Finding the best driver for a box that I have

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have an existing cabinet that I got from someone for FREE who was downsizing to an apartment. He took out the drivers to repurpose them. Each cabinet has 2 separate air spaces with speakers connections at the back to connect each 15" driver separately allowing either a series or parallel wiring scheme based on what drivers are selected. I'm driving each cabinet with a QSC CX302 bridged to mono (1200w @4ohm, 700w @ 8ohm)

I took measurements and to the best of my ability to measure the internal volume of each part of the cabinet is 6.4 cu ft. There are 2 triangular shaped ports that tune the cabinet to around 32Hz with both ports open, but if you block one of the ports effectively cutting the port size in half it drops the tuning to 24Hz. Even with one of the triangular ports blocked the opening is still large enough to avoid port noise.

I've done some research and found a few drivers that might work. Since I have to buy FOUR drivers, cost is a concern. I want to get the best performance at minimal cost. I'll list the ones I've run simulations for and what the results were. I certainly don't know all of the drivers in existence, so I hope that someone in this group might know about some that I haven't considered. This is what I'm evaluating so far along with the pertinent parameters. Of course the end result with take into account the increase in overall db from using dual drivers. I list the impedance to consider what the amp will drive. If I use an 8 ohm driver I'll wire in parallel for an overall 4 ohm load. But if I use a 4 ohm driver I'll wire in series for an overall 8 ohm load.

Dayton DCS385-4 $94.30 4 ohm 300wRMS F3=25Hz 88db efficiency

Dayton UM15-22 $189.81 4 ohm 800wRMS F3=24Hz 86.4db eff

American Bass DX-15 $90.48 4 ohm 400wRMS F3=25Hz 88db eff

American Bass XO-1544 $90.43 2 ohm 400wRMS F3=26Hz 88.5db eff

American Bass XD-15 $121.57 4 ohm DVC (I'd wire it to get 2 ohm per driver) 700wRMS F3=24.6Hz 87.5db eff

Based on what I've found so far I'm leaning toward the AB XO-1544 because I can wire it to get a 4 ohm overall load and get more power from the amp. There is negligible difference in an F3 of 24,25 or 26Hz so the 88.5db eff combined with cost is driving my decision.

Is this the best driver or does anyone else have other ideas?
 

Attachments

  • 20170410_191825.jpg
    20170410_191825.jpg
    703.3 KB · Views: 154
  • 20170410_191818.jpg
    20170410_191818.jpg
    964.6 KB · Views: 147
  • 20170410_192028.jpg
    20170410_192028.jpg
    939.1 KB · Views: 147
  • 20170410_191810.jpg
    20170410_191810.jpg
    538.7 KB · Views: 140
I was told that it was the Klipsch Cornwall woofer (K-33-E). I ran sims on that one and it did exactly what Klipsch had in mind for the driver. The low end rolls off around 40hz but the efficiency is in the high 90s. Since I'm powering these separately with their own amp I'm not as concerned about efficiency and I'm looking for more low end extension.
 
Some general observations:
1) Way back when, Richard Greiner wrote an Audio Engineering Society (AES) paper investigating series and parallel wiring schemes, one of the conclusions of which was to avoid series wiring. Imagine an exaggerated oversimplified example of one woofer resonating at 40 Hz with a big impedance peak, it will "take all the power" compared to the other which lets pretend has its resonance frequency at 50 Hz. Hence I'd favor 8 ohm per driver solutions.
2) Greiner's paper also found the best was all paralleled in common enclosures. BUT some prefer separate cavities, so if one driver blows it does not take the other with it. On the other hand, if I was in your situation, I wouldn't be trying to subdivide the cabinet ha ha.
2b) On another note, from some experience with a multichannel amp driving several woofers in an automotive application, I do NOT like different amp channels driving different woofers in a common cavity, since I had a situation where one woofer moved a lot more.
3) Most amps I really doubt deliver much more actual (not rated) peak clipping power into lower impedances. Something like the QSC might be an exception. I guess there is an AES paper waiting to be written...
3b) You have TWO of the QSCs? Just checking.
4) Efficiency numbers don't apply well to subwoofers. Dick Small created that as a midband spec, and it only vaguely relates to the actual sensitivity at low frequencies. You really have to model the sub, and NOT in some program that simulates the box as an electrical filter. You need something like LEAP that actually takes into account the moving mass, the coil's semi-inductance, and spits out a much truer response, and use that to look at the whole sensitivity curve at low frequencies. If that's not really possible, then looking at your numbers I'd say they are all pretty close so I wouldn't worry too much. I would block one port and get an 8 ohm (or 4x2 DVC) woofer with the most excursion. Frankly, I'm feeling at 6.4 cubic feet I wonder if some Eminence or more PA-style woofer would work better, but I'm not up enough on current drivers to recommend something.
 
head_unit - in response to each comment....

