High Frequency Rolloff in software packages

can anybody recommend a loudspeaker design program that correctly accounts for the high frequency rolloff of drivers and their respective enclosures?

my textbook lists this frequency being equal to RE*MMS/ (2*pi*LE*MMD) but i haven't seen it being implemented in any of the programs i've used. mostly interested in the phase implications rather than magnitude.

thanks

jt
 

CHRIS8

Disabled Account
2001-12-12 8:47 am
VA, USA
"can anybody recommend a loudspeaker design program that correctly accounts for the high frequency rolloff of drivers and their respective enclosures"

??? Please clarify.

Nor amplitude or phase response anamolies are significantly inaccurate when using just about any of the popular softwares, along with a mic with a proper calibration file. You can aquire such a mic with cal file for amplitude and phase correction for only $28 from Kim G. off of madisound board.

However, I have a feeling you are NOT referring to this. Are you asking about LF driver rolloff predictions? If so, a simple equation using published inductance and mass specs will not be able to theorize an actual usable HF output. In addaition to inductance, the actual limiations will also depend on cone stiffness relative to mass, cone curvature and internal damping properties of the diaphragm. This will define HF behaviour, as far as extension and breakup modes are concerened. It is much easier to just purchase the mic capsule i refernece above and use some freewaer such as speaker workshop to make 'real' measurements instead of guessing. Or, you can even use a radio shack SPL meter, making sweeps at 0, 15 and 30 degrees off axis outside. If you refuse to do either of these things, a much better alternative still(as compared to the sole use of the formula you cite)is to just use the published amplitude/impedance plots.

The formulas you are referencing---by themselves, are about as practical as raw textbook crossover formulas without the nesecarry amplitude/impedance/phase data of the drivers you plan to cross. LOL.

-Chris
 
The formulas you are referencing---by themselves, are about as practical as raw textbook crossover formulas without the nesecarry amplitude/impedance/phase data of the drivers you plan to cross. LOL.

yes, that would be ideal. but I also don't have 10 grand to spend buying drivers and test equipment. I am not looking for a realistic high frequency magnitude response. but the theoretical phase response can help me narrow down which slopes and frequencies to consider for use in the crossover. so I only have to buy a few capacitors.

I decided to just write a VB program to work through all the calculations for me.

jt
 

CHRIS8

Disabled Account
2001-12-12 8:47 am
VA, USA
You can get ok results by tuning by ear, if need be. However, you are the one insisting that YOU need phase response data. It's not expensive to do this, btw.

"yes, that would be ideal. but I also don't have 10 grand to spend buying drivers and test equipment"

$28 - mic capsule
$5-coaxial cable for mic
$3-1/8" mini plug
$1-plastic pen(body)
$2-Super Glue(glue capsule to pen body)
$4-Shrink Tubing
$25-mic preamp parts
$0-Freeware such as Speaker Workshop, or your choice of shareware such as Loudspeaker Lab, etc.
Total=$68 or so

Actually, if you have a somewhat linear(mic gain stage) soundcard and amplifier, you can even get away with not using the mic preamp and get 'OK' phase measurements as-is. Now your down to only $43. :)

$10,000 may be a bit of an exagerration. LOL. Of course, I am presuming you have a Windows based PC with a full duplex soundcard.

"but the theoretical phase response can help me narrow down which slopes and frequencies to consider for use in the crossover"

Unfortunately, your theoretical HF rolloff amplitude/phase responses based on textbook theory alone will not be useful in reality. The only phase information that will be able to be predicted accurately is LF rolloff amplitude and phase information, in a given enclosure alignment...if T/S paramters are accurate. Even that iis a big IF. LOL.

-Chris
 
jteef, chris8 has a point. You may assume ideal filter components in your calculations. However, everything else such as the minimum phase response, calculated transfer functions etc. can only be considered informative if they were calculated from measured data. Otherwise, your VB code is just another XoverPro clone :) But yeah the programming practice is well worth it :)

There's no way to accurately calculate the behaviour of drivers above 200Hz-300Hz without using Finite Element Analysis or at least derivatives of FEM.