Baffled by baffle step

Dave Jones · 7th May 2004, 01:35 AM

Okay, I'm ready to cut MDF for a set of speakers. They will be fe107e (full range) in a box with width = 5.5", height = 16 7/8". The speaker is 4" in diameter, and will be placed 6 3/4" from the top of the box. I've calculated a baffle step correction circuit two different ways, namely, using a paper that I found on the net written by Martin King, and also using the edge.exe program. The first method give an inductor of 1.5mH, while the second gives 2.75mH.

HELP!

Svante · 7th May 2004, 02:26 PM

I think that the equation you used (4560/W, right?) originates in Olson's work in the 50's. He calculated these things for spheres and the results have since been applied to baffles of different shapes. The Edge calculates the response for a flat baffle, and also takes the height and width (or any shape) into consideration.

To me it appears obvious that a high baffle should offer more "support" than a square baffle of the same width, and so the f3 frequency should be lower. This would mean that the BSC calculation cannot be simplified to a formula that only contains the baffle width. Also a flat baffle should give more support than a sphere of the same dimension.

On the other hand, it appears as if the baffle step often is not fully compensated, since there is an effect of the room that also boosts the lower frequency.

I sure welcome a discussion on this!

planet10 · 7th May 2004, 06:10 PM

Quote:

Originally posted by Svante
(4560/W, right?) originates in Olson's work in the 50's

As far as i have been able to determine this equation is indeed pulled from the graphs Olson published. The baffles were either spheroids, cubes, or rectangles, some with significant 45 degree bevels. It is only a very rough starting point.

The new simulation tools coming out do a much more accurate job of accounting for the actual baffle shape.

dave

Svante · 7th May 2004, 06:32 PM

One reference that I see a lot on fora is this:
http://www.trueaudio.com/st_diff1.htm

It clearly uses the diffraction around a sphere to derive the function 380/W (W here in feet, 380*12=4560) and then magically the diameter of the sphere is assumed to have the same effect as the width of a rectangular baffle of any height.

Obviously, I wrote the Edge program, so you should be sceptic towards me, since I probably am biased. But as far as I understand there must be room for an error in the above reasoning.

Dave Jones · 7th May 2004, 06:43 PM

I've fiddled around with the F1, F2 fields in edge.exe some. When I set them at 350 and 700, I get an L of 1.8, and the curve looks at least as good as the "suggested" 2.6. So I'm going with 1.8 mH. Thanks for your help, and thanks for the nice program.

planet10 · 7th May 2004, 06:46 PM

Quote:

Originally posted by Svante
Obviously, I wrote the Edge program, so you should be sceptic towards me, since I probably am biased. But as far as I understand there must be room for an error in the above reasoning.

When is it coming for OS X? ... i just hate to fire up Virtual PC.

Is the Edge based on Juha Backman's paper?

dave

Dave Jones · 7th May 2004, 06:54 PM

Quote:

Originally posted by planet10

When is it coming for OS X? ... i just hate to fire up Virtual PC.
dave

Better yet, port it to Python/wxPython just once, and quit hassling over OS... uh... hassles.

planet10 · 7th May 2004, 07:08 PM

If i ever get time to write my OS X speaker simulator it will be in Python...

dave

SY · 7th May 2004, 08:04 PM

(scratches head)

Look, I love simulation programs (I use BDS, a wintel Excel-based program), but xover/eq parts are fast and easy to get. How about (radical idea here) putting the actual speaker in the actual baffle and actually measuring it before putting the xover/eq together? That way, you can order the exact inductor you need- or, more likely, find out that the eq circuit you need is something different than that predicted by any of the simulation software.

I think that baffle simulation is most useful for designing the baffles in the first place. You can choose a size/shape/driver positioning that minimizes hard-to-correct anomalies. The results won't be rigorously quantitative, but qualitative is all you really need at that point in the design process.

Bill Fitzpatrick · 7th May 2004, 08:27 PM

Quote:

Originally posted by SY
How about (radical idea here) putting the actual speaker in the actual baffle and actually measuring it before putting the xover/eq together?

Great idea but most don't have the bucks for the measurement software and hardware.

