Attached are pics of the construction process for my two 20" diameter by 4' high sealed sonotube subwoofers. Each sonotube is capped with a 15" TC-2000 driver and will be powered by a Behringer EP2500 and DCX2496. More photos and some details are available in my photo gallery: Sealed Sonotube Subwoofer.
Sealed Sonotube Subwoofer
1 - 17 of 17 Posts
I don't have the amp yet. So calibration and everything else comes later. The rough design I put together in WinISD doesn't really show what the end result is going to be.
Here's set of graphs:
The blue line indicates how each of the subs should perform, before EQ boost, with 600W to each driver. All based on very loose estimates of 5ft^3 internal volume, which results in a Fb of 41.42Hz and Q of 0.713. A 3rd-order Butterworth highpass filter is included at 5Hz to account for electronics drop off.
The yellow line is the same enclosure, but with a Linkwitz Transform (approximated by the DCX2496, hopefully) applied. This isn't necessarily the anechoic response I am going to end up with, since I'm going to apply whatever filters are necessary to get flat in-room response to as close to 5Hz as possible.
The drivers can each take 1000W peak, and in fact they aren't being driven to their xmax in the yellow line graph. I don't know what the EP2500 peak output is. Couldn't find it anywhere. But you can add ~2dB to the SPL graph if each driver gets 1000W instead of 600W. And these graphs are for one sub, not two.
If the Linkwitz Transform doesn't need to have an F3 as low, with room gain, then the SPL also goes higher. For example, switching to an F3 of 18Hz moves the 600W SPL from 102dB to around 108dB.
Basically, that's why I'm saying the models aren't worth a whole lot right now other than to serve as basic validation for the design. Because the actual result is going to depend entirely on what filters I apply to get flat in-room response.
Here's set of graphs:
The blue line indicates how each of the subs should perform, before EQ boost, with 600W to each driver. All based on very loose estimates of 5ft^3 internal volume, which results in a Fb of 41.42Hz and Q of 0.713. A 3rd-order Butterworth highpass filter is included at 5Hz to account for electronics drop off.
The yellow line is the same enclosure, but with a Linkwitz Transform (approximated by the DCX2496, hopefully) applied. This isn't necessarily the anechoic response I am going to end up with, since I'm going to apply whatever filters are necessary to get flat in-room response to as close to 5Hz as possible.
The drivers can each take 1000W peak, and in fact they aren't being driven to their xmax in the yellow line graph. I don't know what the EP2500 peak output is. Couldn't find it anywhere. But you can add ~2dB to the SPL graph if each driver gets 1000W instead of 600W. And these graphs are for one sub, not two.
If the Linkwitz Transform doesn't need to have an F3 as low, with room gain, then the SPL also goes higher. For example, switching to an F3 of 18Hz moves the 600W SPL from 102dB to around 108dB.
Basically, that's why I'm saying the models aren't worth a whole lot right now other than to serve as basic validation for the design. Because the actual result is going to depend entirely on what filters I apply to get flat in-room response.
Attachments
Josuah, cool project. Yes you're right about taking the simulations with a grain of salt as far as maximum output and FR is concerned. You never know what effect room gain has until you measure it properly.
Depending on the size of your room you should be pleasently suprised by room gain. My sealed Tumult without EQ had a much lower in-room f3 than the anechoic simulations suggested. So, I hope room gain is as much on your side as it is on mine. Without it I wouldn't be satisfied with the FR...Simulations suggested an f3 around 33hz, my real-world F3 is around 18hz without eq.
Depending on the size of your room you should be pleasently suprised by room gain. My sealed Tumult without EQ had a much lower in-room f3 than the anechoic simulations suggested. So, I hope room gain is as much on your side as it is on mine. Without it I wouldn't be satisfied with the FR...Simulations suggested an f3 around 33hz, my real-world F3 is around 18hz without eq.
Joined
·
37 Posts
Wow,
I didn't know that a TC-2000 modeled so differntly than an RL-p15 d2 in a sealed box. I thought they were practically the same driver. Here is the model for my old sealed RL-p15.
As you can see, 100L and heavy fill results in a Qtc=0.493. Taking away the fill it only goes up to ~0.55. 100L = 3.5cuft, which is a long long ways away from your 5cuft.
