Modeling cone excursion in an open baffle

[owdi]

I'm tossing around an idea for a speaker with an OB midrange, and I wanted to model the cone excursion, to see if I can get away with a 1st order crossover.

Any way to simulate this in WinISD? I tried using VERY LARGE boxes, but got strange results.

Dan

[BobEllis]

Linkwitz has an excel sheet (spl_max1) that will show you the required displacement to get a given SPL - sort of backs you into it.

As Linkwitz points out, diplacement increases at 12dB per octave. If you filter it electrically at 6 dB/octave you're going to have increasing displacement at lower frequencies.

[owdi]

That doesn't sound like what I need. While the midrange will be mounted in an OB, I will not be using it below the frequencies where output drops off. Here's the modeled OB response:



The midrange will cover 200-2000hz.

In an OB, the fs of the driver will be around 45hz. Since the impedance peak is two octaves below the crossover point, I was hoping a single capacitor would be sufficient to protect the driver from overexcursion.

Maybe a single 80uF cap for a 1st over highpass xover at 160hz would do the trick. That would give me about -6db at 200hz, and keep the phase shift below 45*. Driver is the Vifa XG18.

Dan

[Svante]

Originally posted by owdi
I'm tossing around an idea for a speaker with an OB midrange, and I wanted to model the cone excursion, to see if I can get away with a 1st order crossover.

Any way to simulate this in WinISD? I tried using VERY LARGE boxes, but got strange results.

Dan

Using a very large box should simulate the cone amplitude of the OB correctly, but not the response.

Edit:

Here is the response of a made-up driver in an OB and in an infinitely large closed box. The red curve is the cone excursion for both systems.

[BobEllis]

Notice that the excursion at constant voltage drive increases by what looks like a nice 12 dB/octave slope right up to resonance. So reduce the drive by 6 dB/octave you're still going to get excursion that increases down to drive resonance. With 2 octaves between crossover and resonance, it is likely to limit your output.

[owdi]

Aha.. now I understand. Thank you both.

Running a few more models in WinISD shows a 1st order filter would keep the driver within it's excursion limits, but max excursion would occur at fs, where the OB respons would be about -16db. That's a waste of energy, and it would cause additional distortion.

With a 2nd order highpass filter, max excursion is just below the xover frequency. Much better, but it will cost more to build.

Dan

[Svante]

I am not sure I fully understand what you do with those filters, but I think taht you should optimize for the response first. If the driver limits the output, then you should add more drivers instead.

A driver that is suited for OBs typically have a high Qts, which will resiult in that the bass response is extended, but also that there will be a higer cone amplitude around fs. Reducing that amplitude by means of a filter counteracts the purpose of having a high Qts in the first place.

...or maybe I did not understand what you are doing.

[owdi]

Let me clarify. I decided to try an midrange OB for aesthetic purposes, and to "get rid of the box." The baffle will be shaped to give me exactly the response I need without any filters (when combined with the response of the driver).

Since I already have the response I want, the purpose of the highpass filter is only to increase power handling. I want to do this with minimal phase distortion, and minimal cost. That's why, initially, I wanted to use only a single series cap.

I'm concerned about power handling because this speaker will be used for music and HT. HT is so important to me, I will not use a dipole for bass.

I got this idea after modeling OB response using TheEdge, which has been an incredibly useful program.

Dan