My Drew Daniel's inspired DIY project.

[Baron030]

Well, I have finally gotten very close to finishing the construction phase of my little DIY project. So, I decided to take a little break and take some pictures to share.

As you can see, it just a plain brown box. Why, I didn't even try to make it look like any existing JBL product. So, this enclosure is a total sleeper...

Baron030

[Baron030]

Check out this view of the back, nothing looks unusual here, right?
Baron030 :)

[Baron030]

Well, the system is quad amplified, and I am feeding it with just a wee-bit of power.
An Ashly XR4001 4-way electronic crossover feeds to a K2, K1, and 3 D-75a amplifiers.
So, in my small apartment, clipping is never a problem.
Baron030

[Baron030]

Here is another view of the back, and with something that you don't see every day.
I will give you a little hint, it's not a port.
Baron030 :)

[Baron030]

OK, this really does give it all a way...It's a lift handle.
These boxes started out weighting in at 75 lbs. each, with just the JBL drivers alone.
So, lift handles really are a helpful addition.

[Baron030]

Here is an inside view from the back.

[Baron030]

And here is a close up view of the Doggy Box, which will house a 2012H driver.
It has a net internal volume of 0.5 cu. ft. and 2 - 2.03125” diameter x 2.1" long ports.
The WinISD program calculated the port tuning to be 98.5 Hz.
Baron030 :)

[Baron030]

Here is a close up view of the LF enclosure section, which will house a 2226H driver. The LF enclosure section has a net internal volume of 5.0 cu. ft. and 3 - 3.937" diameter x 10.5" long ports. After running some test, my plan is to shorten these ports down, until the enclosure is tuned to 40 Hz. precisely.

[Baron030]

A close up view of the LF enclosure section again. This time with the fiberglass installed.
I must thank Zilch for the 3M spray adhesive advice. :cheers:
The 3M spray is really the best way to install fiberglass.

I also sprayed several light coats of flat black paint on the areas that might be visible from outside the enclosure.

[Baron030]

Here is a close up of the HF enclosure.
The plywood and cork ring is to provide support for a 2446H "Big Boy" driver.

[Baron030]

OK, I will finally, stop teasing all of you, and show you a front view with the grill frame removed. I made the opening for the 2405 "Slot", 2382A horn, and 2012H 10" driver, so close together that the drivers actually kiss each other. And the opening for the 2226H is only slightly off-set, its’ outside edge is vertically aligned with the 2382A horn.

[Baron030]

Hopefully, in the next few weeks, I will be moving the components from these temporary quarters and into the new boxes.
Baron030 :)

[speakerdave]

Baron, you are an inspiration and an admonishment to people who blog too much. Very nice indeed.

David

[Robh3606]

Very Nice!!

He's what got me going as well. You going to run subs as well??

Rob:)

[spkrman57]

I will be watching with interest for the final product!

If you were to leave the tuning of the ports flexible, you might use 2235 in the future for deeper LF response. Just a thought!

Thanks for sharing with us!:applaud:

Ron

[yggdrasil]

Very nice.

Great work on those boxes. :applaud:

[Chas]

You're going to use those things in an apartment? Holy crap! Nice work...:)

[Baron030]

I am very surprised that none of my neighbors ever complained about the noise coming from my power tools. These boxes were constructed in my kitchen dinette area, over a period of nearly a year. That’s a lot of noise in a very small area.

Since, I don't own or have access to any large bench sized tools. All of the work was done with a hand held circular saw and a plunge router. Each piece of wood was first rough using a hand held circular saw. And then a router was used to mill all the wood edges straight and smooth. This process involved clamping 4' long straight edge ruler to the stock and then using the ruler as a "rip fence" to guide the router. While, it does take a lot more time, using router to make precise cuts, then it would using a table saw. The finished results can be just as precise as with any large bench tool. Unfortunately, this "rip fence" & router technique does limit your projects to only 90 degree butted-type wood joints. A 45 degree locked-miter or spine-miter joint really does require a large bench tool and a lot of skill to successfully pull off.

About the only angled cuts that I made in this the entire project, involve the construction of the doggy box. The back wall of the dog box is angled, so that the depth varies from about 4 to 8 inches. And the process of milling the back edges of the dog box to a compound angle was very involved. It required the mounting a router onto a large scrape of plywood and then mounting some 1" x 2" rails onto the box for the plywood to ride on during the milling process.

