Thoughts on peaky response of this setup?

[n8skow]

Pair of SA-12s, common chamber enclosure.
3.5 cubes after all displacements with a single 6" aeroport, ~10.5" long

Getting a rather peaky response out of this enclosure in the back of a 2000 Celica hatchback - I'm looking mostly at the below tuning response which is dropping off fairly quickly and was curious if anyone had any thoughts to share on this...

At first I assumed cancellation issues - and have plotted 2 different response graphs with the enclosure firing forward - and backward to compare the two. There was a noticeable ~0.5 db gain across most of the spectrum - but still the very steep drop off below tuning. If I recall, 12db per octave is the typical roll-off? I didn't test low enough frequencies to actually confirm how much roll off I'm getting - but in a visual graph - it looks crazy steep.



And here's the power plot

[SPLEclipse]

Whoa...need to open that up a little more. There's only a 3db range between the highest and lowest points on your response, which is extremely small. Looks to me like you're response falloff is about right, but that actually surprises me. Normally even with a high tuning you get a flat or gradual roll-off because of stored energy and transfer function. Which brings me to: how did you measure this? Looks like you might have done peak-hold individual tone tests, which doesn't translate well to FR graphs because it ignores the time domain and is subject to more user error.

[profundus-sanus]

you have a 40hz tune. 12db per octave would mean you'd be down around 6db at 30hz and 12db at 20hz, roughly.

also if you're doing peak spl graphs at full tilt with any amount of power on the subs your results will be exaggerated by increased cone control at port tuning keeping the sub more linear.

also it could have something to do with you port to woofer orientation. maybe the port needs to be off axis of the woofer.

[n8skow]

Yeah - checked each frequency manually with a TermLab, DMM and clamp.
I have a DATS that I could plot the impedance response with - but at only 1-watt of power input - I won't really have a way to plot SPL at the same time.

The 3db high and low points doesn't bother me near as much as the 3db drop from 38hz-32hz - which seems extreme to me - but I am only a noob in this low frequency 'musical' stuff and have much to learn...
=)

Originally Posted by SPLEclipse

Whoa...need to open that up a little more. There's only a 3db range between the highest and lowest points on your response, which is extremely small. Looks to me like you're response falloff is about right, but that actually surprises me. Normally even with a high tuning you get a flat or gradual roll-off because of stored energy and transfer function. Which brings me to: how did you measure this? Looks like you might have done peak-hold individual tone tests, which doesn't translate well to FR graphs because it ignores the time domain and is subject to more user error.

[HACKMUNCH]

Last time I checked ported enclosures roll of at aprox 24db per octave. The reason why its peaky is you tuning is very close to your sealed resonance. Yes is more efficient but bandwidth suffers.

[n8skow]

Testing at roughly 60-75% of available power here.

The port to woofer orientation is a bit problematic in this vehicle - (90 degrees off axis from woofers) and extends behind one woofer - with the inside flare directly below the 2nd woofer. I previously had a slot port (49 sq. in.) - that was long enough to extend behind both woofers. I gained about a solid .5 db dropping port area to the single 6" aero though.

Originally Posted by profundus-sanus

you have a 40hz tune. 12db per octave would mean you'd be down around 6db at 30hz and 12db at 20hz, roughly.

also if you're doing peak spl graphs at full tilt with any amount of power on the subs your results will be exaggerated by increased cone control at port tuning keeping the sub more linear.

also it could have something to do with you port to woofer orientation. maybe the port needs to be off axis of the woofer.

[n8skow]

I hadn't considered that yet - glad you mentioned it Hackmunch.

I can go a little bit bigger on the enclosure volume (lowering the volume fs) and keep the tuning roughly the same to widen the bandwidth? Guess I need to fire up WinISD and find a happy medium on the efficiency/bandwidth see-saw.

Originally Posted by HACKMUNCH

Last time I checked ported enclosures roll of at aprox 24db per octave. The reason why its peaky is you tuning is very close to your sealed resonance. Yes is more efficient but bandwidth suffers.

[n8skow]

Another note - I had done the port to the driver side for SPL purposes (on high notes, my car likes woofers forward and port to driver side)... I'm curious how much things would change with porting on the same surface as the woofers (all facing back). Lower tuning seems to be a whole different beast here - having to throw a lot of my previous experience with this car out the window... =/

[HACKMUNCH]

No if you make enclosure bigger its gonna make things worse. You either A tune enclosure lower to widen bandwith and loose a little spl or B shink enclosure to raise sealed resonance and leave tuning alone.

[n8skow]

Thank you for the input fellas...
Definitely some more experimentation to do here...
=)