Ultra Light Weight Bass Cabinet

[Bass Unique]

I've seen quite a few OEM versions of the 3012HO, all with a lower Fs than the 3012HO, but otherwise very similar.

Just for those people who want a more technical explaination, if you add mass (more Mms) you can lower Fs but it would kill the broadband output (that's basically the LF version with its different cone and depth). To lower Fs and maintain output, you probably want to lower the stiffness only (surround and spider most likely modified) and leave the moving mass similar (stiffness relates to Cms). But that means the final system resonance is more dictated by the box size which dominantes the stiffness so a larger box is needed to prevent the system resonance climbing.

It is difficult to change one without affecting another and it is likely that the compromise would be Q or the linearity at high output. Also, if you change the cone surround to lower stiffness, it can cause the breakup mode of the speaker to be more abrupt - then you have to consider changing the cone shape to tailor the upper end to prevent such a severe response (peak then dip).

Alternatively, change the dustcap and coil former to a new material which is lighter and stiffer. (or winding material). This may be related to why the HO is only 400W not 450W although less excursion also limits cooling. This would lower mass slightly and could maintain Qts if a lower stiffness is employed. It could also boost dustcap emmissons at the top end extending response up by 500 Hz or so. Then, use a crossover to a tweeter around 3.5 k and all is well.... ok for a commercial manufacturer but the dustcap has to be strong enough for typical bash it DIY people to not wreck the speaker in 5 mins.

Its a balance as a driver designer (I have done it) and unless you are happy with a relatively minor change, you end up changing two or three items to complete the modification and maintain an approximate balance to suit the masses.

I have not even mentioned top plate, back plate or thicknesses of coil and top plate yet (plus materials) - thats far more indepth and worthy of a different thread if people want to know.....

 

[Petebass]

I have not even mentioned top plate, back plate or thicknesses of coil and top plate yet (plus materials) - thats far more indepth and worthy of a different thread if people want to know.....

Very interesting post! I'd love to hear more and I'm sure others would too.

But before people start getting grand ideas about ordering a custom driver, my understanding is there is a minimum order of 50 speakers. Is this correct? And if so can you split that into say 25 x 8 ohm and 25 x 4 ohm drivers?

 

[billfitzmaurice]

I have not even mentioned top plate, back plate or thicknesses of coil and top plate yet (plus materials) - thats far more indepth and worthy of a different thread if people want to know.....

The difference between the 3012HO and LF is elegantly simple: the LF has an extended rear cup, allowing a longer voice coil for longer excursion. Otherwise the frames/motors are the same, and all that's swapped about to get various results are the moving parts, ie., cone, dome, suspension and voice coil. With only 5 different cones, domes and coils alone that's 125 permutations. I expect Eminence could come up with a thousand variations if they wanted to.

 

[DukeLeJeune]

Very interesting post! I'd love to hear more and I'm sure others would too.

But before people start getting grand ideas about ordering a custom driver, my understanding is there is a minimum order of 50 speakers. Is this correct? And if so can you split that into say 25 x 8 ohm and 25 x 4 ohm drivers?

I don't think so; 50 identical units was the minimum...

The last official word I got on minimums for neodymium drivers from Eminence was 250 units (prepaid!) that all use the same magnet, which can be further sub-divided into minimum subsets of 50 units that all use the same "soft" parts (cone+surround+voice coil assembly). If you do the arithmetic, 250 units x double the old price = a lot of money.

 

[bobcruz]

Oy! I'd be looking very hard at other manufacturers. Someone (rpsands?) suggested that there's a Beyma neo that specs between the 3012LF and HO, and it's cheaper too.

 

[Mike Arnopol]

That was me. I was just modeling and the Beyma looks very interesting. Really does model in between the ho and lf. And the mids extend much higher and without the peaks if their specs are to be believed. And $45 cheaper than the LF at US Speaker. It's a bit lower in power handling, but I don't take complete stock in these figures.

 

[Robby Hoinsky]

Hey Rob,

Thanks for bringing all of this great detail to your build! I love the tests and documentation.

I saved all your pdf's and I'll look through them. If you have any others to share, I'd love to see them as well.

A couple of questions. Looking at your accelerometer measurements, you say that you did a a sweep, and then you show the panel resonances at 95 Hz and 130 Hz. Can you explain how that test was done, if you don't mind. Do you have any time based measurements, like an impulse response or waterfall plot? That would be more helpful to understand the damping of the panels.

Not sure if you mentioned it, but how many layers of what weight glass did you use? Any idea of the resin/glass ratio?

I see in your cost list you have damping listed, but no item shown. Did you use any damping materials?

About the baffle, after trying every imaginable way to attach powerful vibrating machines (i.e. drivers) to a foam core composite, I settled on routing a plywood ring into the back of the foam. It is very easy to do with a circle jig and a router, but could be free handed or even just surface mounted on the back. I'm also using 2" Baffles and decoupling it from the cab as much as possible to reduce mechanical vibrations.

Thanks again for sharing your journey! A composite build is very challenging and every seemingly simple step needs to be thought through completely and reimagined to get it right. Well worth it when it is done.