Since, I have had this new system up and running for nearly a year now. I have not had any complaints about load music. But, if I ever wanted to get out of my lease, it would be a very simple matter of playing “Pink Floyd’s Time”, at about 3 o’clock in the morning.

Actually, “Pink Floyd’s Time” and a few tracks on “Van Helen’s 1984” album, are the only sources that I have found so far that will cause the “Green signal present lights” to flicker on the Crown D-75 amplifier that is feeding the 2405s.

Oh, and responding to Spkrman57’s question about the possibility of swapping out the 2226s for some 2235s. If at some point in the future, I should start feeling “Bass Deprived”. I would rather follow Drew Daniel’s advice and add some sub-woofers. Don’t you think that 4 – 2242Hs or a pair of those new 2269Gs would do a better job of fattening up the bottom end?

Baron030

[Steve Gonzales]

Damn those are sweet Baron030!. Something to be extremely proud of and a real treat to see here in detail. I just love this kind of thread, it really gets the creative juices flowing. Congratulations on a fine project. Steve G. :applaud: :applaud: :applaud:

[Don Mascali]

Oh, and responding to Spkrman57’s question about the possibility of swapping out the 2226s for some 2235s. If at some point in the future, I should start feeling “Bass Deprived”. I would rather follow Drew Daniel’s advice and add some sub-woofers. Don’t you think that 4 – 2242Hs or a pair of those new 2269Gs would do a better job of fattening up the bottom end?

Baron030

Now that's the ticket... TOO much is just right!

Beautiful work.

[Baron030]

Over this last weekend, I finally finished installing all of the drivers into the new enclosures. The attached photo is a composite of two separate images and it gives you panoramic view of my living room.

With the construction phase of this project nearly completed, I can now focus on the tuning and tweaking this new system. Also, I will have more time to post and share what I have learned of far.

Currently, I have the system is up and running with everything pretty much roughly dialed in.
So, it’s sounding pretty amazing. :dancin:

Needless to say, this is a major improvement over my old 030 system.
Baron030 :rotfl:

[Don Mascali]

Very nice work. :applaud:

[johnaec]

With the components you're using, those would make also killer speakers for a band! :rockon1:

John

[spkrman57]

You really do want to break your lease!:blink:

Ron

Quoted from Baron030:

Oh, and responding to Spkrman57’s question about the possibility of swapping out the 2226s for some 2235s. If at some point in the future, I should start feeling “Bass Deprived”. I would rather follow Drew Daniel’s advice and add some sub-woofers. Don’t you think that 4 – 2242Hs or a pair of those new 2269Gs would do a better job of fattening up the bottom end?

[JBLnsince1959]

Oh, and responding to Spkrman57’s question about the possibility of swapping out the 2226s for some 2235s. If at some point in the future, I should start feeling “Bass Deprived”. I would rather follow Drew Daniel’s advice and add some sub-woofers. Don’t you think that 4 – 2242Hs or a pair of those new 2269Gs would do a better job of fattening up the bottom end?

Baron030


:applaud: ;)

Great job....:bouncy:

[glen]

Talk About A Lease Breaker Special

You're going to use those things in an apartment? Holy crap! Nice work...:)
I dunno, he may have a ways to go yet...

http://audioheritage.org/vbulletin/showpost.php?p=80341&postcount=64

(http://audioheritage.org/vbulletin/attachment.php?attachmentid=11431&stc=1&d=1133801619)

[Turnitup]

Awesome Craftmanship! I wish I was your neighbor, so I could hear'em
everyday!

[Baron030]

The plan that I had in mind when I started my project was have to the 2226 drivers enclosed in a box with net internal volume of 5 cubic foot and have the bass reflex ports tuned to 40 Hz. This would follow the recommendations on the JBL 2226H spec sheet.

Using the WinISD program, I calculated that I would need to have 3 ports with an inside diameter of 3.94” and a length of 9.4 to reach my target tuning of 40 Hz. Well, being a little unsure of the WinISD programs accuracy, I decided that I would initially leave the ports a little long at 10.5 inches. And then I could do some tests and trim the ports shorter as needed.

When I started testing my enclosures the 10.5” ports, I expected the tuning to be around a WinISD program predicted frequency of 38 Hz. But, my test results clearly showed that the tuning to be a about 31 Hz. :confused:

The only explanation that I have for this error is that I forgot to take into account on the effect of using R-11 fiberglass insulation on all of the enclosures inside surfaces. :banghead: :banghead: :banghead:
The fiberglass has clearly increased effective acoustic volume of the enclosure dramatically. So, the really big question is this. What to about this virtual volume issue?

