PDA

View Full Version : 2235H speaker box plans



Pengvim
01-24-2010, 09:11 AM
Hi.

I'm in the planning project of building a 2-way active speaker system, each consisting of one JBL 2235H per speaker and a hornloaded compression driver. I've got a few different drivers, and Altec 511 horns.

I'm trying to get the cabinet no wider than 20" (50cm), and I'll try to mount the woofer a bit high to "lift" it a bit over the edge of my sofa (it's kinda covered by it on the left side).

Does anyone have some cabinet plans, or tips?

speakerdave
01-24-2010, 10:34 AM
Hi.

I'm in the planning project of building a 2-way active speaker system, each consisting of one JBL 2235H per speaker and a hornloaded compression driver. I've got a few different drivers, and Altec 511 horns.

I'm trying to get the cabinet no wider than 20" (50cm), and I'll try to mount the woofer a bit high to "lift" it a bit over the edge of my sofa (it's kinda covered by it on the left side).

Does anyone have some cabinet plans, or tips?

Yes: follow the total internal volume recommendations of the box design programs or the JBL PRO Enclosure Guide obtainable by google. You can change the shape to fit your space as long as you avoid extremes, but keep the volume and porting the same. Avoid internal dimensions that are the same or integer multiples of each other.

Your horns will overhang the sides. Did you know?

Ruediger
01-24-2010, 10:44 AM
Hi.

... and I'll try to mount the woofer a bit high to "lift" it a bit ...



Woofer and horn should be as close together as possible.

Ruediger

Loren42
01-24-2010, 02:07 PM
A 20" wide baffle will give you a diffraction step loss at 228 Hz. That means, at 228 Hz the bass will roll off between 3 dB to 6 dB per octave as you go down in frequency.

The exact amount depends on room acoustics, but the 228 Hz is the point where the speaker cabinet stops radiating sound in half-space and begins to radiate into full-space. Essentially, some of the bass's sonic wave will wrap around the cabinet and radiate backwards. Obviously, you loose some of the bass.

Ideally, you would like to size the front baffle width to correspond to the crossover point of the speaker, but I doubt that will be the case with your design. Wider baffles lower the diffraction point according to this formula:

f = 4560/Bw

f = Frequency point
4560 is a magic number
Bw = Baffle width in inches

I use a 2235H in my 3-way speaker I built. My cabinet is 6 cubic feet internally after subtracting the volumes of the internal braces, ports, and displacement of the 2235H woofer.

I have two 4" flared ports 6" long. The box is tuned to 30 Hz with the volume/port dimensions cited.

Two ports are used to keep port noises down because the 2235H can move a lot of air.

Stuffing is nominal with 1" polyfill stapled to about 80% of the interior walls.

My woofer is crossed over at 400Hz with a 12 dB/Octave Linkwitz-Riley slope.

Crossovers are the heart of the speaker system. You will need to design a custom crossover if you are using a passive crossover. If you go active, as you stated, then you can bi-amp and improve the system significantly.

Pengvim
01-24-2010, 03:32 PM
Your horns will overhang the sides. Did you know?

Indeed I did. ;)

Thinking about trading them for a pair of 811'. That is, if the 2235H's can go up to 800Hz.

Pengvim
01-24-2010, 03:48 PM
The exact amount depends on room acoustics, but the 228 Hz is the point where the speaker cabinet stops radiating sound in half-space and begins to radiate into full-space. Essentially, some of the bass's sonic wave will wrap around the cabinet and radiate backwards. Obviously, you loose some of the bass.


Can't I counter that with a little EQ'ing? Damping the treble a bit, or boosting the bass?

Could it help to place the speakers close to the wall?



I use a 2235H in my 3-way speaker I built. My cabinet is 6 cubic feet internally after subtracting the volumes of the internal braces, ports, and displacement of the 2235H woofer.

I have two 4" flared ports 6" long. The box is tuned to 30 Hz with the volume/port dimensions cited.

