Keystone Sub Using 18, 15, & 12 Inch Speakers

Hi Djim,

Sorry for being unclear about what I ment, maybe this will do it:

Regards,
 

Attachments

  • Keystone Fold_bottom mouth.jpg
    Keystone Fold_bottom mouth.jpg
    11.8 KB · Views: 2,564
Hi weltersys,

This is all very interesting. It looks like your keystone mouth converts the driver-to-mouth horn section into a tuned chamber, as you would find it in a bandpass. Did you take any measurements w/ the mouth at very end of the horn (i.e.: the bottom of the enclosure)?

Regards,
Using the bottom of the enclosure as the mouth exit would make the cabinet 45 inches deep and only 22.5 inches tall, which would not work for the type of gigs I work, so I never cut the shell to try that configuration.

I have not used 45" deep cabinets since around 1987, when I found that they were too deep to fit in front of the fire curtains in most of the theaters we worked in.

All my tests were with the mouth on the front of the cabinet, same as most of the simple one or two fold THs one sees in the Collaborative TH threads, a few examples below.

The mouth, as shown in your modification of my fold diagram below, would be almost exactly the same area as the Keystone exit.

"Tapped horns" are much like a pair of tuned ducts attached to one speaker.
Reducing the mouth area does create bandpass results, as can be seen in my prior post showing "step down" response, but the Keystone exit is not reducing the mouth area.

In the case of TH similar to ones (other than the modified one showing the mouth on the bottom) below, a keystone shaped exit will probably afford some improvement in reduction of upper dips and peaks.

Art Welter
 

Attachments

  • Front exit.jpg
    Front exit.jpg
    55.2 KB · Views: 2,632
I had thought that the dip/peak thing was improved by keystoning because the driver in your TH was deeper in the horn path. It's easy to observe this with the sliders in HR. I would not expect much of an improvement where L34 is short...
I found a lot of things that were not as expected when I did actual cabinet changes and measurements.

The keystone exit shape did work somewhat like shortening L34 in Hornresp.

You may be correct that starting with a short L34 may minimize the positive effect of a keystone shaped mouth, until someone tries it, we won't know for sure.
It would be an easy experiment for anyone with the typical "tall boy" TH (and some test gear) to try.

Art Welter
 
Post #23/42

Hi Art,

I didn't mean to imply that you should use the box in a different aspect, but was just pondering if you had among your many measurements one all the way at the end of the horn path; I would assume, that your TH18 300 sq measurement in Post #23 (the brown trace) is with the mouth at the bottom of the front of the cabinet, i.e.: in a normal TH mouth position. By the way in Post #23 you have the keystone exit as: "...6 inch wide at the top, 12.5 at the bottom, and 25 tall, 306.25 square inches...". I did a quick check, and it ends up to be 231.25in^2.

Regards,
 
Hi Art,

I didn't mean to imply that you should use the box in a different aspect, but was just pondering if you had among your many measurements one all the way at the end of the horn path; I would assume, that your TH18 300 sq measurement in Post #23 (the brown trace) is with the mouth at the bottom of the front of the cabinet, i.e.: in a normal TH mouth position. By the way in Post #23 you have the keystone exit as: "...6 inch wide at the top, 12.5 at the bottom, and 25 tall, 306.25 square inches...". I did a quick check, and it ends up to be 231.25in^2.

Regards,
Oliver,

Your assumption is correct, the 300 square inch measurement is with the mouth at the bottom of the front of the cabinet.

Thanks for pointing out my math error, it makes things even more interesting.

The 231 square inch Keystone exit behaves quite differently than a 200 or 250 square inch opening on the bottom front.
Below are traces with 200 and 250 sq.in., the TH18 S is the 231.25 square inch Keystone exit. The BCBR38Fb is a ported cabinet, tuned to 38 Hz, also using a BC18SW115-4 speaker at the same drive level.

TH18 S Glued is the finished Keystone, it appears the addition of the front panel glue block, a 3/4 inch restriction, caused an almost 6 dB drop in the 180-225 Hz area.
There are two front braces midway along the exit that also may be contributing to that loss.

Interesting to see the 180 degree difference in phase at 40 Hz putting the opening on the bottom has.

My traces used a 20 Hz 4th order Butterworth high pass and 1K Hz 4th order Butterworth low pass.
The Keystone phase and frequency response is very similar to a DSL TH-118.

The blue trace is a pair of Langston Holland’s 4Ω TH118's, green is a pair of 8Ω TH115's with the addition of 20Hz 4th order Butterworth high pass and 78Hz 4th order Linkwitz Riley low pass filters.

PSW Sound Reinforcement Forums: Product Reviews: Sound Reinforcement => Danley TH118 Review and Measurements

The white and purple traces are his same cabinets with no processing.

Art Welter
 

Attachments

  • Keystone Phase.png
    Keystone Phase.png
    191.7 KB · Views: 2,509
  • L.Holland'TH-1115,118 raw.png
    L.Holland'TH-1115,118 raw.png
    168 KB · Views: 2,308
  • L Holland's TH processed.jpg
    L Holland's TH processed.jpg
    62.6 KB · Views: 2,174
I have now used the Keystone Subs on several gigs and 14 bands ranging from acoustic soloists to hip hop, rock, blues and world beat.

I have no earth shaking difference in sound quality to report compared to the previous subs I used, Lab 12" in ported cabinets. Basically, what went in came out sounding the same, response was smooth over the coverage area.

