Can you help me design a box?

[septimus]

I just need to know the volume needed to obtain an optimally flat response -in car- for a sealed enclosure. that means calculated with cabin gain. I have a protege wagon if cabin size can be taken into account. If i had to take a guess, I'd say its about 130ft^3.

Wow, I just googled average interior volume of a car and hit wikiepedia.

http://en.wikipedia.org/wiki/Vehicle_size_class

Scroll down to USA regulations, small station wagon says under 130, over 120.
I think I deserve a cookie or something.


Anyway. I have a mac and am too lazy to drag the pc laptop out of the closet and find some software for it.

Here are the specs on my sub. I will be feeding it 300 watts.

Here are the technical specs;
Dual 4 ohm
Revc = 7.8 Ohm
Fo = 22.6 Hz
Sd = 350.0 cm^2
Krm = 2.1538m Ohm
Erm = 1.085
Kxm = 193.204m H
Exm = 0.588
Vas = 71.3 Ltr
Cms = 409.87u M/N
Mmd = 129.248m Kg
Mms = 133.013 g
BL = 19.032 T.M
Qms = 3.290
Qes = 0.37
Qts = 0.333
No = 0.22%
SPLo = 85.36 dB

Thanks in advance!

[npdang]

From the linearteam forum:

"For everyone wanting to predict "Cabin Gain" (their vehicles transfer function)
by Josh Sanders

Assuming we have a perfect sound source.

(1) Treating the car as a small, empty, rigid, air tight, closed, box without standing wave resonances

This will give you a 12dB/oct raise in low frequency when the sound starts to act in pressure mode rather than wave mode. This is at frequencys below the lowest room resonance. Frequency of lowest room resonance = ( speed of sound / ( 2 x longest length in the enclosure))

(2) Treating the car as a small, empty, rigid, LEAKY, closed, box without standing wave resonances

The SPL gain in (1) will be attenuated as the pressure will not be directly proportional to the sound source.

(3) Treating the car as a small, IRREGULARLY STUFFED AND LINED , rigid, LEAKY, closed, box without standing wave resonances

Same as (2), but now the higher frequencys will be attenuated more than the lower frequencys.

(4) Treating the car as a small, IRREGULARLY STUFFED AND LINED, rigid, LEAKY, VENTED, box without standing wave resonances

Above the vent frequency the losses in (2) are reduced. At the vent frequency there is gain and below the tuned frequency there is severe attenuation. This effect is reduced because the vent has a large resistance, and the enclosure is leaky. With all of the windows up and doors shut, the tuned frequency will be infrasonic. (3) effect this the same as it effected (2) so The response is similar to (1) with infrasonic frequencys severely attenuated and the gain no longer approaches infinity at DC. With a window wound down or a door open ajar, the tuned frequency moves up into the audiable spectrum. The result is a noticeable gain at the tuned frequency and severe attenuation below the tuned frequency.

(5) Treating the car as a small, IRREGULARLY STUFFED AND LINED, rigid, LEAKY, VENTED, IRREGULARLY SHAPED ENCLOSURE

Similar to (4) but now at different locations inside the car there will be different gain/attenuation at ¼ wave frequencies. This effect is reduced by the irregular shaped enclosure, irregular stuffing inside the enclosure, and lining on the enclosure walls. If the frequency is such that the enclosure is acting in pure pressure mode as stated in (1) then there are no standing waves below the frequency of lowest room resonance.
The response of standing waves in a car is beyond the scope of this article and therefore will not be predicted.

(6) Treating the car as a small, IRREGULARLY STUFFED AND LINED, SEMI-RIGID, LEAKY, VENTED, IRREGULARLY SHAPED ENCLOSURE

Similar to (4) for small signal simulation. For large signal there is gain when the enclosure walls resonate out of phase with the sound source, and attenuation when the enclosure walls resonate in phase with the sound source. The amount of gain/attenuation is incresed as SPL inside the car increases.

Conclusion

Small signal prediction of car with windows up:

An inverted hipass with
f = Frequency of lowest room resonance = ( speed of sound / ( 2 x longest length in the enclosure))
Q = 0.707
Order = 2 = 12 [db/Oct]
infra sonic frequencys will not accurate

With WinISD 0.50a7
infra sonic frequencys will not accurate but better than above.

internal car length = l [m]
Linkwitz transform
f0 = c / 2l
fp =15
Q0 =0.707
Qp =0.707

Eg
internal car length = 1.9m
Linkwitz transform
f0 =90
fp =15
Q0 =0.707
Qp =0.707

internal car length = 2.2m
Linkwitz transform
f0 =80
fp =15
Q0 =0.707
Qp =0.707

internal car length = 2.5m
Linkwitz transform
f0 =70
fp =15
Q0 =0.707
Qp =0.707

internal car length = 2.9m
Linkwitz transform
f0 =60
fp =15
Q0 =0.707
Qp =0.707

internal car length = 3.5m
Linkwitz transform
f0 =50
fp =15
Q0 =0.707
Qp =0.707

Refrence: Seigfried Linkwitz
http://www.linkwitzlab.com/images/graphics/enclosure-spl.gif"

[chad]

Cool info!

I like to take a test enclosure, measure it in an anechoic area (can be done outside on a calm day on a nice stand) then toss it int he car and measure it. Wola, the response of your car from data interpolation!


-OR- the smartass Chad response... Fill your car with water, dump it out and see what it holds :)

Chad

[septimus]

lets say we just assume 130ft^3. Internal car length 2.5m.

Could someone give me best internal box dimension for a simple sealed box?

I have no testing equipment.

[solacedagony]

Cool info!

I like to take a test enclosure, measure it in an anechoic area (can be done outside on a calm day on a nice stand) then toss it int he car and measure it. Wola, the response of your car from data interpolation!


-OR- the smartass Chad response... Fill your car with water, dump it out and see what it holds :)

Chad

Not trying to steal your thread, but by data interpolation, do you mean comparing the change from the anechoic area into the car? So if you get 92db @ 60hz anechoic, and 95db @ 60hz in-car, you have +3db @ 60hz from cabin gain?
/offtopic

[septimus]

Yeah. You could actually compare the two response graphs and then interpolate a transfer function formula from that to calculate what a given speaker systems response would be for the entire measurement window.

[demon2091tb]

This is really intresting, anyone actually done this and gotten rather close to what were calling the transfer function and actually build the box based on the cars transfer function.......to get that flat response, and how close were they after?

Just intresting stuff.......if you don't have an EQ to make up for it.....even though an EQ is a patch, getting it pretty close on just by doing a few extra steps would be worth it i'd think. But no equipment here either.

[septimus]

When I built my first DIY project, a NHT 1259 sub and plate amp for my stereo from madisound, I had used some freeware box modeling software that had an input for cabin gain/room size. Now I have a mac, and don't feel like getting virtual PC or anything like that. I'm sure some of those box modeling programs have it. Does anyone know of one?

[septimus]

Anybody?

bump.