One possible solution would be to remove the R-11 fiberglass and then replace it with a thinner layer fiberglass. This would hopefully shrink the virtual volume back down to match it’s actual volume.

Or I could leave the R-11 fiberglass alone and accept that the enclosure is now “virtually” much larger and possibly come up with a slightly different enclosure tuning.

Lately, I have been experimenting with different ports lengths and I am starting to see a pattern emerging. After, running tests with 9.25” and 8.0” port lengths, it appears that my test results are very closely matching the WinISD predicted results for an enclosure volume of about 7.5 cubic feet. The 9.25” port tuning came in at about 33Hz. and the 8.0” results are posted below, which are coming just in a little below 35Hz.

And from just listening to my system, I have found that the 8” port tuning sounds a lot better than the other port lengths that I have tried. So, this would suggest that I am getting a little closer finding the correct enclosure tuning.

But, I am not sure what I do next. Should I pull out some fiberglass or shorten the port tubes further? :blink:

So, any advice here would be more than welcome.
Thanks, Baron030

[Zilch]

BB6P says 3 x 3.938" vents at 8.065" long will give 40 Hz tuning with typical (all sides but the front baffle) fill.

"Heavy" fill lowers it to 37.95 Hz.

I'd leave the R11 in place and continue trying shorter vents until you achieve the tuning you like best for your room and listening preferences, which may be something other than the target 40 Hz.

The insulation is doing more for you than just adding volume, i.e., damping.

R11 is kinda low density, so I'd say you have the equivalent of typical fill in place, and should leave it there.

It's common for the simulated port lengths not to produce the exact tuning predicted in the built box. There are other variables involved, and the empirical method you're using is almost always a part of the design finalization process.

Note: Fmin provides an "approximation" of the box tuning only, so don't consider it absolute. There are more precise means of determining the actual Fb. For example, at what frequency is woofer cone movement minimal?

Clearly, you are close to where you want to be. Just tune to taste, now.... :thmbsup:

[spkrman57]

And also, you can adjust one of the ports by itself and tune it that way instead of messing with all 3 ports.

Ron

[Rudy Kleimann]

Something is wrong with your math here. No other reason than that. Not to worry too much, though, unless your net volume plus damping yields an effective volume that is way off the mark. A little off here and there isn't going to matter much, if at all. Cabinet size is often adjusted from the theoretical to the practical for aesthetic or other reasons, namely size or fit in a particular installation. Your ears will tell you... they are your best test instrument;) , and the receiver of good vibrations that make the musical experience... never lose sight of this fact while busying yourself with the math and science of it all:banghead: .

As Zilch pointed out, actual testing of the finished system is the only way to see what you really, really ended up with. This includes the ear test:D

A couple of points:

Insulation, particularly fiberglass, adds virtual volume by converting the energy in the air from adiabatic (constant heat) to isothermal (constant temperature) which brings about a reduction in the speed of sound. loosely filling a box with insulation can easily increase the effective volume 20%; theoretically it can increase it by 40%. If you were shooting for 5CF, but your tests indicate 7.5CF, something is wrong here, either your physical dimensions are innacurate, or your test method to determine box tuning frequency is wrong. Insulation improves damping, which controls ringing and overshoot when the driver is fed a transient signal. R-11 is fine; I wouldn't use any less than this on the cabinet walls. This should add about 10% virtual volume, or as Drew has commented in the JBL enclosure guide, should about amount to about the same volume displaced by driver(s), bracing, and ports. A little thicker should add all that much more.

The same size port in a larger box will yield a lower tuned frequency. As Zilch pointed out, the design programs will get you close -if your physical dimensions and volume calculations are accurate. Always start long on the port as long as it is not too close to any obstruction i.e. a cabinet wall or bracing, and cut it to length to hit the actual tuned frequency desired. If your box volume is too big (or you think it is), fill the void with bricks, sandbags, or other solid, inert, incompressible material and give it a listen.

