skinny baffle, you have to get the integration of the drivers' crossover network and physical placement on the baffle (keeping them close together) to work in their intended frequency range to smooth out the off-axis response as well).

 

[Greg Monfort]

>Will a narrower baffle help with anything?
====
It will require more baffle step compensation than a wider one.
====
>Common wisdom is that a narrow baffle helps with imaging....
====
At the expense of a flat FR. A narrow baffle with electronic compensation has no better imaging than a wider one that's been damped. Indeed, most folks I've known prefer the performance of the damped wider baffle since there's fewer electrical components in series with the driver.

GM

 

[Hank Frankenberg]

Whoa, I'm learning something here. All the "wisdom" I've read over the last 10 years, in Speakerbuilder mag and elsewhere, has been favoring the narrowest possible baffle (with closely spaced drivers) and many commercial designs seem to have gone that route, NHT for example.

So Greg, you're saying bottom line that narrow is best for imaging, but requires more crossover components which messes with FR? Let's keep this discussion going.

 

[Jon Hancock]

Hi Hank,

I'm going to stir up the pot and throw a few points into the discussion.


Baffle step occurs when a loudspeaker transistions from radiating into a a half space (forward from an "infinite" plane) to full space. Because the distance most speakers are from adjacent boundaries is large compared with the size of the speaker baffle itself, the effect of the floor as an additional boundary is a much lesser affect. This behavior is true for any speaker with a finite size baffle, operating in a stand alone enclosure or baffle. The drop in the response as you move from half space to full space radiation is 6 dB, though at lower frequencies it will be modified by the presence of other boundaries, such as a rear wall or floor.

Some speakers are designed to always radiate into a half space- such as the series from Allison Acoustics, or most "professional" monitoring speakers. The latter are typically built flush into the walls of the studio. Then, no baffle compensation is required or used.

Diffraction is an effect which occurs when the wave radiating from an acoustic source (loudspeaker) mounted on a baffle reaches the impedance discontinuity at the edge of the baffle, and causes a secondary radiation launch, delayed in time from the primary source.

The use of small cabinets will have two effects:

1) it moves the baffle step transistrion frequency to a higher frequency range

2) it moves the diffraction induced frequency response errors to a higher range.

For any free standing cabinet, baffle step compensation will have to be used to achieve flat axial response; changing the cabinet size only changes the frequency tuning of the crossover to achieve this, (due to the change in the frequency from which the cabinet transitions from being a half space radiator to full space), but it doesn't change the required number or type of components. With intelligent network topology design, no additional components are needed anyway.... For a typical large bookshelf or tower 12" wide, the basic baffle step drop starts around 1 kHz, and will be finished down around 200-250 Hz. To change these frequencies significantly requires doubling or halving the dimensions.

For a small cabinet, the diffraction effects which are moved up in frequency can be minimized by using techniques such as large radius corners (which minimizes and distributes the acoustical impedance discontinuity), and damping materials on the baffle. Compensating for baffle step in a midrange or tweeter used on a small panel can be more difficult than doing the same thing for the operating range of the woofer on a larger panel. OTOH, a large panel speaker can be built to manage directivity and diffraction effects quite nicely, also. Check out Dunlavy's, to see what I mean.

Regards,

Jon

 

[Dan Wesnor]

This is a bad idea. Essentially, you are creating a round baffle the same radius as the driver frame - the frame becomes the baffle. This will cause worst-case diffrection.

 

[Hank Frankenberg]

Jon, thanks for kickin' it up a notch. Now we're cookin!
I'm with you on the edge defraction distortion occuring at a higher frequency range with narrower baffle width, which is what I have assumed is at least part of the reason for the trend to narrowest possible baffles in recent years. Better to have distortion that you have to live with, moved up out of the midrange or as far up into the midrange as possible, IMO.

You're saying that the crossover component count isn't really necessarily affected by this. Okay, I understand that also.

Rounded-over corners: I've used a 3/4" radius roundover bit and I have a new 1 1/4" monster bit (MLCS also has a 1 1/2"!). I don't believe that anything smaller than 3/4" helps at all, maybe not even 1". What's the expert opinion on this?

Hi Dan, what's a bad idea - rounded cabinet edges? Please explain.

 

[Jon Hancock]

Hi Hank,
There's no substitute for making your own tests and and measurements, but consider getting that 1-1/2" bit- a good working radius for diffraction control by shape alone is in the 1-1/2" to 2" range, IMO&E.
I fully agree with Dan's comments- this is why a round cabinet or round baffle is basically a no-no- the diffraction affect is not distrubted in time, but occurs all at once.
To play with these issues and learn a bit more, I strongly endorse playing with BDS (Baffle Diffraction Simulator). It requires a late model version of Excel, though.
BDS homepage
Best regards,
Jon

 

[Ron D Core]

Now I'm confused. So is a baffle with large round edges a good thing or a bad thing. I was thinking about constructing a rounded corner with a radius of about 3 inches with some 6" PVC fireline (really thick).

 

[Greg Monfort]

>So Greg, you're saying bottom line that narrow is best for imaging,
====
Yes, in that without compensation, the FR has a rising rate so the harmonics that gives us the sense of space are emphasized.
====
> but requires more crossover components which messes with FR?
====
As JH noted, if the design incorporates it from the get-go then no, but it seems to me that more often than not with DIYers and some manufacturers it's added as an afterthought. It used to be common for them to just add adjustable L-pads to the tweeters even though they often were XO'd too high to get the job done properly. The few current mini monitors I've auditioned don't sound like they even bother to account for it. I assume because you're expected to sit inside 1m from them.
====
>This is a bad idea. Essentially, you are creating a round baffle the same radius as the driver frame - the frame becomes the baffle. This will cause worst-case diffrection.
====
The frame and/or surround is one already. Anyway, the foam damps it down in his example if the right density is chosen, though I wouldn't bother with the extra work of either recessing or raising it. Placing the foam on top and overlapping the faceplate works for me.
You're right though in that round baffles are bad unless the driver is offset on it to average out the standing waves over a wide BW.
====
>I'm with you on the edge defraction distortion occuring at a higher frequency range with narrower baffle width, which is what I have assumed is at least part of the reason for the trend to narrowest possible baffles in recent years. Better to have distortion that you have to live with, moved up out of the midrange or as far up into the midrange as possible, IMO.
====
Well, I'm not. When you move it up into the BW where our hearing is most efficient it becomes even more of an imperative to damp it out by some means. An advantage of the higher transition point is that with increasing frequency acoustic energy decreases, so it's easier to attenuate, and I assume is one of the reasons for the 'shrinking' baffle.
====
>So is a baffle with large round edges a good thing or a bad thing.
====
Good, though this wasn't your original question.
====
> I was thinking about constructing a rounded corner with a radius of about 3 inches with some 6" PVC fireline (really thick).
====
Sounds like a plan, just be sure and fill the voids with something inert so there's no possibility of any cavity resonances.
GM