Robby

 

[Robby Hoinsky]

Any answer to the above?

 

[LowBstring]

depends on the weight of the cloth, the materials, etc.

I like to use light enough weight cloth that at least two layers are needed. It makes better corners.

Also, I'd vac bag anything like this.

 

[robthebassplaye]

Hey Rob,

Thanks for bringing all of this great detail to your build! I love the tests and documentation.

I saved all your pdf's and I'll look through them. If you have any others to share, I'd love to see them as well.

A couple of questions. Looking at your accelerometer measurements, you say that you did a a sweep, and then you show the panel resonances at 95 Hz and 130 Hz. Can you explain how that test was done, if you don't mind. Do you have any time based measurements, like an impulse response or waterfall plot? That would be more helpful to understand the damping of the panels.

Not sure if you mentioned it, but how many layers of what weight glass did you use? Any idea of the resin/glass ratio?

I see in your cost list you have damping listed, but no item shown. Did you use any damping materials?

About the baffle, after trying every imaginable way to attach powerful vibrating machines (i.e. drivers) to a foam core composite, I settled on routing a plywood ring into the back of the foam. It is very easy to do with a circle jig and a router, but could be free handed or even just surface mounted on the back. I'm also using 2" Baffles and decoupling it from the cab as much as possible to reduce mechanical vibrations.

Thanks again for sharing your journey! A composite build is very challenging and every seemingly simple step needs to be thought through completely and reimagined to get it right. Well worth it when it is done.

Robby

Hi - Sorry for being out of pocket for so long - I was in the Middle East working, just got back...

You are, of course correct, an impulse reasponse would be much better at showing the actual response of the panels. The problem is, I don't know how to impart that into the panel. I was testing various panels attached to an enclosure with a driver in it, and using that to excite the panel. An electrical impulse, presented to the driver, by the time it got to the panel, would really just be a pile of mush (to use a technical term...). I actually tried doing that by dropping a weight onto the panels tied with a string, creating an arc, and that should have imparted an impulse to the panel, but I could not make it reproducable (ie that is I got different results every time I did it), so I didn't put that on line (but I have the data)....

I used 2 layers of glass everywhere, and probably a lot more resin than I should have. I didn't do this by weight, but I would say that I used about 10oz per panel surface (so that means, each inside, outside, top, bottom, etc... flat surface). Nothing scientific about it, basically that is just what it took to saturate the glass.

I did use damping material, I had some acoustic 2" foam laying around from another project, and that is what I used to line the inside (back, sides, top, I left the bottom untreated, and that is where I mounted the xover).

Your comments about the mechanical attachment of the baffle should create a good conversation. Mechanically isolating that baffle (and the associated motion of that baffle due to the drivers and the internal pressure) is not a trivial analysis. Fundamentally (and in theory), it is very hard to isolate that baffle from the rest of the enclosure. That is the standard tradeoff between a shock mount and a hard mount. A shock mount done incorrectly will be a disaster. I chose to hard mount the baffle, so it is an integral part of the rest of the enclusure from a structural perspective.

Your approach of using plywood behind the composite is a really good idea, I think I would incorporate that. But, I don't think I would allow any compliance between the baffle and the rest of the cabinet, it just has the potential of creating too many phase, noise, and other problems.

Anyway, This is a great discussion, thanks for asking those quesions, and I certainly look forward to any folow on conversations about this.

 

[robthebassplaye]

Rob, I wonder if a major side-effect of your high-fidelity design is that it will force you to improve your playing technique...many commercial cabs cover a multitude of sins!

HaHa! I apologize for missing your post before, but you are absolutey correct, what you say has proven true for me. If I'm playing in my Peavey 18"x2x10" cab, I can slop around a lot. Or even in my GK 2x12, or RedHead. But in this cab, I hear every single thing that is going on, so a little pull on, or damping here or there by accident is totally audible.

Good call! (Not that I'm getting any better at playing yet).

 

[kraigo]

Also, because of the material, there is a lot less phase distortion.

<snip>

Rob

I'm rereading this thread. Can you tell me what you mean by "less phase distortion" and why that is? I wouldn't expect cab material to affect phase distortion.

Thanks Rob. I'm still learning about this stuff after years of being an enthusiast.

KO

 

[LowBstring]

I'm not sure what he means, but imo if the cabinet parts are vibrating, the sound from the speakers are passing over this vibrating part and "the pure, clean sound of the speaker" is distorted by bouncing off the vibrating part. The vibrations impart additional layers of sound that are out of phase with the speaker sound, causing all kinds of muddy interference.

You want your sides and baffles to be very rigid so sound passes over them, or bounces off them, but does not excite them.

 

[kraigo]

In my opinion, as long as the rear port is at least a quarter of a wavelength away from the rear wall, there should not be any noticable effects.

This is the second time you've posted this but it doesn't tell me a quarter of a wavelength at what frequency? First port resonance frequency makes about the most sense. Box tuning would mean "simply leave your cab about 10 feet from the wall..."

