AVS Forum banner

1 - 20 of 446 Posts

·
Registered
Joined
·
1,984 Posts
Discussion Starter #1 (Edited)
EDIT: This design is no longer experimental, it is proven to work. Feel free to build and enjoy.
Please pay attention to wiring listed here and in the Dev index, as it effects cone excursion.

Index link:
Devastator Quad JBL 12" 48"x28"x28"

I usually wait until builds are complete to post a comprehensive build thread, but this one is not all about me. We have enough interest in this one, and I feel that I owe it to the contributors to keep it more real-time. I will edit the main posts as progress continues, and add new posts to flag it when significant updates occur.

This is an experimental design by John, @Red Five, utilizing four of the “$30 JBL” 12 inch drivers in a quarter wave “ported horn” loading. It is experimental as we don’t know how multiple drivers located at different distances in the horn path will act.

First off, a huge thank you to those who have contributed drivers and funds to make this experiment happen. Let’s get a round of applause for @MTBDOC, @gpmbc, and @Chris Popovich!

As for my involvement in this process, I just haven’t yet learned to not get involved. John was looking for someone willing to build the prototype. I already had all my tools out for past and future builds, and was picking up material for additional builds. Though I don’t have a use for the cabinet, I am very interested in the concept and believe it will help to open up more design possibilities. Therefore, I volunteered for the designated idiot/monkey-with-tablesaw position.

Now for the design itself:


Many more details and the cutlist will be posted if this design proves to be successful. As of now, this has a “Don’t try this at home!” disclaimer on it.


@Ricci. I’ve seen some of your awesome horn designs. Since you’ve dealt with multi-driver horns, do you have any thoughts or advice on this? Specifically variable excursion between drivers?
Thanks!
 

·
Registered
Joined
·
1,984 Posts
Discussion Starter #2 (Edited)
Now for the actual build process.
Disclaimer: This build is my version of rapid-prototyping. Mistakes are made, and fit-and-finish is not up to my normal standards. This cab will be left in raw-wood form.

Step one: purchase and haul some materials.

There are multiple projects in that load.
Current material-count in the Quad Dev is three 4x8’ sheets, two 8’ 2x4’s.
Another disclaimer: There are two more straps on the 4x8’s and 2x4’s. I prefer to actually make it home with the wood.

Time to cut some wood:


Yes, an outfeed table would make life easier.

To avoid offset measurement, I made a quick circular saw jig with 1/8 masonite and a cheap 4’ beam level.

I plan to rebuild it for the wide side of the shoe, to avoid balancing the saw.

I broke down some of the 4x8’s for easier handling on table saw:


I modified my table saw for up to 30” cut capacity. All larger cuts are made using circular saw and jig:


2x4’s and rails and braces were cut to length on the miter saw:


Next, I laid out the baffle and drilled driver centers with a 1/8” bit.


I cut shallow recesses at driver outside-diameter to make location easier:


Perform the driver cutouts:


Not really visible, but I added a 1/8” roundover inside the cutout to protect the surround:


Flip the baffle and lay out horn rail locations.


Mark the rail for driver clearance, and clamp on a guide for 1/2” cut depth.


Use a top bearing flush trim bit to cut rail clearance:

Use the first rail as a template for the second.

To avoid any differential pressure between drivers or horn passages, I decided to window the rails over the drivers.

I laid it out, drilled holes at the corners, and cut with a jigsaw.

I applied roundovers to try to minimize turbulence.

Ugly, but effective.

Here is the completed “flat pack”:
 

·
Registered
Joined
·
1,984 Posts
Discussion Starter #3 (Edited)
Assembly time!
I prefer to use PL Premium 3x construction adhesive for these builds. It does a very good job of sealing gaps, and has a much longer workable time than wood glue.

This was my first experience with the Kreg pocket hole joining system. I found it to be a quick and effective tool for enclosure assembly. The “Kreg Kreep” can be adjusted for by shifting the piece beforehand and using controlled torque to pull into place.
Precision assembly should be clamped to prevent creep.

I chose not to use any visible pocket holes, because I don’t think it looks good. This did greatly increase the difficulty of the build. I used a few brads as needed on the outside of the cab.

First, I assembled the horn rails onto the baffle, and the port rails onto the cabinet bottom.

On Devastator designs with a port back, the bottom port rails should be attached to the bottom port board instead.

Main cabinet assembly:
Lay out a side panel, the top panel, and the front panel. Cut pocket holes as appropriate:


Glue up these panels.

Make sure to carefully align all panels on the front face of the cabinet. This is the only face where overhang can’t be easily fixed with a router and flush trim bit.

Cut pocket holes in the baffle assembly:


Mark the back of baffle locations on the side and top panels to aid assembly:


Glue on the baffle:


Pocket-hole and glue on the port board:

Ignore the extra glue and pocket holes on front edge. “Somebody” tried to install it sideways.

Install the bottom port board. I clamped it, then used brads through the side panel, and screws into the port rails behind the baffle.


