Noob question: baffle step compensation

[HeatMiser]

Greetings.

I'm in the beginning planning stages of a relatively low-cost, simple, sealed 2-way project (supplemented by external subwoofer), tentatively using a Peerless 850439 midwoofer and either a Peerless 811815 or Vifa XT25T dual-concentric tweeter with a serial type crossover based on one of several circuits available on the web.

I'm wanting to incorporate a simple baffle step compensation filter based on John Murphy's paper here.



The calculations are straightforward enough, leading to 8 ohm values for both resistors and approximately 1 uF for the inductor, based on the baffle dimensions I have in mind. I'm left with a couple of questions however:

- Is it an accepted practice just to plunk a filter like this in line with an existing crossover? Or is the simple circuit shown above merely a theoretical representation, and the compensation should be "built in" by adjusting the values of the main crossover?

- If a filter like this can just be dropped in front of the rest of the crossover, what kind of power are those resistors going to see? It looks like it might be substantial, even though much of the lower bass is going to run through the inductor rather than being shunted through the resistors.

By the way, I'm aware that room loading and other factor will enter into the equation, but I'm trying to keep things simple for the purposes of this rather basic question. You may expect more complex questions later...

Thanks in advance for any advice; and of course other opinions on my proposed project are welcome.

[SY]

Unless you have drivers with a constant impedance vs. frequency, this circuit will not do what you want it to do. It's normally easier to do a passive (or even active) R-C circuit at line level (i.e., between pre and power amp), or to adjust topology and values in the speaker xover to get you the desired transfer function.

[HeatMiser]

SY, thanks for the reply; that makes sense. I guess I have more research ahead of me in the crossover department.

[SY]

"antisuck.com." What a domain name! Made my day, thanks.

[stokessd]

Actually the resistor in parallel with the inductor will form a baffle step compensation of sorts. IE short at low frequencies and "R" at high frequencies.

If you are going to build a set of speakers from scratch, you probably want to get together some sort of measurement software so you aren't blind. textbook crossovers rarely work properly for real world drivers

Also, the xt25 is a touchy driver, it doesn't like to play low (think Sd of a 3/4" tweet). It's a great tweet but harder than some (IE ss 9500) to get working well.

Sheldon

[HeatMiser]

SY, glad you found my domain name as amusing as I did when I registered it; I should probably do something with it some day.

stokessd: with regard to the compensation thingy, that's the theory. I think SY makes a good point that the response won't be as predictable as one might hope due to driver impedance irregularities. Measurement software is definitely in my future; I haven't ordered any parts yet, I'm just trying to get a feel for ways I might tweak things once I get started. Thanks for the advice about the xt25 - being that I am a beginner, I'll probably go with the Peerless tweeter since I've heard they are not too bad and they don't cost much.

[SY]

Measurement software is definitely in my future; I haven't ordered any parts yet, I'm just trying to get a feel for ways I might tweak things once I get started.

Well, as you probably already know, you can cob together some much-better-than-nothing measurement capability for the cost of a Panasonic mic capsule and maybe 5 bucks of electronics- assuming you've got a soundcard. And those are dirt cheap these days. That adds up to less than the cost of a crossover inductor.

[apassgear]

Serial type crossover are very impedance dependent. So you may run into trouble trying to add a pasive step compensation circuit.