Two ports are used to keep port noises down because the 2235H can move a lot of air.

Stuffing is nominal with 1" polyfill stapled to about 80% of the interior walls.

My woofer is crossed over at 400Hz with a 12 dB/Octave Linkwitz-Riley slope.


And all this adds up to? Do you achieve an extended low frequency, in the (perhaps a bit extreme) way I want?

Loren42
01-24-2010, 06:55 PM
Can't I counter that with a little EQ'ing? Damping the treble a bit, or boosting the bass?

Could it help to place the speakers close to the wall?



And all this adds up to? Do you achieve an extended low frequency, in the (perhaps a bit extreme) way I want?

You can compensate for the loss, but using tone controls is not the way to go.

This article explains the concepts and lists a simple circuit to do this. The circuit should be between the preamp and power amp.

http://sound.westhost.com/bafflestep.htm

Moving the speaker closer to the wall may and probably will have other unwanted effects. You could mount the whole cabinet in the wall and that would solve the baffle step issue. It has been done before, but not a very convenient method.

Resolving the baffle step diffraction issue helps make the speaker perform closer to the theoretical flat frequency response you want. However, room acoustics play a big part of the total picture and can make a great word class speaker system sound like poop if the room is not a good environment. Solving that requires adding "room treatments" such as bass traps and other sound deadening material to reduce the nodes and antinodes that occur. You may notice that sitting in one part of the room sounds poor compared to standing in some other area of the room. Murphy seems to always work such that the ideal place to listen to speakers in a room is standing upside down on the coffee table when the table is mounted to the TV.

However, I digress. The first order of business is to try to make your speakers behave the best that they can, i.e., flat frequency response. Once the speakers are correctly set up, then work on the room to get the best acoustics you can.

Some people resort to an equalizer, but the equalizer can only adjust the frequency response for one spot in the room. 18" in any other direction someone else hears a bad EQ. So, the room treatment is the right way to go.

Lastly, do not try to electronically boost the bass below the cabinet resonance point in a ported enclosure. You will blow the woofer.

Pengvim
01-25-2010, 11:31 AM
Isn't it possible to use some kind of cloth on the bafflefront, to absorb the waves that wrap around the cabinet?

But this baffle step compensation could be done in an active crossover, couldn't it? In a Behringer DCX2496 perhaps.
Or, when I'm building my own analog active crossover, I could make it an extra step..


Cabinet resonance point is the same as the tuning frequency, right?

Loren42
01-25-2010, 03:22 PM
Isn't it possible to use some kind of cloth on the bafflefront, to absorb the waves that wrap around the cabinet?

But this baffle step compensation could be done in an active crossover, couldn't it? In a Behringer DCX2496 perhaps.
Or, when I'm building my own analog active crossover, I could make it an extra step..


Cabinet resonance point is the same as the tuning frequency, right?

Trying to absorb the wavefront will simply aggravate the problem as you absorb sound energy that you want to be able to hear.

The best method is to design the front baffle so that it is either "infinitely" large, round off the baffle's edges/corners with a sufficient radius to minimize diffraction, or add a compensation circuit before the power amplifier.

Yes, when I spoke about the cabinet's resonance point I really meant the point it was tuned to.

Pengvim
01-25-2010, 03:52 PM
I think I will try out a compensation circuit first. I might try a rounding of the edges after a while, but I'm thinking that might be a bit harder to make as I have no experience (what so ever) with woodwork. I do know electronics though.

I'm thinking about something like this: (this is more of an illustration, most of this will probably be designed very differently)
http://i158.photobucket.com/albums/t105/Pengvim/delefilter%20design/aktivt_delefilter1.png?t=1264459951


First draft, with a crossover frequency of 569 Hz. I haven't set the attenuation yet, and I'll probably just remove the stage for the LF.
I think I'll skip some buffer stages in a final design though. If necessary, I'll add a delay stage.
Line-driver values are "stolen" from Linkwitz. :D

I'd like to try an Sallen-Key topology for the LP- and HP-filters, instead of the passive networks I've drawn.


