Boson Revised, TL Version

[JasonP]

Well, after some serious modeling in Martin's MathCAD, I've got a design for the Boson I am happy with. While I am straying from the total open baffle speaker, I can't face the low-end issue anymore and simply backed off :o I've also settled on a controlled dispersion, since I know for a fact this design will be placed near the side wall (18-24" from them in fact) and I hope that controlling (limiting) the off-axis response will lower the impact of early reflections and imaging.

I still have a couple DA270-8 aluminum 10" woofers, and modeling them in MathCAD with a 60" transmission line has given me this, in-room, with baffle compensation, at 50 watts. At this level it is still within xmax all the way down to 20hz. As I pump it up past here, it starts to get a little wild below 30 hz with excursion, as expected with the amount of output it is generating.



The top end will be handled by an open-baffle Peerless 6.5" HDS with a GFC cone, the 830990 plus one yet unspecified tweeter. Likely the EOS6 horn with DNA-150 compression driver.

[JasonP]

Re: Boson Revised, TL Version

Here is the box design in sketchup, not quite the final proportions but overall the same, without the side panels attached to show the interior.

[JasonP]

Re: Boson Revised, TL Version

I started on construction today, using: 3/4" quality birch ply for the rear, top, and back panels - 3/4" solid edge glued pine for the side panels - one seriously awesome solid piece of Purpleheart at 11 x 50 (will be shortened) for the front baffle.



The exterior was glued up with Titebond II and secured with pocket holes via the Kreg Jig (man that thing is nice). The front baffle is going to be removable, but I'm unsure how I am going to handle securing it yet.

Another big question is how I am going to finish the side panels. I'm pretty sure the birch is going to go dark, really dark but I have no idea what to do with the sides yet to make it blend with the purpleheart baffle.

[Paul K.]

Re: Boson Revised, TL Version

Your modeled responses look very good but the cabinet drawing is confusing (to me). It appears that either there are two woofers in the same line, one near the beginning of the line and the other at the opposite end, or is the large, circular hole in the back the line's terminus/port, or will there be passive radiator there? Sorry to seem dense.
Paul

[JasonP]

Re: Boson Revised, TL Version

Thanks for looking this over Paul, I appreciate it! To answer your question, the rear opening is the terminus of the line. I made it round because since I've got the circle jig I've been in perfect circle cutting heaven! Also, it seemed an easy way to roughly match it to the SD of the woofer. Also an easy way to cover the hole with a grill & speaker cloth to prevent anything getting in there.

I didn't mention all the details above, but its a tapered line, roughly 2 to 1. That modeled with more bass response than the straight line did. There is also some serious stuffing in the model, like .7 lbs per cubic foot.

[Paul K.]

Re: Boson Revised, TL Version

Thanks for the explanation. What is the modeled system tuning frequency, which should be pretty much the 1/4-wave resonant frequency of the tapered line, just out of curiosity? BTW, I normally use a stuffing density of 0.75 lb/ft3, and up to 1.0 for most all TLs I design, so I wouldn't call 0.7 as particularly serious, except that with a taper of only 2:1, you're not getting much reduction of the magnitude of the response wiggles.
Paul

Thanks for looking this over Paul, I appreciate it! To answer your question, the rear opening is the terminus of the line. I made it round because since I've got the circle jig I've been in perfect circle cutting heaven! Also, it seemed an easy way to roughly match it to the SD of the woofer. Also an easy way to cover the hole with a grill & speaker cloth to prevent anything getting in there.

I didn't mention all the details above, but its a tapered line, roughly 2 to 1. That modeled with more bass response than the straight line did. There is also some serious stuffing in the model, like .7 lbs per cubic foot.

[Psycoacoustics]

Re: Boson Revised, TL Version

Hi Jason: Have you ever worked with Purpleheart before? The reason I ask is because I have. Just on a few small things. Certainly nothing that big. Anyway, I've found that when it's in it's raw state it is very "Purple"! However, the more you sand it to get it smooth, the more the purple color goes away. I've given away all the things I've built with it, so I don't have anything on hand to look at, but I do remember being disappointed after sanding to like 220, and than finishing with Watco Danish oil. That bright purple gets really muted!

I would certainly suggest testing you finishing procedure to make sure you'll like the final product. May I ask what that hunk of wood cost? Good luck, Mark

[JasonP]

Re: Boson Revised, TL Version

Paul, I noticed Martin's sheets went up to 1 lb. per cubic foot, so I figured I was still safe. Though when the sheet told me I was going to use 2.4 lbs per line I boggled at the amount I assume from my vented box experience that the point of highest acoustic impedance is the resonance of the line, which look at the mathcad sheet puts it right above 25 hz, maybe 27 or so.

I didn't model tapers in excess of 2 to 1, since even that seemed hard to pull off in a folded line like this, maybe that was silly? I mean, as it is, I have to add a second piece to the box inside to make sure the line isn't effectively "growing" right after the turn.

[JasonP]

Re: Boson Revised, TL Version

Hi Jason: Have you ever worked with Purpleheart before? The reason I ask is because I have. Just on a few small things. Certainly nothing that big. Anyway, I've found that when it's in it's raw state it is very "Purple"! However, the more you sand it to get it smooth, the more the purple color goes away. I've given away all the things I've built with it, so I don't have anything on hand to look at, but I do remember being disappointed after sanding to like 220, and than finishing with Watco Danish oil. That bright purple gets really muted!