I also was using a Behringer DEQ2496 to boost the low end. a +6db shelf filter at 39hz worked well.
I didn't know that a TC-2000 modeled so differntly than an RL-p15 d2 in a sealed box. I thought they were practically the same driver. Here is the model for my old sealed RL-p15.
As you can see, 100L and heavy fill results in a Qtc=0.493. Taking away the fill it only goes up to ~0.55. 100L = 3.5cuft, which is a long long ways away from your 5cuft.
I also was using a Behringer DEQ2496 to boost the low end. a +6db shelf filter at 39hz worked well.
Attachments
EP 2500 output power (1 kHz, 0.1% THD) per channel stereo: 500 watts RMS @ 8 ohms, 750 watts RMS @ 4 ohms, 1,200 watts @ 2 ohms. Output power (1 kHz, 0.1% THD) bridged mono: 1,500 watts RMS @ 8 ohms, 2,400 watts RMS @ 4 ohms.
You are aware of the incredible amount of power required to boost 10hz 18db, right? Be careful or you will clip your way into new drivers, fast.
Lets say you are cruising around with explosions going off around 60hz loud enough to eat up 25w. To play any 10hz content at the same level will require 1800w. At 50w nominal we are up to 3500w for 10hz. Umm... 100w at 60hz is 7000w at 10hz.
Just a reminder not to go boost crazy. Keep the boost level in mind and you will be fine. Have you checked out the FRDs LT spreadsheet?
You are aware of the incredible amount of power required to boost 10hz 18db, right? Be careful or you will clip your way into new drivers, fast.
Lets say you are cruising around with explosions going off around 60hz loud enough to eat up 25w. To play any 10hz content at the same level will require 1800w. At 50w nominal we are up to 3500w for 10hz. Umm... 100w at 60hz is 7000w at 10hz.
Just a reminder not to go boost crazy. Keep the boost level in mind and you will be fine. Have you checked out the FRDs LT spreadsheet?
Joined
·
37 Posts
Ah, you're putting two drivers in each tube? That makes more sense, and sounds about right for a Qtc=0.7 (Though I probably would have shot for an even lower Qtc). I saw two tubes, so I thought you were giving each driver your 5cuft.
Additionally, my Behringer DEQ2496 only allows a max boost of +12. And yes, lots of power required.
Additionally, my Behringer DEQ2496 only allows a max boost of +12. And yes, lots of power required.
I've got the clip limiters turned on on the EP2500. And I've been watching the lights on the DCX2496. So far, they've never gone above the second bar. Right now I've got a 28dB boost going using a pair of LP shelf filters each set at +14dB. It's chained as follows:
pre -> input 1 -> output 1 -> input 2 -> output 2 -> EP2500
Inputs 1 and 2 are linked with regards to their DSP settings.
Here is the response graph I've got so far, after applying about 11 EQ filters somewhat manually and somewhat suggested by Room EQ Wizard. But I'm not sure what's right yet because the 10Hz drop could be either a null or just represent issues at 11Hz that needs to be cut (unfortunately no outboard hardware EQ has 11Hz filters that I'm aware of). I'm going to need the newer Room EQ Wizard to know.
pre -> input 1 -> output 1 -> input 2 -> output 2 -> EP2500
Inputs 1 and 2 are linked with regards to their DSP settings.
Here is the response graph I've got so far, after applying about 11 EQ filters somewhat manually and somewhat suggested by Room EQ Wizard. But I'm not sure what's right yet because the 10Hz drop could be either a null or just represent issues at 11Hz that needs to be cut (unfortunately no outboard hardware EQ has 11Hz filters that I'm aware of). I'm going to need the newer Room EQ Wizard to know.
Attachments
I'm using the new digital Radio Shack SPL meter with the correct calibration file loaded. Also my soundcard response loaded. You can see the two lines on the graph. I'm leaning towards the 11.5Hz being a peak, rather than the 10Hz being a null.
Now that REW 4.0 has been released, I can post the <10Hz measurements I took of these subwoofers. I'm running two LP shelf filters to boost low, and about a dozen PEQs IIRC. I used Sonny's ECM8000 calibration mic and extrapolated to -12dB at 5Hz. Here's the graph.
Attachments
1 - 17 of 17 Posts
Join the discussion