My plan with regards to the crossover/filter will probably be to experiment with crossover-frequencies and such using a Behringer DCX2496 or a similar alternative, and then implement it in something like this schematic.

Loren42
01-25-2010, 07:56 PM
I think I will try out a compensation circuit first. I might try a rounding of the edges after a while, but I'm thinking that might be a bit harder to make as I have no experience (what so ever) with woodwork. I do know electronics though.

I'm thinking about something like this:

http://i158.photobucket.com/albums/t105/Pengvim/delefilter%20design/aktivt_delefilter1.png?t=1264459951

First draft, with a crossover frequency of 569 Hz. I haven't set the attenuation yet, and I'll probably just remove the stage for the LF.
I think I'll skip some buffer stages in a final design though. If necessary, I'll add a delay stage.
Line-driver values are "stolen" from Linkwitz. :D

I'd like to try an Sallen-Key topology for the LP- and HP-filters, instead of the passive networks I've drawn.


My plan with regards to the crossover/filter will probably be to experiment with crossover-frequencies and such using a Behringer DCX2496 or a similar alternative, and then implement it in something like this schematic.

Is it me or do you have the H_out and L_Out reversed?

Pengvim
01-26-2010, 12:04 AM
Is it me or do you have the H_out and L_Out reversed?


They are.

Pengvim
01-26-2010, 01:22 AM
Perhaps it could be possible to construct the speaker to play "with" the backwall?
I have a pair of Carlsson OA-50 set up in the livingroom, until I've gotten some DIY'ing done, and they play surprisingly deep bass from a tiny cabinet with a 6,5" midbass.
Perhaps it could be possible to do something like that, but with the 15" 2235H's?

Loren42
01-26-2010, 06:22 AM
Perhaps it could be possible to construct the speaker to play "with" the backwall?
I have a pair of Carlsson OA-50 set up in the livingroom, until I've gotten some DIY'ing done, and they play surprisingly deep bass from a tiny cabinet with a 6,5" midbass.
Perhaps it could be possible to do something like that, but with the 15" 2235H's?

If you have the test equipment that is one way to go about it, but it can be a little like throwing darts with a blindfold on. Without test equipment it is too subjective for my tastes.

I come from an engineering/scientific background, so my approach to problems tends to put a lot more cerebral effort up front first.

The basic process is to theorize, design, build, test & verify. Using those steps you generally take the shortest path to the best solution. It really is the best course to first understand the problem, then select an appropriate solution.

Then again, one of my Six Sigma instructors always told us, "If you can't be good, be lucky." :)

If it were me, I would try to learn as much on the subject as I could, then make a plan, and then execute that plan. You tend to learn a lot more that way and if you are persistent, succeed. Obviously, keep talking to people in the know to get feedback and new ideas.

Have you bought Dickason's 'The Loudspeaker Design Cookbook'?

Ruediger
01-26-2010, 07:48 AM
The 2235 datasheet shows it's response in a 140 l enclosure in a hemispherical freefield environment. Actually the response rises starting at 300 Hz. In the 4430 xover (elsewhere in this forum) the woofer section is overdamped and compensates this rise.

Ruediger

Pengvim
01-26-2010, 10:54 AM
If you have the test equipment that is one way to go about it, but it can be a little like throwing darts with a blindfold on. Without test equipment it is too subjective for my tastes.

I come from an engineering/scientific background, so my approach to problems tends to put a lot more cerebral effort up front first.

The basic process is to theorize, design, build, test & verify. Using those steps you generally take the shortest path to the best solution. It really is the best course to first understand the problem, then select an appropriate solution.

Then again, one of my Six Sigma instructors always told us, "If you can't be good, be lucky." :)

If it were me, I would try to learn as much on the subject as I could, then make a plan, and then execute that plan. You tend to learn a lot more that way and if you are persistent, succeed. Obviously, keep talking to people in the know to get feedback and new ideas.