I would certainly suggest testing you finishing procedure to make sure you'll like the final product. May I ask what that hunk of wood cost? Good luck, Mark

Thanks for the advice Mark! It is indeed the first time I've worked with it. I did some reading beforehand though, since I know its some stuff you want to avoid inhaling or getting splinters all over yourself with. I hadn't heard about the finish dulling like that after sanding, but I do know it'll turn brown as it is exposed to UV light and finally settle into some purple-brownish color.

I talked a bit about the piece of wood I nabbed at Woodcraft in the original thread. It was a whopping 13" by 9 foot board, 4/4, before the first cut. Not its in two pieces, one 50" long and trimmed to 11" wide, and the other unworked so far. Its almost perfectly flat and true along that whole length, except for some minor scuffing/curving at one end, where it looks like it is an effect from it being the end cut of the large piece of wood they worked. It wasn't cheap, the big board was 110$ out the door. I look at it as a long-term investment, given how nice the grain is on it and how uncommon it is to find so perfect a large board.

[Psycoacoustics]

Re: Boson Revised, TL Version

From what I know,,, you got a good price! Pat yourself on the back! I've never seen a piece that large. It's funny, from the pic I would have thought it was a 5/4 board. Anyway, like I said, test your finishing procedure first! My guess though is you'll have a really unique baffle, that'll look beautiful! Again, good luck and keep us informed. Mark

[Psycoacoustics]

Re: Boson Revised, TL Version

Hello again: I did find a piece that I sanded and test finished. Your right, it does turn a little brown w/ purple highlights. I don't know if these pics will help, but here ya go. One is one side of the board w/test finish, (Watco, and Briwax) the other side is just sanded. Mark

[Paul K.]

Re: Boson Revised, TL Version

Well, the larger the taper, the more the ripples in the response are reduced, and the shorter the line needs to be for a given 1/4-wave resonance. Also, the larger the taper, the less stuffing is needed, both in density and length. When I use tapered lines, I start with a taper of 10:1, and I've built some folded lines with a taper of 25:1 that worked quite well. Of course, you have to keep an eye on the velocity in the terminus so it doesn't get so high as to cause "port" noise.
Paul

Paul, I noticed Martin's sheets went up to 1 lb. per cubic foot, so I figured I was still safe. Though when the sheet told me I was going to use 2.4 lbs per line I boggled at the amount I assume from my vented box experience that the point of highest acoustic impedance is the resonance of the line, which look at the mathcad sheet puts it right above 25 hz, maybe 27 or so.

I didn't model tapers in excess of 2 to 1, since even that seemed hard to pull off in a folded line like this, maybe that was silly? I mean, as it is, I have to add a second piece to the box inside to make sure the line isn't effectively "growing" right after the turn.

[JasonP]

Re: Boson Revised, TL Version

Took a free-air measurement of the GFC cone peerless last night with Holm, placing the mic about 3-4 inches from the cone dustcap, about as loud as my mic could handle.



Distortion looks nice, and if it wasn't for that peak at about 2.6 khz of 3rd order distortion (red is 3rd, green is 2nd, all tall orders were below -70) it could operate to 3.5 khz likely without issue. I was impressed at how muted the breakup was in this measurement, the peerless specs show it to be 6-8 db above the normal operating level. I wonder if that is just an artifact of measuring this way, dunno. If it isn't, the breakup here isn't hard to handle at all.

Overall nice driver. Its the first I've listened to free-air and felt it sounded pleasant, though that could just be the overall smoothness you can see in the graph.

I'm not sure how the cone is constructed, but it looks like woven fiberglass backed by uncoated paper. If that is it, it really seems to work in this case. I've never seen the Scanspeak glass fiber cones, so I wonder if there similar (with the paper backing).

[JasonP]

Re: Boson Revised, TL Version

I think I finally settled on a tweeter for the Boson: A Vifa OT19 faceless tweeter in a waveguide of my own construction. I fired up MathCAD to model a curve I've been looking at for a while now, which implicitly one would think to have good qualities in a waveguide - the Agnesi curve (or Witch of Agnesi as it is sometimes known). I rendered it in MathCAD and then applied a squish to it based on the Golden ratio. Finally I have this:



It will be about 1" deep total, and 4" in diameter, the increments are inches. I placed the straight red line where the 1 and 1/8" cut out will be for the OT19. Next I will see if I can scrollsaw out a thin piece of wood, maybe 1/4 or 3/8, with this shape. I'll place some polymer clay on a box, and use this wood form to press out the shape - press and turn, repeat until it looks ok and smooth as needed by hand. I think I'll get a workable waveguide from this to test.

[JasonP]

Re: Boson Revised, TL Version

I picked up two polymer clays today, Premo Black Polymer and Polyform Model Air. The latter was easy to work like playdoh, so I modeled out a very (VERY!) rough version of the waveguide like so after a few failed attempts.



Yea, I said real rough. Its not meant to be for any application beyond seeing how workable the material is after it air dries in 24 or so hours. I want to see if I can drill out the 1 1/8" hole in the center, trim the edges and so forth with tools or not. I did try this about five times using different methods, and finally to get one that worked:

I rolled the clay to a reasonable thickness (say a little less than 1/4") with a rolling pin,
place that over a 4+ inch small bowl,
pressed the center down by hand to loosen it up a bit,
and finally used my scroll saw cut curve profile to press it into the right form.

After that, and I realized it would be hard to make it perfectly smooth this way, I started drawing on the surface with a tiny clay tool rake. Hand drawn waveguides here! Get your hand drawn waveguides here!

I hope this material is good, because it is only 6$ for 2 lbs, and I used one-quarter of that for this guide. What would be a materials cost of like 1.50$ Nearly off the dollar menu. Do you want fries with that? !