Open baffle bass leakage

[DanielT]

A speaker element placed in the middle of an open baffle. How wide does the open baffle need to be so that frequencies down to 300 Hz are supported by the baffle and not extinguished? At what baffle width are they extinguished?

How do you calculate this? Scenario a loudspeaker element centered on the center of an open baffle. 300 Hz break off point. Do you then count on half a wavelength when calculating baffle width?

Do you calculate the distance of the baffle from the center, the middle of the speaker cone, or from the edge of the speaker cone?

 

[Don Hills]

I think this page should answer most of your questions:
https://www.linkwitzlab.com/models.htm

 

[DanielT]

I think this page should answer most of your questions:
https://www.linkwitzlab.com/models.htm

Thanks, that's nice of you, but can't you just tell me how I'm supposed to do it? It is about wavelength at a certain frequency combined with baffle width. It should be a fairly simple calculation. Or?

 

[DVDdoug]

Here is a wavelength calculator. At a half a wavelength the waves are 180 degrees out-of-phase, so that's a "round trip" from front the back. Then as you go up in frequency 1.5 wavelengths will be out-of-phase again, etc. But at higher frequencies all of the reflected waves in the room combine in more complicated & unpredictable ways.

 

[DanielT]

Here is a wavelength calculator. At a half a wavelength the waves are 180 degrees out-of-phase, so that's a "round trip" from front the back. Then as you go up in frequency 1.5 wavelengths will be out-of-phase again, etc. But at higher frequencies all of the reflected waves in the room combine in more complicated & unpredictable ways.

"At a half a wavelength the waves are 180 degrees out-of-phase, so that's a "round trip" from front the back"

Okay if we assume that. A speaker element placed in the middle of a baffle, say a 10 inch element. From where, how do you calculate baffle width at frequency 300 Hz before it is cancelled? From the middle, center of the speaker cone or from the edge of it?

We ignore medium and temperature. 300 Hz , 1.1 meter wavelength, half that 55 cm. BUT does that mean a 55cm wide baffle, with the 10 inch driver ,example above? I assume so but I still have to ask.

 

[sam_adams]

BUT does that mean a 55cm wide baffle, with the 10 inch driver ,example above?

Shapes other than circular compound the issue because the radius, 'b', is not constant.

 

[Plcamp]

before it is cancelled

It isn’t ‘cancelled”. It rolls off in efficiency as you lower the frequency below the baffle width. You can always equalize it back up, noting the xmax limits of your driver. That’s why you see big cone area drivers for bass…to raise efficiency back again. Mine are twin 15” woofers on 21” wide baffles with a 8” fullrange driver acoustically crossed in at 650 hz. Could use an 8” at lower frequency if a separate tweeter handled highs.

Some folks use no baffle at all.

 

[DanielT]

I stop beating around the bush. Like this. I have two Faital Pro broadband elements. And a pair of Peerless SLS10 bass elements in boxes. Now I thought I'd test my Faital Pro together with SLS10. That in the form of an open baffle, for the Faital Pro. Fun to test dipole, I thought.

I already have the SLS10 in closed speaker boxes. I will not change them. Crossover I can set as I want, but I prefer not to push the Faital Pro below 200 Hz nor the SLS 10 above 500 Hz.

4FE32 (8Ω) by FaitalPRO

FaitalPRO combined the latest R&D methods and technologies with the most advanced industrial automation techniques and all FaitalPRO drivers are made in Italy with the highest level of quality, consistency and durability.

faitalpro.com

Crossover filter, 24 dB slope with an active crossover filter so I can choose the crossover frequency myself. Baffle width, not so important but not too narrow. It might as well be, almost, "unnecessarily" wide.

This is what the SLS bass boxes look like. It's a bit messy in the picture, because it's....messy. I'm testing a bit but anyway on the SLS10 I thought I'd try placing the Faital Pro.

 

[Don Hills]

As others have said, when you have a real world driver that has finite diameter, things get complicated. You no longer have one length from front to back. You have a range, from near edge to near edge through to the distance from far edge to far edge. For a 10 inch driver on a 20 inch diameter baffle, the distances therefore range from 20 inches to 60 inches. I refer you again to Linkwitz' page, specifically section A3. The graphs there show what happens as you vary the baffle diameter from 0 (just the bare driver, or b=a) to 8x the driver diameter. So there's no "right" size for the baffle. And as is pointed out in section B, a circular baffle is less than ideal. A square or rectangular baffle will smooth the response. Lastly, section D discusses conditions specific to your case of a midrange driver on an open baffle.
So start with the desired lowest frequency (300 Hz?) and the speed of sound , about 13500 inches per second. 13500 / 300 = 45 inches, so a 45 inch diameter baffle. If that's wider than you want, use Linkwitz' graphs to see the effect of making it narrower.