Have you bought Dickason's 'The Loudspeaker Design Cookbook'?

I have that kind of background to in a way, as I'm taking a bachelor in electronics. I haven't had that much acoustics though, so I don't know much about that (yet).

I haven't bought Dickason's book, but I probably should place an order on amazon later. Is there anything else I should pick up while I'm at it? ;)


There is always the possibility of building a new enclosure after learning more though. I'm a bit busy in the daytime at the moment, so I don't have unlimited time for hobby-learning unfortunately.
So I'm thinking of building something a bit more quick&easy first, and then perhaps something a bit more fancy in the summer...

Loren42
01-26-2010, 12:03 PM
I have that kind of background to in a way, as I'm taking a bachelor in electronics. I haven't had that much acoustics though, so I don't know much about that (yet).

I haven't bought Dickason's book, but I probably should place an order on amazon later. Is there anything else I should pick up while I'm at it? ;)


There is always the possibility of building a new enclosure after learning more though. I'm a bit busy in the daytime at the moment, so I don't have unlimited time for hobby-learning unfortunately.
So I'm thinking of building something a bit more quick&easy first, and then perhaps something a bit more fancy in the summer...

Since you are steeped with the learnings and usage of math, then Robert M. Bullock's 'Bullock on Boxes' will give you a pretty good mathematical view of acoustics.

If you want something even more complicated, look into Transmission Line Enclosures. I am itching to pick up the following software and do some virtual experimentation with TL design, which is supposed to offer superior bass response to a vented enclosure.

http://www.audioxpress.com/bksprods/products/sof-tlwrx.htm

Lastly, you need to get a good handle on crossovers. Dickason will give you a very good primer on the subject, but designing a good crossover is pretty involved, as you may already know a little from filter network design.

The crossover is the heart of the speaker design. I recently upgraded from a Part-Express special to a LEAP designed 3-Way for my 2235H, Audax PR170M, and Morel MDT-37. The difference was quite audible and really brings the system quality up to that of a world class speaker.

Expect to spend a few hundred dollars on the crossover components. However, Bi-amping or even tri-amping using an active crossover is even better than a high quality passive crossover. That's another subject in of itself.

Pengvim
01-26-2010, 01:15 PM
I've had a whole subject (a third of the subjects of last semester) about filters. Didn't get good grades, but learned much though. Designing a good crossover from the bottom takes a lot of time though, but making the filter is probably the least of my problems. :)

Transmission Line is intriguing, but I think I'll start with a simple bass-reflex design.
Trying to keep it simple in the beginning, and I have a hunch that a TL makes for more woodwork..:D

I'll start by ordering those three books, and a Behringer or BBE digital crossover. (Those are the two I've seen that looks reasonably priced and goodsounding)


How much power do you use to drive the 2235H's btw? I've got a lead on a pretty sweet poweramp with a 4x100W rating.

Loren42
01-26-2010, 02:27 PM
I've had a whole subject (a third of the subjects of last semester) about filters. Didn't get good grades, but learned much though. Designing a good crossover from the bottom takes a lot of time though, but making the filter is probably the least of my problems. :)

Transmission Line is intriguing, but I think I'll start with a simple bass-reflex design.
Trying to keep it simple in the beginning, and I have a hunch that a TL makes for more woodwork..:D

I'll start by ordering those three books, and a Behringer or BBE digital crossover. (Those are the two I've seen that looks reasonably priced and goodsounding)


How much power do you use to drive the 2235H's btw? I've got a lead on a pretty sweet poweramp with a 4x100W rating.

Yes, TLs will probably consume more wood. Not that the cabinet is really that much more complex, but a TL tuned to a 20Hz 1/4 wavelength tapered tube is going to be about 14' long when unfolded if memory serves. That is a lot of wood, but the real complexity is in the design. Simple is a good starting point.

The 2235 is rated for something like 93 dB/W SPL. I use a 50 - 60 WPC tube amp that I built, which is plenty of power to drive my speakers to the point of hearing damage, I am sure.