Well, I thought I would start a build thread about my next subwoofer project. This is sort of an evolution of my now defunct 2.12.18PR sub, which had pretty good extension to around 13-15Hz (see more details on the 2.12.18PR at the project page of my web site).

The goals of this subwoofer are:

Here are the results of some initial modeling. I used the program "Basta!" to make these plots, and have confirmed the models using Jeff Bagby's Woofer Box and Circuit Designer:

Red lines = excursion; black line = SPL

The above graph is the frequency response I can expect at "maximum output", which is around 107dB in the passband. The output at the low end is limited by the PR excursion first, the driver excursion limit around 8Hz second, and lastly by the available amplifier power (see next post for more details on this). I plan to use the EP2500 bridged, and use the drivers in series. I won't need the full power that the EP2500 can deliver, and this should keep the current draw low enough that I can get away with the "quiet fan mod". I plan to install the amplifier inside the sub cabinet and use acoustically treated tunnels to supply air to (and exhaust air from) the amp.

In order to get the above response, I use the following:

The last filter, the 14Hz LP filter, shapes and flattens the response. When using these drivers in a 240L box, the models predict that the response will fall off at about 6dB/octave below around 40-50 Hz until the box tuning, and then at 24dB/oct below that. I found that, by using the 14Hz first order LP stage, I could flatten out the response so that F3 is between 10Hz and 11Hz. I increase the corner frequency of this stage to 20Hz if I want a more gradual roll off starting at 30 Hz. I hope that this will allow for some room to tune the response to the listening space. In reality, I should use a shelving LP filter but I could not model this accurately. This will only affect the response above 50Hz and will change the Q and type of LP filter I use in the crossover to the main speakers.

Next post: how I arrived at this design.

-Charlie

The goals of this subwoofer are:

frequency extension to around 10Hz

smooth low end roll off

minimum passband response: 105dB

low distortion

I will be using the following:smooth low end roll off

minimum passband response: 105dB

low distortion

TWO Dayton Audio RSS390HO-4 drivers

ONE Acoustic Elegance (AE Speakers) 18" PR with 2100g mass

Behringer EP2500 power amp

ONE Acoustic Elegance (AE Speakers) 18" PR with 2100g mass

Behringer EP2500 power amp

Here are the results of some initial modeling. I used the program "Basta!" to make these plots, and have confirmed the models using Jeff Bagby's Woofer Box and Circuit Designer:

**FULL POWER FREQUENCY RESPONSE AND CONE EXCURSION:**

Red lines = excursion; black line = SPL

The above graph is the frequency response I can expect at "maximum output", which is around 107dB in the passband. The output at the low end is limited by the PR excursion first, the driver excursion limit around 8Hz second, and lastly by the available amplifier power (see next post for more details on this). I plan to use the EP2500 bridged, and use the drivers in series. I won't need the full power that the EP2500 can deliver, and this should keep the current draw low enough that I can get away with the "quiet fan mod". I plan to install the amplifier inside the sub cabinet and use acoustically treated tunnels to supply air to (and exhaust air from) the amp.

**ACTIVE CROSSOVER NETWORK:**In order to get the above response, I use the following:

12Hz Q=1.4 second order HP filter

5Hz first order HP filter

85Hz Q=1.4 second order LP filter

14Hz first order LP filter

The 12Hz filter provides a boost around the box tuning and a cut at lower frequencies. It's basically creating a 6th order PR system, which limits the cone excursion below tuning (see the nice symmetric cone excursion in the plot above) and flattens the response. The EP2500 has a built in 5Hz first order filter so I included this in my model. The 85 Hz LP filter will be part of the crossover network leading to the main speakers and is adjustable.5Hz first order HP filter

85Hz Q=1.4 second order LP filter

14Hz first order LP filter

The last filter, the 14Hz LP filter, shapes and flattens the response. When using these drivers in a 240L box, the models predict that the response will fall off at about 6dB/octave below around 40-50 Hz until the box tuning, and then at 24dB/oct below that. I found that, by using the 14Hz first order LP stage, I could flatten out the response so that F3 is between 10Hz and 11Hz. I increase the corner frequency of this stage to 20Hz if I want a more gradual roll off starting at 30 Hz. I hope that this will allow for some room to tune the response to the listening space. In reality, I should use a shelving LP filter but I could not model this accurately. This will only affect the response above 50Hz and will change the Q and type of LP filter I use in the crossover to the main speakers.

Next post: how I arrived at this design.

-Charlie

## Comment