How exactly are you determining the tuned frequency of your box? You should be able to get close by feeding the woofer a sine wave at 100-200 watts, carefully observing cone movement as you sweep down through the frequencies below 60 Hz. At the box tuned frequency, port output is at maximum (as felt by your hand) while at the same time, the cone movement is at its' minimum (practically stops moving). The accurate method involves an oscilloscope and a resistor (PM me if you have a 'scope and want the setup instructions).
As you go below the tuned frequency, the cabinet ceases to load the woofer, and cone movement increases rapidly as you continue sweeping down. At this point, STOP! If you continue sweeping down, woofer excursion increases to the point where the voice coil can crash into the rear plate of the magnet and damage the coil. It can also be hard on the spider, surround, or even the cone itself. I've seen cones curl back near the suspension from over-excursion and even develop a circular fatigue line in the paper cone near the dust cap. This quickly leads to driver failure from severe over powering or driving below the ported box frequency. Just because the coil can handle the power doesn't mean you can't break the speaker...

[Rudy Kleimann]

And also, you can adjust one of the ports by itself and tune it that way instead of messing with all 3 ports.

Ron

I don't know about that... I'd plug a port, but I wouldn't have different length ports... seems wrong to me.

[Baron030]

Your responses made me really question my test results. So, it forced me to go back to “The Book”. “The Book” in this case is “Testing Loudspeakers” by Joseph D’Appolito.

The excel chart that I included in my last post was created using a voltage divider technique described in Joseph’s book. For test equipment, I used a freeware computer program called SigJenny and an inexpensive digital multimeter. I have tested accuracy of the SigJenny program, using oscilloscope and I have found it to be very accurate. So, I know that the SigJenny program is not at fault here. Since, I don’t have my computer hooked up to my stereo. I used the SigJenny program to create my own DIY Port Testing CD with each track spaced just 1 hertz apart.

Getting back to my excel chart, all of the data contained in it is fairly accurate. But, I have completely misunderstood it meaning. After taking some additional impedance measurements, this time with the ports closed, I have now gathered all of the information necessary to properly calculate the enclosure tuning. The missing bit of information is the closed-box resonant frequency of my enclosures, which just happens to be 48 Hertz.

Fb = box tuning frequency
Fc = the closed-box resonant frequency
Fh = the frequency of the higher-frequency impedance peak
Fl = the frequency of the lower-frequency impedance peak
Fm = the frequency of the minimum impedance between the two peaks

Fb = square root of ( (Fh x Fh) + (Fl x Fl) – (Fc x Fc) )
So, plugging in the data values for my left enclosure into the formula above, the box tuning works out to be:
38.74 Hz = square root of ((58 x 58) + (21 x 21) – (48 x 48))
And for the right enclosure tuning works out to be:
37.23 Hz = square root of ((57 x 57) + (21 x 21) – (48 x 48))

These numbers come in fairly close to BB6P predicted values for an enclosure with a “heavy” amount of fill. Since, my project has a dog box, there is larger amount of surface area relative to its volume, which may explain why my project comes in with a heavier amount of fill in it.

Normally, the Fm value is very close to the Fb value. But, in my project, the Fm value appears to be much lower than the Fb value. And that is where I got really confused. :confused:

Over this last week end, I did try to observe at what frequency the cone movement was at its minimum. But, with separate test tones for each frequency, I found this test method next to impossible to judge accurately. So, I think I will stick with the voltage divider method for now.
Thanks again Zilch and Rudy, you both deserve a cold frosty one. :cheers:
Baron030

[Zilch]

Heh. You are "Golden" on tuning, then.... :thmbsup:

[JBLROCKS]

:applaud: :applaud: :applaud: :applaud: :applaud:
FANTASTIC JOB!!!

[UreiCollector]

I'm very impressed! These must sound incredible!!

[spkrman57]

I'm curious what the final result on the port length ended up at.

Ron

[Baron030]

Hi Ron

I have gotten side tracked with the holidays and other projects for a long time now. So, I haven't gotten around to gluing the tubing in place and calling this project done yet. Currently, I got the ports trimmed down to 7.5" long. And my last tests showed that the tuning is still a little below 40 Hz. But, at this point, I realize that I am so very close to hitting the bull's eye that I am not sure if I want to trim the ports any shorter or not.
Maybe, with some better test equipment, I could be convinced that another ½" may have to come off. But, I am not sure that it would make any real difference in the sound.

The 2226s now have a great punchy sound and I don't feel the least bit bass deprived, yet.

