Baffle Step question

[mazeroth]

I just purchased a pair of the Hi-Vi B3N 3” full-range drivers to make a pair of test computer speakers, after reading John Krutke’s review of them here:

http://home.new.rr.com/zaph/audio/minitest/

Being new to speaker building I just dug deep into baffle step correction last night and really got a grasp on how it works. I even tested my Altec Lansing computer speakers, which also use 3” drivers. The front baffle is about 5” wide so I would listen to it for a few seconds normally then put a CD jewel case on each side of it and BAM, it sounded so much fuller! I’d love to employ a full BSC circuit but figuring the costs it would almost run as much as the Hi-Vi driver ($9.50) and I do plan on making 5 sets of these for my whole family for their computers. Here comes my question:

Jim Salk has a nice page about BSC and the wavelengths at different frequencies. For example, 750hz is roughly an 18” wavelength. If I’m using a 3” speaker, theoretically I’ll need about a 36” baffle to keep 750hz and up away from needing BSC? (do I add in the 3”?) I’m not going to make a 36” baffle, but just trying to grasp the concept! If not 36”, how large? Also, say it IS 36” needed, if the baffle is that wide how much will I lose an octave down? Will 375hz be -1db, -2db etc…?

If these weren’t going to be mass-produced (if you can call 5 that heh) I’d probably do a 0.5 woofer or mount an identical one on the back (do I make it 0.5 as well?). I have searched on Google and can’t find a BSC calculator that asks you to input baffle width and speaker size to show you when the frequencies start to drop off. What I’d like to do, and help me out here, is find out where (say 1khz) and add a 0.5 woofer to the front and use a 1st order to cut the highs out of the 0.5 woofer. Anywhere on the web I can find a program that will tell me where to put the 1st order?

Sorry for all the questions and thanks for ANY input!

[rabbitz]

A good starting point for the rear 0.5 woofer xo is:
Baffle Step F3=115/Baffle width in metres. For your 5" baffle start off calculating the 1st order xo on 905Hz.

Calculate your xo using the driver's impedance at the xo point and not the nominal impedance as it will be more accurate.

[Svante]

I think you should skip the easy formulas based on baffle width, since there actually is a dependence on baffle height too. I would of course recommend you to try http://www.tolvan.com/edge , since I wrote it , but there are other programs too that you might want to try.

[mazeroth]

Thanks Svante, I'll give that a look in a minute!

Since these will be computer speakers and my desk is right up against the wall, will I even need to worry about a BSC circuit? Is there a rule that says you can be "x" distance away from a wall before needing to compensate?

Thanks!

[Grahamt]

If these are on the desk and right up against the wall, the speakers are pretty much firing into quarter space. You shouldn't need to compensate for BS.

[Svante]

If the speakers are 20 cm from the wall, the first minimum will occur near 345/(4*0,2)=430 Hz, above this frequency, the approximation "the loudspeaker is near the wall" will not hold. So, the solution is IMO not to skip the BSC completely unless the speakers are truly mounted "in-wall". The question is how much BSC there should be, and there probably is no simple answer for that, particularly if the speakers are moved a bit.

[planet10]

A good overview (if i do say so myself) with links...

http://www.t-linespeakers.org/tech/b...intro-bds.html

dave

[happy.gringo]

I have some of the Onkyo buyout 4X6 speakers with whizzers. The 4 ohm ones. There are no available specs on them, except that they are Zohm 4 ohms, RE 3.55 and fs is 105hz. I want to use two per box and make 2 triangular boxes to put in the back seat of my car. I want to wire them in series, but use one just for baffle step compensation and to boost what will be very weak bass in a sealed box. Using Rabbitz formula above, I calculated that the baffle step speaker should begin to roll off at 453hz for a 10 inch baffle. I have no way to determine the actual impedance at this frequency, but assuming that it is somewhere between 4 and 5 ohms, couldn't I just parallel an 80 microfarad capacitor with the first speaker, which would shunt all the signal to the second speaker above that frequency? Is there a better way to do this, considering I have no way to measure these speakers? Thanks

[planet10]

Originally posted by happy.gringo
to put in the back seat of my car. but assuming that it is somewhere between 4 and 5 ohms, couldn't I just parallel an 80 microfarad capacitor with the first speaker, which would shunt all the signal to the second speaker above that frequency?

You have the right idea, but in a car, with a "wall" backseat right behind the box you aren't likely to have any bafflestep. The driver will help at the lowest frequencies.

In the end you are going to have to play with the cap size.... you also have to consider that car amps typically like low impedance. At low frequencies your amp is going to put out les power (into 8 ohms) but will be more efficient, so there could be some synergy that works out.

dave

[soongsc]

Originally posted by Svante
I think you should skip the easy formulas based on baffle width, since there actually is a dependence on baffle height too. I would of course recommend you to try http://www.tolvan.com/edge , since I wrote it , but there are other programs too that you might want to try.

What do the "Edge Sources" and "Speaker Source Density" do? The more the better?