Natalie P towers, with a twist

[mrstampe]

After a lot of reading and advice from the forums, I’ve decided to start out my 5.1 DIY surround with a set of Natalie P floor standing left and right channels. Eventually, these may move to the rear if I decide to build Dayton TMWWs for L/R and a Dayton WTMW center. Also, an SDX15 sub is in the works.

I plan to build the ported 50L version of the Nat Ps, but I’d like to put a twist on the design… or maybe a bend. I’m reasonably experienced in woodworking, though this would be my first try at panel bending. This can be accomplished by form pressing thin laminates into thicker panels, by saw-scoring MDF or Baltic Birch, by stacking “polylam” walls, or buying manufactured bendable ply. I'm not married to these plans, so I am open to creative ideas other than the standard rectangle.


See a few sketchup ideas at the end.


Bending techniques…

form bending



scored



bendable ply



translam

Image not available



Questions:

1) The Swerve preserves the intended Nat P baffle dimensions, but the Vibe and the Wave are instead widen to 10” on average with curved edges. Will this affect the sonic qualities of the speaker?
2) Does anyone have experience with these bending techniques? Any advice?
3) Some building the 50L towers seem to prefer a higher tune (6” long, 3” diameter port = ? tune) to the designers recommendation of an 11” long, 3” diameter port for a 28Hz tune. Any thoughts on that?
4) Which design do you prefer?

[JonP]

After a lot of reading and advice from the forums, I’ve decided to start out my 5.1 DIY surround with a set of Natalie P floor standing left and right channels. Eventually, these may move to the rear if I decide to build Dayton TMWWs for L/R and a Dayton WTMW center. Also, an SDX15 sub is in the works.

I plan to build the ported 50L version of the Nat Ps, but I’d like to put a twist on the design… or maybe a bend. I’m reasonably experienced in woodworking, though this would be my first try at panel bending. This can be accomplished by form pressing thin laminates into thicker panels, by saw-scoring MDF or Baltic Birch, by stacking “polylam” walls, or buying manufactured bendable ply. I'm not married to these plans, so I am open to creative ideas other than the standard rectangle.


See a few sketchup ideas at the end.


Bending techniques…

form bending



scored



bendable ply



translam

Image not available



Questions:

1) The Swerve preserves the intended Nat P baffle dimensions, but the Vibe and the Wave are instead widen to 10” on average with curved edges. Will this affect the sonic qualities of the speaker?
2) Does anyone have experience with these bending techniques? Any advice?
3) Some building the 50L towers seem to prefer a higher tune (6” long, 3” diameter port = ? tune) to the designers recommendation of an 11” long, 3” diameter port for a 28Hz tune. Any thoughts on that?
4) Which design do you prefer?

There will be an influence on the design any time the box volume, (not being changed here) tuning if vented, or the dimensions of the front baffle are changed significantly. The shorter distance is more sensitive, especially in a tall tower. One inch is more of a change proportionately to the width, where one inch on a 40" height is less of an influence. I think the wavy styles can be the same effective width, if the "waves" are "in phase" with each other, where if they aren't , the effective width will change. Hard to say exactly what either case will do, though the same width probably is equivalent to a straight sided box.

Another thing, the tweeter offset could be changed by where a wave is... You'd want to make them mirror images of each other so the tweeters are placed the same, and will have the same diffraction/dispersion pattern, or they might sound noticeably different.

I think they're all pretty cool ideas... think I like the more gentle, full length waves the most... I they'd look a little understated rather than overstated in real life. Probably easier to build, too...

Now if you, and your decor, are into overstated... well, that's a different story! Anyway, it's good to see something different from a plain box... :T

[Gir]

I think the wavy styles can be the same effective width, if the "waves" are "in phase" with each other, where if they aren't , the effective width will change. Hard to say exactly what either case will do, though the same width probably is equivilant to a straight sided box.

I think this will change the baffle step, just like offsetting the tweeter does. I'd be a little careful going crazy with the waves, but I must say, that last design does look pretty sweet!

In about a month I'll be starting my CNC mill (working on designing it right now), and once that's done I'll be pumping out some crazy box ideas, such as engravings along the front/sides, translam, and other crazy shapes that are difficult to obtain.

