6th order bandpass design?

SubNut

Member
2001-08-20 10:32 am
For years, I have never figured out the dual-reflex box, where there are two chambers and both are ported. I just got a program call WinISD and am a little paranoid that it won't work right just like all the others. I practically spend endless hours perfecting every square inch of the box and take into account for all the displacements (spkr, port, etc). I guess my question is: "Has anyone used WinISD for building a 6th order box and been truly happy with the performance?"

Here's the driver specs:

Fs 34
Qts 0.262
Vas 5.20
Xmax 2.0 inches (peak-to-peak)
Re 3.4 ohms
Sd 169.65 cu in
Disp 0.43 cu ft
Sens (1w/1m) 96 db

It is an 18 inch Cerwin Vega Stroker sub and I know it's
not the very best but it fits the need. I believe in low frequencies, but if it's flat down to 30 hz - I am very pleased. Would like a little gain (3 db) at 35 hz.

Again, "Has anyone built a 6th order box and been happy with the sound? If so, is it WinISD or another program?"
 
hi.

you my see my comments on bandpass cabinets and win-isd in another thread.

about 6-th order bandpass cabinets we did make some tests with these (also...) some years ago.

we had a test cabinet with the driver mounted on a wall between 2 chambers and we had adjustable ports in each chamber.

measuring the output (with a pc measurement system) we couldnt find any port on the second chamber that didnt impair the output (let alone giving a speaker a ride for its money, mening its excursion was quickly reaching unsound levels with both chambers ported)

this may explain the ill faith of the jbl cabinets using just this topology and my general impression of win-isd (and other simulation progs) being useless in bandpass enclosure design.

in most (if not all) cases a simple bass reflex cabinet worked far better than a bandpass cabinet of same total volume.

bye k madsen - cadaudio.dk
 
I've used both WinISD and LspCAD Pro to simulate speaker enclosures. WinISD gives very little actual data, and does a poor job of simulating high volume levels. LspCAD Pro seems to give excellent results, but I've never built a box from it.

I've modeled some 4th and 6th order enclosures with LspCAD Pro. In every case, the impulse response was poor. The speaker seemed to lag, and resonate. Lower order enclosures seem to have much less of a problem with this sort of behavior. I'm not sure exactly what this would sound like, but I'm fairly sure that it won't sound as good as a sealed or PR enclosure.

I tried to use the specs you supplied, and found they weren't adequate. I ran a few sims on a Stroker 18S (WinISD had enough specs to work with, but they weren't compatable with yours), and it looks like your driver won't give deep bass in either a sealed or PR enclosure. It starts rolling off quite early. This is typical of a car audio driver, where the small cabin volume compensates for the early rolloff.

In a 6th order, your driver can get down to 30Hz, without problem, but you loose 8-10dB of efficiency. The impulse response for this driver isn't as bad as some I've seen, but it's not real good. By the time your port size is 8" (I'd go at least that big with your driver, maybe 12"), the port length is INSANE. It might be better to use PRs instead of ports.

If you're looking for good, deep bass in a 3-4 cu. ft. cabinet (that seems to be where you are), I'd try a single Shiva or NHT1259. The construction is simpler, the driver is probably much less expensive, and the efficiency ends up higher. Also, as it would be a sealed enclosure, the bass should be tighter.
 

kelticwizard

diyAudio Moderator Emeritus
2001-09-18 2:33 am
Connecticut, The Nutmeg State
SubNut:

Just a general guideline for ported boxes:
In the real world, a speaker with a Qts of .4 should go in a box equal to it's Vas, tuned to it's Fs, and be 3 db down at that point.

A speaker with a Qts of .3 should go in a box that is one half it's Vas, and will, if properly tuned, have an F3 half an octave above, (1.4) it's Fs.

A speaker with a Qts. of .5 should go into a box twice it's Vas, and with proper tuning, should be 3 db down one half octave below it's Fs. That is .7 times Fs, NOT .5 times Fs, which would be a full octave below.

This is for a ported box, not double chamber.

As soon as you saw the Fs for the Cerwin Vega was 34 Hz, and the Qts was .262, you were in trouble.

Speakers with Qts lower than .4 are meant to go into enclosures much smaller than the Vas, and have cutoffs much higher than Fs.

According to AVI design, this Cerwin Vega speaker will give very flat response in a box that is 1.25 cu. ft. internal volume, tuned to 52 Hz, and be 3 db down at 62 Hz. That is an octave above what you were hoping.

When shopping for speakers, keep these numbers in mind:
a speaker of .4 Q goes into a box that is Vas large and goes down to Fs. I find it to be a nice quick guideline.

These are approximations. You can stick a speaker with a Q of .35 in the same box as one with .4 and it will be pretty close.

[Edited by kelticwizard on 09-22-2001 at 07:16 PM]