 

[kraigo]

I just thought that the panel most likely to vibrate would be the one with the most mass cut out of it i.e the front baffle. I thought that bracing that panel would be as good a place to start as any, and as it turns out I got lucky didn't have to go any further before the problem disappeared.

Did you just run a foam/fiberglass piece from the center of the baffle to the front of the cab? When you say "braced back to front", how? Curious and all.

KO

 

[Petebass]

Did you just run a foam/fiberglass piece from the center of the baffle to the front of the cab? When you say "braced back to front", how? Curious and all.

KO

Yep that's pretty much what I did. Here's a pic of the same bracing as used on a 2x12 I built recently.

 

[robthebassplaye]

I'm rereading this thread. Can you tell me what you mean by "less phase distortion" and why that is? I wouldn't expect cab material to affect phase distortion.

Thanks Rob. I'm still learning about this stuff after years of being an enthusiast.

KO

Phase distortion can mean a lot of different things to different people, but what I mean is that the response will not be linear with frequency, so that the phase shift is not constant over the frequencies of interest. Thats my definition, perhaps not everyone else has the same definition, but it's the effect I want to describe.

It's a second order effect... In other words, the way I have defined it, there are different "phases" of a sine wave, which align differently at different frequencies. When those sum up (in a standard plywood cabinet, which will resonate and have phase differences), that is what I am calling phase distortion.

In this build, what I was referring to is that the contribution of the cabinet (due to the extreme stiffness of the walls) is minimal, therefore, there is not a lot of different phasing (of any significant magnitude) going on, and that, in my opinion, really reduces the phase distortion.

At least, thats my take on it...

 

[robthebassplaye]

This is the second time you've posted this but it doesn't tell me a quarter of a wavelength at what frequency? First port resonance frequency makes about the most sense. Box tuning would mean "simply leave your cab about 10 feet from the wall..."

Yes, in theory, what I have read is that as long as it's less than a quarter wavelength away from a reflective surface, there are no significant artifacts. And that frequency is the first port resonance.

So, let's assume it's 40Hz, then the wavelength is about 28 ft. A quarter wavelength then is 7 ft.

I have to say that I have not been able to hear this effect at all, I have experimented with having my cabinet close to and far away from a rear wall, and I cant tell if there is something different going on.

When I was building the cabinet, I also read some papers that stated that the front and rear ports had to be within a quarter wavelength of each other (which this cabinet is).

I wish I knew how to measure any effect, perhaps that would help.

But, my bottom line is that a cabinet with a front and rear port that are as close as these 2 are to each other works just fine, and can sound good.

 

[5StringPocket]

Nice design and very thorough analysis in this thread. There's something satisfying about following a project from inception to completion and having it turn out well. I've followed composite builds by Petebass, Zac, and Robby Hoinsky as well and would like to try a build of my own shown below. I'll start my own thread on this so as not to derail but wondered if I might bring up a few construction ideas and questions?

1) The Owens Corning XPS Foamular seems to be the DIY core of choice for composite builds. I'd like to use 1.5" for the baffle and 1" for the rest but Home Depot seems only to have 1/2" or 3/4" around my area. (I'd really rather not have to laminate to reach the right thickness.) Does a router work with XPS board to round edges, cut circles, etc?

2) From my reading, the West Systems Epoxy with 8.95 oz Style 7781 satin weave E-glass fabric from Link Removed might be a good choice for epoxy and fabric. I also got that roughing up the board with 60 grit paper before bonding and running a fillet of epoxy along the joints is recommended because the heavier fiberglass cloth won't get all the way down into tight corners. I've had good luck with roll on Duratex finishes and would like to use it for this cab as well. How would you prep epoxied fiberglass for a Duratex finish (bondo and sand, gel coat)?

3) Plywood reinforcement rings with tee nuts on the inner baffle and sides for mounting speakers and handles also make sense for supporting load-bearing fasteners. With a 1.5" core, 2 layers of 8.95 oz glass cloth inside and out, and a 1/4" thick figure 8 plywood backing for speaker mounting reinforcement would a front to back brace still be needed with a pair of 3012HO drivers?
Thanks.

 

[billfitzmaurice]

I have experimented with having my cabinet close to and far away from a rear wall, and I cant tell if there is something different going on.

You should. When a radiating plane is 1/4 wavelength from a rear wall the reflected wave meets the original wave 180 degrees out of phase and a cancellation notch results, as much as 24dB deep. And that applies not just to port radiation but to cone radiation as well below the baffle step, where the polar pattern is omni-directional. This is called Allison Effect, and it pretty much demands that neither the port nor the baffle be in the zone from 2.5 feet to 7 feet from a rear wall.

When I was building the cabinet, I also read some papers that stated that the front and rear ports had to be within a quarter wavelength of each other (which this cabinet is).

It would take a gigantic cab not to be within the port pass band.

In this build, what I was referring to is that the contribution of the cabinet (due to the extreme stiffness of the walls) is minimal, therefore, there is not a lot of different phasing (of any significant magnitude) going on, and that, in my opinion, really reduces the phase distortion

That's true of any cab that's properly braced. IMO those which aren't are defective. Unfortunately many are just that.