The efficiency of the Keystone cabinets did make a positive impact. The outdoor shows were at a venue with inadequate wiring, resulting in voltage dropping from 117 volts with the system powered up to as low as 107 volts on bass peaks.
Working the same venue for previously with the ported cabinets, the sub amps were clipping and I still wanted more level, the low voltage actually reduced peak output by about 6 dB.
With the Keystone Subs, never hit the clip lights and had more level, and less voltage drop.

Although it is hard to say definitively without a side by side comparison, it did seem that the directivity of the tapped horn design is useful. Acoustic upright bass and a grand piano parked within two feet of the subs required less EQ to get more gain before feedback than I would have expected using bass reflex subs.

Low frequency output was great, even on some trance dance music I heard nothing in the headphones lower than what the subs were putting out.

Art Welter
 
Drawing for Keystone Sub

Hi Y'all,

Had a little extra time this weekend, and came up with a drawing of Art's sub. Art may have some information to improve the accuracy (as I don't really know what he actually build), but at least this puts the basic construction hints and data in one place.

Regards,
 

Attachments

  • Keystone_sub_dwg_1.pdf
    51.7 KB · Views: 1,069
Last edited:
Hi Y'all,

Had a little extra time this weekend, and came up with a drawing of Art's sub. Art may have some information to improve the accuracy (as I don't really know what he actually build), but at least this puts the basic construction hints and data in one place.

Regards,
Oliver,

Good job with the drawing, pretty close to what I built.

As built, the expansion is even from S1 to S5, there are no side expansion ramps as shown on your drawing upper right.

The baffle has a 1 degree angle at the bottom, horn A has a 5 and 3 degree angle.

The actual cabinet uses nine braces inside. Two “wing braces” connecting the baffle to the Keystone front panel two “brace 1”, two brace 2, one brace 3, one brace 4, one brace 5. All braces have angles, best determined when lofting the cabinet.

The 3/4” x 3/4” vertical cleats around the removable Keystone grill (29 and 7/16” tall) were omitted in my line drawing.

Art Welter
 

Attachments

  • Keystone Sub.jpg
    Keystone Sub.jpg
    40.2 KB · Views: 1,834
Hey Art, it's obvious you went through many iterations on this design. Was there anything in your data that gives a guideline for the aspect ratio of the keystone?
Post #36 shows the various exit shapes tried.
The aspect ratio of the Keystone was arrived at empirically, measuring the response, then changing one size and shape at a time, using clamps and pieces of plywood to change the shape.
 
Hi Y'all,

Had a little extra time this weekend, and came up with a drawing of Art's sub. Art may have some information to improve the accuracy (as I don't really know what he actually build), but at least this puts the basic construction hints and data in one place.

Regards,

excuse my novice status, but is there any point in making the first 2.5 feet of the 'horn' tapered instead of parallel?
 
Hi 60ndown,

"...is there any point in making the first 2.5 feet of the 'horn' tapered..."

I wish I could give you a definitive answer to that question. In the case of the 18SW125 I would say: yes. Anyway, the section from S1 to S2 can be used to tune the response (the dips) around the first of the two major peaks just above the normal passband of a tapped horn. It also affects the height of the peaks and valleys leading up to that response area (we're messing w/ a few different aspects of the horn geometry here). It all depends very much on the specific driver used. Weltersys empirically developed an enclosure, that according to his many measurements works with a wide variety of drivers, so my gut feel is that you could leave that section straight, and you would not hear the difference in the passband. On the other hand, putting some stiffeners between the driver baffle board, and the parallel horn divider can't hurt. What I recommend: prior to cutting wood, put (or have somebody do it for you) your enclosure design into Hornresp using measured Thiel/Small parameters for the drivers you decided to use, and work from there. Even if you cannot model the Keystone exit in Hornresp, you can still get a general idea as to what may be going on.

Now that's fun. :)

Regards,
 
Originally Posted by 60ndown
if i wanted to get a flavor of what your keystone box sounds/works like, would i be completely off 'bass' trying this?

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-036

Specs don't show the Bl (magnetic strength) a very important specification.
The Xmax is low, so even if it has enough magnet strength, the speaker won't get anywhere near as loud as two Lab 12, or the BC18SW125-4.
That said, it will fill a hole until you can afford a real speaker.
 
excuse my novice status, but is there any point in making the first 2.5 feet of the 'horn' tapered instead of parallel?
I evidently was not clear enough in post 49 where I posted a picture of the actual plans.

Although Oliver’s (tb46’s) plan in post 48 is fairly close to the way the Keystone was built, other than the exterior dimensions none of the dimensions in his plan are the same as the actual cabinets I built.

The first section of the horn does not use parallel walls, it is tapered.

The expansion is uniform, as close to a straight line “V” as possible considering the turns, from beginning to end.

Art Welter
 
Hi 60ndown,

Weltersys empirically developed an enclosure, that according to his many measurements works with a wide variety of drivers, so my gut feel is that you could leave that section straight, and you would not hear the difference in the passband. On the other hand, putting some stiffeners between the driver baffle board, and the parallel horn divider can't hurt. What I recommend: prior to cutting wood, put (or have somebody do it for you) your enclosure design into Hornresp using measured Thiel/Small parameters for the drivers you decided to use, and work from there. Even if you cannot model the Keystone exit in Hornresp, you can still get a general idea as to what may be going on.

Now that's fun. :)

Regards,
Oliver,

The cabinet was built after many iterations with Hornresp.
The response came out similar to the simulations, but none of the sims could tell me what I needed to figure out empirically regarding the exit size and shape.

Other than left and right, top and bottom, there are no parallel horn walls in the Keystone as built, and there is extensive bracing, an absolute must for a good sounding high powered sub.

Art Welter