Install other side panel:

Due to warps in the ply, variable clamping and tweaking was needed to pull the cabinet into square. Notice attitude-adjustment deadblow hammer.
The front edge was pinned with brads, then pocket screws were tightened behind baffle.

Time for the bracing. Having built devastators before, I adjusted the bracing on the fly. These changes have now been added to the design files.

Flip the cabinet face-down, and mark fromt-to-back brace locations:


Pocket-drill braces:

There is quite a few extra and extraneous holes pictured. I adjusted the bracing design half way into drilling.

Glue in front-to-back braces:


Glue in vertical braces, and lower cross brace:

Ignore the cluster%#€£ going on with the lower brace here. I drilled them for a different assembly order without checking lengths. The design has been changed in the plan.

Glue in remaining braces:


Add back panel:

Make sure edges are well sealed.

Lay out and add bracing to hatch:


Congratulations, its assembled!
Let dry for 24hrs minimum.
 

·
Registered
Joined
·
1,984 Posts
Discussion Starter #4 (Edited)
Finish and final assembly:

Scrape excess dried glue out of hatch recess, and test fit hatch height and width. I trimmed by 1/16” height and 1/8” width for fitment. Don’t worry about corner fitment yet, just overall dimensions.

Apply 1/4” roundover to outside edge and corners of hatch panel. Do not roundover inside yet (oops).


Set in hatch panel upside down. It should not be tight on edges or corners now, and sit flat on 2x4 bracing. The hatch bracing will make it easily removable later.

Scrape glue from seams, then use router and flush trim bit to remove excess material from back of cabinet:


Remove hatch panel. Apply roundover to inside edge of hatch.

Flip cabinet, and flush trim remaining edges as needed. Be careful of port and horn openings.

Sand any rough seams/edges, then apply 3/8 roundover on all edges:

Ignore the big gouge in the port. Uhh... I’m pretty sure the ply came like that. I would never get in a hurry and go around the wrong corner with the flush trim. That would just be stupid.

Lay out screw pattern on hatch, and drill with 1/8” bit (assuming #10 screws):


Set hatch in place, and mark which way faces up. Predrill screw holes with 1/8” bit:


Remove hatch cover, and enlarge holes with 3/16” drill:


Cut terminal cup recess and cutout with router and circle jig:


Since I’m using 2 pieces of 12ga speaker wire, I used crimp terminals to join and attach to terminal cup. Then I soldered it together.


Insert terminal cup, predrill, and fasten:


Use a driver as a template to predrill the mounting holes:


Mount the drivers. I insert the screws with an impact gun, but tighten by hand.


Hook up wiring. I cut it short enough that it can’t rattle against anything. If you want to leave it longer, pipe insulation works well.


Verify correct resistance:


Add the gasket tape for the hatch:


Secure the hatch. Again, I run the screws in with an impact gun, but tighten by hand.


Here it is, one complete quad-Devastator, in the raw and ready to rock.
 

·
Registered
Joined
·
1,984 Posts
Discussion Starter #5 (Edited)
Measurement time!

I put together a test cart using a pair of Stewart HPA-1000 amps, an audiosource preamp, a minidsp 2x4 balanced, some old infinity speakers, and my modeling laptop.
I also set up a “control” sub, a PA460 tuned to 27hz, 3.5cf net.


The amp I’m powering the quad with is rated for 1000w at 4ohm.

I used this setup to break in the JBL’s free-air. I used REW to generate a 20hz tone, and turned it up till the driver was moving decently. I let it run for 30sec or so per driver. The suspension felt much looser afterwards.


First, a I took a close-mic measurement:


I set a 19hz 2nd order BW HPF in the minidsp. I left the LPF bypassed.

Sweep was run from 0 to 200hz.

Blue is measured response.
Green is hornresp sim using JBL data.
Orange is hornresp sim with ChrisP’s driver measurements and Josh’s semi inductance calculations.
Thanks for the data @Chris Popovich and @Ricci.

It looks like this one hit the semi-inductance simulated response nicely.
 

·
Registered
Joined
·
2,184 Posts
I am really excited for this thread :D It is the perfect horn formula, cheap subs, big box in exchange for a lot of output with a rational amount of power.
 

·
Registered
Joined
·
3,640 Posts
I am really excited for this thread :D It is the perfect horn formula, cheap subs, big box in exchange for a lot of output with a rational amount of power.

So if this is a success, when will the 3x3 Dev come out? 4500watts power handling, twice the size :p
 

·
Registered
Joined
·
2,184 Posts
So if this is a success, when will the 3x3 Dev come out? 4500watts power handling, twice the size :p

LOL. I am not sure if anything that ridiculous will occur, but I could look into a Dual SAN 21" box that is a 16Hz tuned monster. Something sort of like a SKHorn or MAUL...
 

·
Registered
Joined
·
829 Posts
LOL. I am not sure if anything that ridiculous will occur, but I could look into a Dual SAN 21" box that is a 16Hz tuned monster. Something sort of like a SKHorn or MAUL...