Baron030 :)

[Joe Alesi]

Hello Baron,

Well done nice workmanship on your enclosures- I love the compression driver support and detailed routing on the front panel and matching routed grill. Make me want to get into a DIY project.
Best
Joe Alesi

[4313B]

So, the really big question is this. What to about this virtual volume issue?That's why we prototype...

At least you've experienced first hand why I've harped on proper enclosure tuning, virtual volume and enclosure losses countless times to nothing more than blank stares in return. :p

Nice project! :)

Your responses made me really question my test results. So, it forced me to go back to “The Book”. “The Book” in this case is “Testing Loudspeakers” by Joseph D’Appolito.

These numbers come in fairly close to BB6P predicted values for an enclosure with a “heavy” amount of fill. Since, my project has a dog box, there is larger amount of surface area relative to its volume, which may explain why my project comes in with a heavier amount of fill in it.
Good book to have in the personal library.

BB6P does a nice job.
Currently, I got the ports trimmed down to 7.5" long. And my last tests showed that the tuning is still a little below 40 Hz. But, at this point, I realize that I am so very close to hitting the bull's eye that I am not sure if I want to trim the ports any shorter or not.
Maybe, with some better test equipment, I could be convinced that another ½" may have to come off. But, I am not sure that it would make any real difference in the sound.

The 2226s now have a great punchy sound and I don't feel the least bit bass deprived, yet.:applaud:

[Dualbios]

Love this project really is helping me strive to get mine done

[Baron030]

Hi Dualbios

Other then enjoying this system, I have not done much more to enhance the system. In a way, I would not recommend the 2012H driver for your project. This driver has a rising response which does require some additional equalization. It has a strong peak at about 1,000 Hz, which requires a fairly high Q series tank circuit to bring under control. So, it’s not liked a lot around here.

Also, the 2380a/2446h combination will require a little high frequency shelving boost in the 3,500 to 8KHz range to get a flat response out to the 8Khz crossover point for the 2405 UHF driver. Here is a link to one of my early attempts at a 2012H EQ network, see response number 6: http://www.audioheritage.org/vbulletin/showthread.php?15658

I know I should post some updates on the EQ networks for this project. So, I will work on that soon. Oh, and for the record I did take red pill and bought “Cliowin 7 Lite”. So, with Clio, I can see just "how deep the rabbit hole goes." I know I am being little silly...

Baron030:)

[badman]

Consider scrapping the venting on the mid. You can start by just stuffing the vents, before trying to optimize the sub-enclosure. I'm assuming the tuning is below the XO, and is there for power handling, but really doesn't need it for any reasonable use.

Any crossover details you can provide?

[1audiohack]

I like 2012's and 2020's but I have never tried to tame them passively. 2123's are a snap though its true.

About that rabbit hole thing, there is no end, it is infinitely deep, how far you go in is up to you. For some of us it's an obsession, I know I can't stop thinking about it, no matter what I'm doing, audio stuff is running in the background in my mind.

Cough up that red pill if you can! :D

[Wardsweb]

Very nice indeed. A lot of work but something to be very proud of. Now just enjoy the fruits of your labor.

[badman]

Cough up that red pill if you can! :D

Nicely said. I currently have 8 12"s on my bedroom floor.

[Baron030]

Hi Badman

Here is a link to my 2012 passive EQ circuit: http://audioheritage.org/vbulletin/showthread.php?t=20823

Baron030:)

[Baron030]

With some renewed interest in my old project, I realized that I had never posted the passive high pass shelving network circuit for the 2446H/2382a combination. Just as a reminder this system uses an active crossover network and separate amplifier channels for each driver. The active crossover network is set to feed a 1,200Hz to 8,000Hz signal range to the 2446H driver. And a 2405 driver handles all frequencies about 8KHz. After running some tests with Clio, I did find that the 2446H/2382a combination does have a slight roll off issue that starts at about 4,500 Hz and builds to about a 5 db drop at 8 KHz. In the circuit pictured below the array of 82uF caps provide for driver protection. And the 5 and 20 ohm resistors form an L-PAD, which at lower frequencies attenuates the signal by about 5.5 db. The 30uf caps bypass the 5 ohm resistor at higher frequencies and this reduces the attenuation of the L-PAD to less than 1 db at 8 KHz. In practice, this simple shelving network does a pretty good job of flatten the 2446H/2382a combination out to 8 KHz.

Baron030:)

[maxwedge]

I like the retro style of your cabinets. Nice job!