7th May 2004, 01:35 AM	#1
Dave Jones diyAudio Member Join Date: Jul 2003 Location: Silicon Valley	Baffled by baffle step Okay, I'm ready to cut MDF for a set of speakers. They will be fe107e (full range) in a box with width = 5.5", height = 16 7/8". The speaker is 4" in diameter, and will be placed 6 3/4" from the top of the box. I've calculated a baffle step correction circuit two different ways, namely, using a paper that I found on the net written by Martin King, and also using the edge.exe program. The first method give an inductor of 1.5mH, while the second gives 2.75mH. HELP! __________________ Davy Jones

7th May 2004, 02:26 PM	#2
Svante diyAudio Member Join Date: Feb 2004 Location: Stockholm	I think that the equation you used (4560/W, right?) originates in Olson's work in the 50's. He calculated these things for spheres and the results have since been applied to baffles of different shapes. The Edge calculates the response for a flat baffle, and also takes the height and width (or any shape) into consideration. To me it appears obvious that a high baffle should offer more "support" than a square baffle of the same width, and so the f3 frequency should be lower. This would mean that the BSC calculation cannot be simplified to a formula that only contains the baffle width. Also a flat baffle should give more support than a sphere of the same dimension. On the other hand, it appears as if the baffle step often is not fully compensated, since there is an effect of the room that also boosts the lower frequency. I sure welcome a discussion on this! __________________ Simulate loudspeakers: Basta! Simulate the baffle step: The Edge

7th May 2004, 06:32 PM	#4
Svante diyAudio Member Join Date: Feb 2004 Location: Stockholm	One reference that I see a lot on fora is this: http://www.trueaudio.com/st_diff1.htm It clearly uses the diffraction around a sphere to derive the function 380/W (W here in feet, 380*12=4560) and then magically the diameter of the sphere is assumed to have the same effect as the width of a rectangular baffle of any height. Obviously, I wrote the Edge program, so you should be sceptic towards me, since I probably am biased. But as far as I understand there must be room for an error in the above reasoning. __________________ Simulate loudspeakers: Basta! Simulate the baffle step: The Edge

7th May 2004, 06:43 PM	#5
Dave Jones diyAudio Member Join Date: Jul 2003 Location: Silicon Valley	I've fiddled around with the F1, F2 fields in edge.exe some. When I set them at 350 and 700, I get an L of 1.8, and the curve looks at least as good as the "suggested" 2.6. So I'm going with 1.8 mH. Thanks for your help, and thanks for the nice program. __________________ Davy Jones

7th May 2004, 07:08 PM	#8
planet10 frugal-phile(tm) diyAudio Moderator Join Date: Oct 2001 Location: Victoria, BC, NA, Sol III Blog Entries: 4	If i ever get time to write my OS X speaker simulator it will be in Python... dave __________________ community sites t-linespeakers.org, frugal-horn.com ........ commercial site planet10-HiFi p10-hifi forum here at diyA


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

7th May 2004, 08:04 PM	#9
SY diyAudio Moderator Join Date: Oct 2002 Location: Austin, TX Blog Entries: 1	(scratches head) Look, I love simulation programs (I use BDS, a wintel Excel-based program), but xover/eq parts are fast and easy to get. How about (radical idea here) putting the actual speaker in the actual baffle and actually measuring it before putting the xover/eq together? That way, you can order the exact inductor you need- or, more likely, find out that the eq circuit you need is something different than that predicted by any of the simulation software. I think that baffle simulation is most useful for designing the baffles in the first place. You can choose a size/shape/driver positioning that minimizes hard-to-correct anomalies. The results won't be rigorously quantitative, but qualitative is all you really need at that point in the design process. __________________ “Listening to records is like ****ing a picture of Brigitte Bardot.” - Sergiu Celibidache

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Does open baffle suffer from baffle step?	454Casull	Multi-Way	15	19th May 2012 03:12 PM
Baffle Step Calculations? Help!	Mutley666	Multi-Way	5	8th April 2005 08:49 AM
baffle Step	rgrayton	Full Range	33	4th December 2004 11:16 PM
Baffle Step + Sub	Aural Dan	Multi-Way	7	9th October 2004 08:23 PM
baffle step help.	AudioGeek	Multi-Way	3	10th February 2004 11:02 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read