I look forward to seeing how yours turns out, I think it will be very cool!

[dlneubec]

Interesting! I like the Wave the most.

How about turning the wave 90º and adapting it so the woofers are on the top of the curves and the tweeter is offset in the valley between?

The cutouts would be a challenge, but they would look interesting and there would be a functional reason for the curves. The baffle width could match the original NatP, but of course the tweeter offset would have to be compensated for in the crossover.

Probably too much work!

[Operandi]

Cool idea, though it looks like an insane amount of work, and probable headaches. Personally the Wave is only one that I care for visually.

The cutouts would be a challenge, but they would look interesting and there would be a functional reason for the curves. The baffle width could match the original NatP, but of course the tweeter offset would have to be compensated for in the crossover.

Form as well as function, neat idea. :T

[Steve Manning]

I like the wave idea you have as far as looks, can’t speak to the acoustics of it. I can speak of the translam method though, it’s lots and lots of work, I’m in the middle of building a pair, myself. As a suggestion though since I like how they look, you could try laminating the layers vertically rather than going horizontal. The Nuforce S-9 is built that way, http://www.nuforce.com/Product-Speaker.htm. I think it could be easier than going the horizontal route if you went the for the wavy look.

Steve

[PoorboyMike]

I don't think the wavy look would be too difficult. If you have access to a band saw, you could cut out the shape on a thick piece of wood. Keep both halves, and sand them until they fit tight. This would then be your form that you would use to press the thin layers into shape.

I talked to friend of mine who builds bows and fancy, curved furniture, and he uses a fast drying polyurethane glue between the layers. He doesn't steam the laminate either, just gets in nice and wet with warm water.

[tpremo55]

A few things to consider.

First, if you use the kerf-cut method to bend MDF, you will need to alternate the side in which the kerfs are to create the wave as you have designed. While this is absolutely do-able, you will need extra treatment for the concave area where the kerfs will face out.

Second, again if you use MDF kerf-cutting method, you will need to experiment with the frequency of kerfs to get a smooth curve. While simple bend equations can give you the stacked surface length of the inner curve, this will not typically result in a smooth curve on the solid side of the board and you will have a polygon-type surface (i personally don't like the polygon).

third, and again with MDF kerf method, water is actually another key tip in a smooth surface. The same technique with a wet washcloth wiped along the solid side just prior to forming the curve will turn out much smoother than the same technique with the dry board. Again, the result is a tendency to crack or become polygon-like at the surface.

After looking at all the options for a recent project with curved sides, I did decide to go with the MDF kerf bending technique. You can see more about this design here:

Images not available

Part of this design required the use of a contoured press to set the curve. MDF bent like this retains too much energy for the integrity of the wood. I further filled all the kerfs with a glue/MDF sawdust mixture to both return the density to near original and to set the curve.

For a smaller surface, I would probably do a stacked laminate. I recognize many benefits to that process. Even if you don't have access to a CNC, you can get close to the same results with a router with 1/4 spiral bit and a collar insert set and a profile template out of hardboard.

[orbifold]

For a smaller surface, I would probably do a stacked laminate. I recognize many benefits to that process. Even if you don't have access to a CNC, you can get close to the same results with a router with 1/4 spiral bit and a collar insert set and a profile template out of hardboard.

As one who used this, (maybe the "another guy," dunno, but mine was MDF) I will say yes, it is work, but I loved doing it and will use it on the center, a Dayton WTMW, and on a 7cuft Tempest SW after that. FWIW, I found that a 1/4" spiral upcut would cut the 3/4" thru in one swat. But this next time I will try a 8mm (about 5/16") Whitside bit with a 1/2" shank, and a larger collar for more clearance. I just say this because you might be able to adapt the method well to that wave if you just make vertical, rather than horizontal laminations, ala Steve Manning's sugestion. No matter what, you're going to have to laminate something to get a curve, and if you look at that picasaweb album, that looks like more work than a translam, to me.
This next time, I will glue 3 or 4 pcs at a time, and nail each one as it is placed before clamp pressure is applied, using a square or a square form. Without nailing first, the pieces move as you clamp, and you have to rout off the whole outside surface. With a few judiciously placed nails, I should only need to sand smooth. Big difference!
BTW, I like the wave :T

[---k---]

It looks very cool. I'm sure it will have some impact on the sound, likely minimal, but if you cool with that; :T

If you did stack laminate, I would be tempted just to cut squares, ofset them, then go back and fill the stagger with bondo and shape. Easier to cut, and think of how solid of a box it would be.