1) very interesting point. I didn't realize that about the series wiring. Does this still apply if both drivers in series are identical? Specifically with the American Bass XO-1544 which is 2 ohm per driver, what would the drawbacks be of wiring that in series? That driver is actually a dual voice coil with 4 ohm per VC, so I could wire each speaker in parallel to get a 2ohm load per driver and then wire that in series for 4 ohm overall. OR I could wire the individual driver in series to get an 8 ohm load, then wire the drivers in parallel for a 4 ohm load. Would the different wiring schemes make a difference?
2) the cabinet is already built and there are two 6.4 cu ft chambers per cabinet. There isn't anything I can do about that. I have two ports per cavity in the cabinet and by blocking one of the ports I can change the tuning frequency. I found that blocking one port completely gives an Fs of around 24Hz which seems to work with the drivers I've found. There will be one woofer per cavity with a single QSC amp driving each cabinet (2 woofers per cabinet)
3) The QSC amp is a solid professional grade amp. I wouldn't drive it into a 2 ohm load but it's rated for 1200w into a 4 ohm load. Yes, I have two QSC amps, one for each cabinet.
4) I've looked at countless (40+) "PA style" woofers and they all have a common trend - high sensitivity and the bass rolls off usually between 50hz and 60hz, even with a big 15". A few will extend into the mid 40s, but I haven't found a PA style woofer that gives significant response under 40hz, hence why I'm looking into more traditional home theater and car woofers.

While this is a general statement and exceptions might apply, I've found that the PA style woofers are the ones that have the accordion style paper surround which might factor into their higher sensitivity. The woofers with longer excursion have foam or rubber surrounds that are rounded.

Based heavily on comment #1, should I keep searching to find an 8 ohm driver?

The modeling that I've used so far is the website Lautsprecher Gehauseberechnung mit Thiele Small Parameter. I know there are software programs that are much more detailed, but this seems to give pretty good results for a rough simulation.
 
I should also mention that my main speakers (left, center and right) are LaScalas, so I have pretty solid bass down to 50hz. I'll probably set the crossover in my receiver around 60hz to give more of the bass in the 60hz - 150hz range to the LaScalas and let the subs take over the last couple of octaves.

One of the few "Pro" speakers I found that gave any kind of low extension was the B&C 15BG100. With both ports open and a Fs of 32hz it has an F3 of 32Hz. It has 2000w RMS power handling, but at a cost of $315 per driver.

The Eminence Lab15 has a response down to 30hz with a Fs of 32Hz. The power handling is 600w RMS, but this comes at a cost of $285 per driver.
 

Attachments

  • 20170830_085342.jpg
    20170830_085342.jpg
    590.7 KB · Views: 97
Last edited:

GM

Member
Joined 2003
The guy told me that it was designed for the same drivers the Klipsch Cornwell uses (K33-E).

OK, found very different K33 specs, so using the ones that made the most SPL, I plugged in a few well proven sub drivers suitable for, but not designed specifically for, prosound vented, bass horn apps, i.e. mobile audio woofers and not surprisingly [to me anyway], Rockford Fosgate has the best bang/buck in a 15" of the one's I know are still available, but even with Amazon's heavy price discount it's still $192.22 delivered and your cab's vent system is way too small, with a 60+ m/s vent mach/1200 W according to Hornresp :(, so can't use its potential, though at 750 W its good to 115 dB/~22 Hz and worst case may have to critically damp the vent if there's any noticeable port chuffing [sims a narrow spike at ~22 m/s]: Amazon.com: Rockford Fosgate P3D2-15 Punch P3 DVC 2 Ohm 15-Inch 600 Watts RMS 1200 Watts Peak Subwoofer: Car Electronics

Otherwise, the above performance is about the best the cabs can do power ratings wise with the lower tuning, so choose a lower 20-25 Hz Fs, >8-10 mm Xmax depending on efficiency.

