Baffle step compensation in an active X/O

[sunil]

Hi,

I hope this is the right forum to post this question.
I'm making an active X/O based on John Pomann's design.

http://www.snippets.org/filters/crossover.htm

I need to calculate the C's & R's for a baffle step
compensation circuit.

I'm using Alex Megann's writeup on:

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

Now from what I understand, the geometric mean for the midpoint
of the step is calculated by the square root of 2 which gives the 3db point, 2 being 6db. How do I calculate the rest ?
Eventually I want to be able to switch from 0 to 1, 1.5, 2, 2.5
& 3db points.

Also, I would like some help from people who are using Active X/O's to know whether it would be worth the effort & if, then should I build the X/O past the 3db point.

Any other ramblings would be welcome.

Cheers,
sunil

[ashok]

Hi Sunil,
I thought I should tell you that active crossovers are worth the time and money you spend on it. Done properly it will be far better than any passive crossover.
For the variable crossover frequency you can use multi position rotary switches. Pots will not be good enough IMO. Making it variable only helps while designing. Later it will not be of much use unless you plan to test other drivers. You could also use small plug in boards with the resistors and caps on it. That way you can get almost any frequency you want. Make the opamp circuit around it have gain. You can then control the Q of the filter from 0.50 upwards. Might be useful .
Cheers.

[sunil]

Ashok,

I've never done this before so I don't want to complicate it too
much. Right now, a simple X/O with baffle compensation & volume control (attenuation) should be good.

I can't figure out how to get the geometric mean for the points mentioned, ie......0.5, 1, 2 & 2.5 db. Since 3db is taken as sq rt of 2 & 1.5 db the fourth root of 2, how do I figure out the rest.

Cheers,
sunil

[ashok]

Hi Sunil ,
I am not sure what you mean by 3, 2.5, 2 ,1.5 etc db.
What one does adjust is the -3db frequency. The -3db point is only a reference point. You have -3db points for any order of crossover. The order indicates how steep the roll off is and the -3db point tells you what frequency it is designed for and is also 0.707 times less than the mid band response ( 1/2 power point).
Thats a crude quick explanation.

So you design for center frequency ( what you really want ) and for a few steps on either side. What you change is the
-3db "frequency". So for a 3Khz crossover frequency you can have steps at 1.5, 2 , 2.5 , 3 , 3.5 , 4 , 5 KHz. Very fine steps may not be essential in your case. Whatever frequency you set , it will be -3db down at that point.

Hope this explanation makes things clearer.
Cheers.

[stokessd]

Do you mean 1-3 dB of baffle step boost? I'm not sure what you are getting at, I know you want low frequency gain, and making it variable can be useful. In my experience the 6 dB loss from 2pi to 4pi isn't really realized because there is also room gain working in your favor, and many rooms and speaker placements never really can get to 4pi.

I use a shelving circuit for baffle step in my prototype active crossovers, and I've used a shelving EQ to great success for my 2pi to 4pi attenuation problem on my dipole ESL's. you can see my circuit here:

http://quadesl.com/speaker/diyesl/esl_xo_manual.pdf


Shelving EQ's are discussed in Jung's op-amp books and if you are really stumped, I can help you out.

Also you want to keep in mind that you may need an all-pass network in your active EQ. An all-pass network is a phase shifter (with no amplitude attenuation) of sorts. You may have problems with different acoustic centers between say a dome tweeter and a midbass cone driver. the acoustic center of the cone driver can be 2" behind the tweeter which is 130 degrees of phase shift between the two at 2.5 KHz. So if you crossover is there, you'll want to lag the tweeter or lead the woofer to make the phase replationships of the drivers match for a smooth transition region.

An all-pass network isn't nearly as useful for a sub crossover because you are dealing with such long wavelengths, that being off by say 2 feet between the sub and the main speakers is like 45 degrees at 80 Hz. That's not ideal but room mode based response anomalies will swamp the phase response based amplitude anomalies.

Sheldon

[sunil]

Sheldon,

I'm using the write-up by Alex Megann as
mentioned in an earlier post. Yes, I'm trying
to get a lift out from Baffle step.

Your piece of work is really neat.

I think there is some confusion here regarding
my question. I got it clarified by Alex Megann,
the author of the article.

I was trying to get the formula for calculating
a voltage ratio from a decibel value db, so I can
have 5-7 different settings, switchable, for
different lifts in the baffle step compensation
circuit.

Since it will be part of the correction being
incorporated in my active X/O. These would
start from 0, then 0.5 & increment in 0.5 db
steps till possibly 4 db.

A = alog10 (db/20)

BTW it is an awesomely simple write-up to
fully understand baffle-step & correct it.

Cheers,
sunil

[5th element]

Try looking at linkwitz site too, www.linkwitzlab.com
under the shelving lowpass, this to me explains how to caculate everything. Hope that helps