[mrstampe]

Nat P Wave

I think I'm leaning toward the Nat P Wave and possibly done in verticle polylam layout. In that case, the finish would be polyurethane over BB verticle layers on the wavy sides and a nice complementary veneer on the baffle. I just wonder if all of those verticle wavy lines would accentuate the curves or be too distracting? If I'm not going to see the polylam layers, I may opt for one of the other methods.

Rough dimensions would be as follows: H40" x W10" x D12", curve wavelength 20" and amplitude 1/2" --> If the sidewall were a soundwave, that corresponds to a audio frequency of 675Hz! Also an internal volume of 56L before driver, bracing and port subtraction.

Also, a higher tune or even a sealed enclosure for the mains may be easier to mate with that SDX15 sub in the planning stages. Any suggestions as to how best to tune and cross mains to the sub?

Thanks for all of the advice!

[1Michael]

I built mine with the 3" Dia, 11" length port and it sounds quite good. I to am going to use this as a rear channel, but right now they are my mains.

[mrstampe]

Getting started!

The Natlie P Wave build begins!!

To summarize the final build plan:
Design: Natalie P MTM
Cabinet: floor standing 50 liter (10" x 12" x 40" exterior)
Waves: wavy sidewalls formed with bendable ply
Port tuning: to be determined
Prefinish: Bondo seams and open grain, sand and seal
Finish: piano black lacquer (if I can figure it out)


Step 1: transfer waves to template
I laid out verticle lines for front baffle and rear wall on 1/4" hardboard as follows: midline left to midline right margins 8 1/2", as well as 1/2" deviation line to left and 0.5" deviation to right for both walls. Then at 5" intervals, I marked out points along the wave at left dev line --> midline --> right dev line --> midline, and so on. Finally, I used an adjustable ship curve tool ($20 at Rockler.com) to trace the curves based on the layout points. (this will all become clearer in the Step 3 picture)




Step 2: cut out template
Since my bandsaw conveniently fried just before this project began, I resorted to my jigsaw -- very slow going and not as accurate, but it gave me a decent rough result. I then sanded to my lines using a drillpress drum-sanding kit ($12 at Rockler.com).




Step 3: transfer template to 3/4" MDF, cut
I applied my template to MDF board with double-sided tape, traced the outline, and cut out by jigsaw. Since I already had a relatively smooth template curve, there was no need for sanding. I just used a bearing-guided flush router bit on a router table. The bearing moves smoothly along the template, while the flush bit duplicates the curve in the MDF baffles.

[Wildcard26]

My these are impressive so far and they are not even mounted as yet.

[Gir]

Excellent work so far. I can't wait to see how these come out. Nice pictures too!

[mrstampe]

Step 4: driver cutouts

Step 4: baffle driver cut-outs
I laid out the driver midpoints on each baffle based on the original designer's plans on HTguide.com:




Then using my handheld router with a Jasper circle jig, a pin placed in the desired radius jig hole and also into the midpoint of each driver location. Mounting plate recesses are routed first. Then, the cutouts are are routed through the baffles. The appropriate router bit for this operation is called a spiral upcut -- best source: MLCS.




Here's how they turned out...

[---k---]

Impressive. :T

[TacoD]

Wow, you do not like the easy way (straight cabinet)

[JonW]

mrstampe,

These look great. :T I can't help but want to call them the "Dr. Seuss speakers."

[mrstampe]

jumping ahead

Okay, I'm going to jump ahead a few steps, then come back to explain the details later...

I've wired the crossover boards and assembled one of the towers, except one of the wavy sides. I have to keep it open to stuff insulation, install the x-overs and to wire everything up. I've learned a lot from the first one, so the second one will go together much easier. That will be the one I chronicle step-by-step with pics.

Any thoughts or suggestions? Will I need post-assembly access to the x-over compartment? I'm open to any advice.

[sprint_9]

Very nice work :T