Yes PLEASE....MAUL style....


Juju
 

·
Registered
Joined
·
616 Posts
This will be interesting... Someone mentioned to me this was happening the other day. I have a few (lol) JBL 12s I only have room for sealed boxes that I need to make. But we will be moving soon and will have more room.... Good luck with the build and interested to see what sort of results you get!
 
  • Like
Reactions: a77cj7

·
Registered
Joined
·
1,984 Posts
Discussion Starter #15 (Edited)
Thread is now updated to current progress.

Thanks for the support!

EDIT: Driver pairs are currently showing Thursday and Friday delivery. Testing should happen this weekend.
 

·
Registered
Joined
·
1,309 Posts
I usually wait until ...

This is an experimental design by John, @Red Five, utilizing four of the “$30 JBL” 12 inch drivers in a quarter wave “ported horn” loading. It is experimental as we don’t know how multiple drivers located at different distances in the horn path will act.
.
.
Now for the design itself:

.
.
@Ricci. I’ve seen some of your awesome horn designs. Since you’ve dealt with multi-driver horns, do you have any thoughts or advice on this? Specifically variable excursion between drivers?
Thanks!

This is a great build "for the masses" that don't have $500 to spend on a driver but still want the Devastator wallop!

I thought HornResp can accommodate multiple drivers along the horn path, or is that not the case with this particular configuration? I guess I'm trying to understand why this is more experimental than most first-time builds based solely on simulated performance.

Thanks for going where no man has gone before! ( I couldn't help it, I just watched the first episode of Picard last night...)

Mike
 

·
Registered
Joined
·
1,984 Posts
Discussion Starter #17 (Edited)
This is a great build "for the masses" that don't have $500 to spend on a driver but still want the Devastator wallop!



I thought HornResp can accommodate multiple drivers along the horn path, or is that not the case with this particular configuration? I guess I'm trying to understand why this is more experimental than most first-time builds based solely on simulated performance.



Thanks for going where no man has gone before! ( I couldn't help it, I just watched the first episode of Picard last night...)



Mike


John @Red Five and I have been warned that HR doesn’t properly predict the excursion variance between drivers located at different locations of the horn path.

Supposedly, AkAbak does, but neither of us are familiar with it.
At this point, we’re going to wait and see what happens before delving into new software.

If anyone AkAbak-literate would like to model this, it would be greatly appreciated.

Chris
 

·
Registered
Joined
·
5,542 Posts
@Ricci. I’ve seen some of your awesome horn designs. Since you’ve dealt with multi-driver horns, do you have any thoughts or advice on this? Specifically variable excursion between drivers?
Thanks!

I do but it's a little late at this point isn't it since you already have it built!?

Most of my thoughts on it are in the MAUL development thread at the DB forum. The quick answer is it can have a major effect on the driver behavior, usually not for the better and can result in a loss of output and power handling. Messes with things like the FR and distortion levels as well. Excursion gets asymmetric between the drivers. It depends a lot on the drivers used and the enclosure design too. The MAUL was supposed to be 4x21Ipal per enclosure but it looked like junk after I modeled the drivers offset as they would be. For whatever reason the T3 19's seemed to be much more "immune". Not entirely sure what causes that difference in behavior between the 2 drivers.

All I can say for certain is the output will be slightly less than indicated by the HR sims but how much is unknown without a much more detailed model. Might be a nothing burger or it might be real bad. Unfortunately you cannot model this in HR. You have to go to Akabak or ABEC3 for that. Does anyone have measured specs from one of these JBL drivers? Ideally you want the semi inductance parameters included too. Those make a big difference. Not trying to be a downer here but it is what it is.

You've already got it built. Might as well fire it up and see what's what. It's a big cab with 4x12's in it so a loss of some output versus the ideal may not even be noticed.
 

·
Registered
Joined
·
1,984 Posts
Discussion Starter #19
I do but it's a little late at this point isn't it since you already have it built!?



SNIP


You've already got it built. Might as well fire it up and see what's what. It's a big cab with 4x12's in it so a loss of some output versus the ideal may not even be noticed.

Thanks Josh!

Yeah, I wasn’t aware if the severity of the potential issues until it was already built. I agreed to build it Friday, so this one moved kinda fast.

I have read the MAUL thread, but believe I missed the development one. I will check it out.

I believe someone has measured these drivers, not sure on semi-inductance though. I’ll look for that data.

Testing this one as-is is the plan. If it fails massively, there will be a rebuild IF the issues can be addressed.

Thanks for all the measurements and designs you’ve made and shared with the world! Completely awesome!

Chris
 

·
Registered
Joined
·
829 Posts
Thanks Josh!

[snip]......

Thanks for all the measurements and designs you’ve made and shared with the world! Completely awesome!

Chris
@Ricci

A HUGE DITTO to that one.....


Juju
 
1 - 20 of 446 Posts
Top