 

[voodooless]

Use this:

Home of the Edge

 

[Juhazi]

Daniel, The Edge simulations will help you with calculations, also off-axis, but you must understand some basic aspects of dipoles

"dipole loss" happens when front/back soundwaves meet in phase and cancel each other. This cancellation overrules edge diffractions and give dipole 8 radiationwith

constant DI (directivity index). This also means sloping down axial response, which must be compensated by equalization 6dB/octave.

Above "dipole peak" peak frequency, the driver behaves like a monopole when we look at spl response, We don't need eq any more, but we dont have constant dispersion either. Also backside radiation/dispersion loses similarity to frontal dispersion when one uses a coil/magnet motor driver (typical cone driver like Peerless SLS series.

Most people designing and using "open baffle" speakers don't seem to understand this at all. And this is why a really good wide ranges dipole loudspeaker must be 4-way with different drivers and baffle width for each way. And now, if you want to try dipole bass, you must use appropriate baffle width with 10" woofers, perhaps U shape if you use deep filter slope to avoid tube resonance... IMO open baffle's "fast and clear bass" is a myth, or at least very much room-dependent and requires very large rooms and speaker distance to walls of several meters (like SL had at his house)...

www.dipolplus.de by Rudol Finke is very good advice

And here axial response of 400 vs. 2000mm wide baffles. In reality, you need infinite baffle to avoid dipole loss in bass, but you won't get dipole radiation then either.

 

[DanielT]

As others have said, when you have a real world driver that has finite diameter, things get complicated. You no longer have one length from front to back. You have a range, from near edge to near edge through to the distance from far edge to far edge. For a 10 inch driver on a 20 inch diameter baffle, the distances therefore range from 20 inches to 60 inches. I refer you again to Linkwitz' page, specifically section A3. The graphs there show what happens as you vary the baffle diameter from 0 (just the bare driver, or b=a) to 8x the driver diameter. So there's no "right" size for the baffle. And as is pointed out in section B, a circular baffle is less than ideal. A square or rectangular baffle will smooth the response. Lastly, section D discusses conditions specific to your case of a midrange driver on an open baffle.
So start with the desired lowest frequency (300 Hz?) and the speed of sound , about 13500 inches per second. 13500 / 300 = 45 inches, so a 45 inch diameter baffle. If that's wider than you want, use Linkwitz' graphs to see the effect of making it narrower.

Thanks for the tips and advice. Section D you say. I'm looking into that.

"A driver becomes directional of its own, when its effective piston diameter becomes larger than 1/3 of a wavelength. For an 8" driver this would be above 558 Hz. "

Then, as in my case with a 4" drives 1116 Hz.Or maybe it was too easy for it to be that way?

This was a good animation:

"Measured Data for a Real Loudspeaker
The animation at right shows actual directivity data measured for a 4-inch boxed loudspeaker.[1] At low frequencies (250 Hz) the speaker radiates sound equally well in all directions. At higher frequencies (10 kHz) the speaker radiates all of its sound in front - the sound level behind the speaker is almost 25 dB lower than the level in front, indicating that much more sound energy is being radiated directly in front and very little behind ."

Use this:

Home of the Edge

Thanks for the tips! Looks good, I'll test and see if I can get that program to work.

 

[Juhazi]

And here simulation of an ideal nominally 4" planar driver in 300mm open baffle

If one uses it above dipole loss range, response will be interesting... And the Faital is far from planar!

 

[DanielT]

And here simulation of an ideal nominally 4" planar driver in 300mm open baffle

If one uses it above dipole loss range, response will be interesting... And the Faital is far from planar!

View attachment 228821

Aha! Interesting and at the same time it looks absolutely crazy fluttering FR... response will be interesting. It was put diplomatically. I trade interesting for crap.

It looks interesting that Tolvan (Twelve)

Home of the Edge

I just have to install it and work with it.

Edit:
Thanks for your simulation!

Edit 2:
Hm, just for a test. It increases my motivation to get UMIK up and running. I can probably find some wooden boards in the free scrap heap at the wood supply store, so it won't cost anything. It might be 60 cm wide or 100 cm. I see what is there. A little sawing, a few screws, glue and braces are easy to fix.