GM
 

Attachments

  • Rockford Fosgate P3-D2 Vs Klipsch K-33E 6.4 ft^3_24 Hz.PNG
    Rockford Fosgate P3-D2 Vs Klipsch K-33E 6.4 ft^3_24 Hz.PNG
    43.1 KB · Views: 35
  • Rockford Fosgate P3-D2 Vs Klipsch K-33E 6.4 ft^3_24 Hz [vent mach limited].PNG
    Rockford Fosgate P3-D2 Vs Klipsch K-33E 6.4 ft^3_24 Hz [vent mach limited].PNG
    70.3 KB · Views: 23
GM - you're right. This driver has some crazy good response. Just to compare apples to apples with other drivers I've been comparing I ran it through the same sim I've used on the other ones. I've attached that file. What I found basically matches what Hornresp told you:

Box volume of 185 liters (roughly 6.4 cu ft)
Rockford Fosgate P3D2-15 Punch P3 ($194.22 free shipping on Amazon)
Fs = 23.42
F3 = 22.53
Dual voice coil 2 ohm per
600w RMS

The best performance from what I had evaluated previously was the Dayton Audio UM15-22 15" Ultimax. I've attached the graph for it also. It gave the following performance:

Dayton Audio UM15-22 15" Ultimax ($180 free shipping on eBay sold by Parts Express)
Fs = 23.42
F3 = 23.6
Dual voice coil 2 ohm per
800w RMS


Given that these are more expensive than the American Bass I had looked at earlier, I'm thinking that I could put ONE driver in each cabinet and leave the other one empty (for now) and have a damn good sounding sub. If the day comes where I get some extra money I could always buy the 2nd sub for each cabinet.


GM - How did you calculate the airspeed to know if port noise would be a problem? The equivalent diameter of the port is 7.33" with a length of 12". I wouldn't think that a port that large would be susceptible to port noise. I found an online calculator at mh-audio.nl - Home which gave an airspeed of 30.34 meters per sec or mach 8.87%. The site said under Mach 10% would not be audible. So do I have a problem with the port size if I block one of the 2 ports to drop the tuning to 23Hz?
 

Attachments

  • Rockford Fosgate P3D2-15 Punch P3.png
    Rockford Fosgate P3D2-15 Punch P3.png
    42.7 KB · Views: 43
  • Dayton Audio UM15-22 15in Ultimax.png
    Dayton Audio UM15-22 15in Ultimax.png
    47.6 KB · Views: 39
Yeah, just block them off, stuffing the cavities with a sack of kitty litter or similar.

In HR, the only way I know is to adjust the power, Xmax to 115 dB [in this case] for it to calculate the correct vent mach plot.

Since you didn't publish the port dims, I used a single 6" dia. pipe, since it was more than adequate for the higher power listed K-33E.

Hmm, my vent area [Av] is considerably smaller, yet quite a bit longer [15.37"] to get a ~24 Hz tuning, so something's not right; at 7.33" dia. = 42.2"^2 requires a 24" long vent, not 12".

5 % mach is commonly used, though Richard Small [the S in T/S] noted that he couldn't hear any at ~4 %, so with a common 344 m/s SoS, 5 % = < 17 m/s is the recommendation for all but the most extreme apps.

GM
 
The inside dimension of the 2 equal sides of the triangular port is 7.5" (hard to tell size from the photos). The area would be 7.5 * 7.5 / 2 = 28.125 sq in. To get the radius we use: sq rt(28.125 / pi) = 2.99", so the diameter would be 6". I don't know where I made the math mistake to get 7.33" equivalent diameter but thanks for pointing it out.

With a 6" diameter port that is 12" long and a cabinet size of 172 cu liters (estimating space lost to driver) it gives an Fb of around 27Hz. To get the 24Hz tuning I'm going to have to cut a 12" long 4x4 in half lengthwise and put them in the corners of the port to reduce the overall size of the opening. That would reduce the opening area by 16 sq in, so the equivalent port size is 5" diameter. That gets the tuning back to 24Hz where it should be.

With the smaller port size the airspeed is up to 16% mach (57m/s) which is not good. I need to find another way to drop the box tuning down to 24Hz. Lengthening the port isn't feasible because the box inside wall makes a 45deg angle where the port ends. Since the port isn't a tube, adding another piece of PVC pipe to lengthen it can't be done. That's a task I'll look at another day. My brain is too tired to think any more tonight.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.