crossover/attenuation guidance?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ex-Moderator R.I.P.
Joined 2005
So, you got your answer amongst other things ... you can expect your speaker to be around 80db, maybe a little more ... thats the answer which you didnt like, but it is a fact

You will loose sensitivity in bass with isobaric, with or without BSC ... but I am not sure how the midrange will be

You may just as well use only one woofer, or maybe with a passive slave

Its not the simple answer, but all part of the sensitivity issue
sreten said:

FWIW after implementing baffle step the allegedly 94dB tweeter
will need attenuating 12dB or so to match the BSC equalised
82dB or so sensitivity (from 87dB) of the isobaric array.


Sreten answered your question right there. Forget the iso setup. This isn't a subwoofer. Assuming a conventional arrangement adn the woofers in parallel: 87db + 6 dB = 93 dB @ 2.83v Now baffle loss will result in 93dB - 6 dB = 87 dB at the lower frequencies. Because of the loss we institute what's known as Baffle Step Correction, which means dropping the upper frequencies so that response is flat. Full BSC would be 6 dB, resulting in an 87 dB @ 2.83v across the woofers' response, in this case the tweeter would need to be dropped 7 dB (94-87=7). Understand? Now you may not opt for full BSC, maybe you go with 4 dB instead. Now the woofers would be about 89 dB in the crossover area, resulting in 5 dB drop of the tweeter.

I know diy has its appeal, but like many things in life, sometimes we must admit we may not have the aptitude to actually do it. I'm sure there are plenty of commercial designs that you can buy at your local dealer that you would be happy with.
BSC is not a discrete circuit that you can "see", it's built into the XO design, even a first order one, by any manufacturer worth his salt.

I guarantee the B&W implements BSC. Thawach you seem very enthusiastic about this hobby, you should check out some of the technical literature to get a good scientific grounding. Loudspeaker Design Cookbook or High Performance Loudspeakers would be a good place to start.

I know you may be right.I suggest him that's not too difficult. He may be do or not . He thinks by himself. But you tell him may be difficult. And tell him to buy. I ask you. Is it hobby? The thing that I do. What am i wrong? About the book don't tell me. I have a lot more than you think. Please stop you mouth. And go away.
If you are not stop reply. I will reply forever.

Hello. sreten
I am sorry to do that with you. I know that you are a good gentle man. you try to help everyone to know.You are the best man really. Today is a christmas day.I hope you are happy.
And 1 january 2008 is a new year(Thailand). I hope everyone on this broad to be happy(send before).

Here is the link about baffle step correction simulator
that i ever used.
Hear is new epos speaker.

Someone tell me please. i don't understand why the epos sound engineers don't interest to design baffle step circuit in those speakers.




  • epos.jpg
    84.7 KB · Views: 190
I plan on crossing them over at 3500Hz, 12dB per octave, but how much will the tweeter need to be attenuated?

keeping in mind, that the two woofers will be wired in parallel, to give 4 ohms, however, they will be loaded isobarically...

does this mean the sensitivity of the woofers, with 2.83 volts will be the same as a single woofer?

First off, you are crossing a woofer that the manufacturer claims extends up to 3khz at 3.5khz, second you are crossing a tweeter where the manufacturer recommends a second order at 4khz at 3500hz. I doubt the woofer is clean up to 3khz as most manufacturers over-extend the upper limit of their woofers for marketing purposes, the "clean response" is usually significantly lower and/or requires a steep xover to reach the claimed upper limit. When a company tells you to crossover ~1.5 octaves above the lower limit there is probably a reason, whether it is distortion or rough/sloping response you might want to trust them on it.

Second, I can't believe everyone is so hung up on the SPL here, the big problem you are going to face is that you are trying to reproduce the bulk of your sound through the frame of the woofer! This is not going to sound good, feel free to listen to a wide band woofer through the front and rear, better yet measure the frequency response both on and off axis.

Isobaric loading is very limited in its usefulness, it reduces box volume by 1/2 (two woofers,clamshell mount), and it cancels* suspension and motor nonlinearities due to the mechanical phase inversion.

* cancels but its not necessarily 100% perfectly canceled,

As for spl, you obviously don't understand the difference between efficiency and sensitivity, cool, just be respectful of those who try and explain it.
How much do you need to pad down the tweeter?
First as was said before if you apply 2.83V into an 8ohm woofer or tweeter the power is 1W, from V^2/R, 2.83^2/8 ~=1 (+-roundoff error) now if you add a second 8ohm woofer (not isobaric yet, later for that) and wire it in parallel to the first you will gain +3dB from doubling the radiating surface area, and +3dB from halving the impedance of the woofer section V^2/Rp or 2.83^2/4 ~= 2W.
So if you apply 2.83V to the 2way system you get 1W and 94dB from the tweeter and 2W and 93dB from the woofers a nice match.

But now you have to factor in the baffle step phenomenon, which will have a low frequency where the system response begins to rise and a high frequency where it levels off (full transition into half space radiation). These frequencies are simple and can be analytically determined for round baffles with no edge radius, and the final result at and above the high freq. the response is +9-10dB higher than at and below the low frequency as defined above. For rectangular baffles the baffle step effect is reduced and spread out over a range of different frequencies whereas the round baffle had one high and one low freq.
I suggest you dl the baffle step simulator (Edge?) for a graphical representation of how this works.
FYI,the f3 going from high frequency to low is analytically/empirically (not sure)determined to be f(3) = 380/W(B) (where W(B) is the baffle width in feet), all of the equations are based on an approximation of a spherical enclosure response ( I don't fully understand that part but who cares) the important thing is to learn how to measure the in box response and design a circuit to deal with the actual reduction. To illustrate why this is important lets look at the above eq with the above two woofers, if they are mounted on a 1 ft wide baffle (for convenience) then 380/1=380Hz.
So 380Hz is the approximate midpoint for the baffle step issue, this means that the music below 380hz will have 1/4 of the energy of that above 380hz (theoretically 1/4, but that's why you must measure it).

Now, what happens with isobaric? First, since you are taking two independently mounted drivers operating in the baffle step range and mounting them onto each other you will now have more pronounced baffle step response (theoretically) and you now lost half of your radiating area (-3dB). So while each woofer still produces 87dB with one watt of input they are not radiating independently, so 2.83V in and 4 ohm impedance= 2W to each woofer and 90dB output (total). This is 2W to the Woofers and 1W to the tweeter. Now with 3-6 dB of loss from BSC you will have to pad the tweeter down 7-10dB depending on final BSC level and where you install the BSC circuit, if it is before the power amp ( line level) then you wont need to pad the tweeter to account for BSC, if you only apply the BSC to the low pass circuit, then you will need to pad the tweeter down accordingly.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.