Introducing the SoundRounds

[dlneubec]

This is my third hybrid omnidirectional project, intended to push the limits of that concept. The hybrid concept is one in which the low and mid frequency drivers are omnidirectional to a point where they cross to the tweeter, where the tweeter is front and the radiation pattern transitions from 4pi into 2pi space. I believe this gives the best of both worlds, with an expansive soundstage, a very open midrange, yet good spatial imaging because of the front firing element. In this case I wanted to push the omnidirectional output to cover as much of the frequency range as possible before making the transition. The 3way, all passive crossover, is not yet final, but the crossover points are going to be about 365hz and 3500hz, which I believe is about as high as a dual cone midrange driver (up and down firing) design can be taken in this kind of design, due to driver baffle interactions compromising frequency response.. The other goal was to be able to get this design into the “budget” category at the IowaDIY, which has been under $300 for drivers and crossover parts. This barely makes it and it will take some electrolytics on the woofer shunt to get there.

The drivers include one Dayton RS225-8, two of the HiVi B3A-B buyouts (the ill fated RS90, I believe), and one Aura NT1 neo tweeter. I originally was using the Dayton ND20-FA, but was having some distortion issues at this crossover point and decided to try the Aura, which I like considerably better. This necessitated the need for a small baffle for the tweeter, where there was none with the ND20-FA, which you may have seen in a previous photo I posted.

The Crossover is 2nd order electrical on the woofer low pass and mid high pass, approximating a 3rd order Bessel acoustic slope. The -f3 points are spread so the drivers sum flat at the crossover point. The woofer is wired in reverse polarity. The mid low pass and tweeter high pass are 3rd order electrical, approximating a 4th order or steeper LR acoustic slope. I’ve done some different things in the midrange net to raise the impedance, which dips to as low as 3.5ohms, but is about 6ohms nominal overall. I have two versions of the crossover I will be deciding between. One has a typical Lpad before the tweeter and the other, posted here, has a series resistor for the tweeter net, on the amp side. The drivers are time aligned vertically for a listening position of 38” ear height at 3 meters. The phase match is optimized for that location.

I have just one prototype speaker built, but I’m very happy, really kinda shocked, with how very much I like it. The midrange is incredibly open and natural sounding and the overall balance is right on the money, IMO. We’ll see if that is still the case when the pair is completed.

Parts Express really needs to go ahead with offering the RS90 driver, IMO. It is a really nice, clean midrange, with a pretty wide band capability and small form factor.

Included is a photo of the prototype, a mock-up of the finished look, followed by the Xover and simulated FR and Z and Phase responses.

Thanks for looking! Comments are welcome.

[WillyD]

How do you like the Aura NT1? I am anxious about having them in my mains, since I've never heard them. I have some faith that they sound just fine though (good enough for me anyway). I am no "audiophile".

[dlneubec]

I like it a lot so far. I think it will be fine as long as you cross it high enough and/or with steep enough slopes. It's a pretty low distortion tweeter. Even Zaph used it in his most recent design, the ZD-3. Is that what you're building? If so, I'd say you have nothing to be concerned about.

[---k---]

Dan,

Looks great. Not sure if I can add anything, but I sure do enjoy seeing your projects.

[Dennis H]

Dan, you're a designing MACHINE, going 'where no man has gone before.' :T

[WillyD]

I like it a lot so far. I think it will be fine as long as you cross it high enough and/or with steep enough slopes. It's a pretty low distortion tweeter. Even Zaph used it in his most recent design, the ZD-3. Is that what you're building? If so, I'd say you have nothing to be concerned about.

Yep, thats what I am building. Good to hear.

And what the other guys said. Your projects are something else!

[critofur]

Cool! The last time I worked on a speaker design that got finished (the Micro Walsh at Ohm), most of the work was put into designing the cabinet, but, you'd have to cut the speaker open and look on the inside, if you really wanted to see what was done.

I'd been thinking about doing a design like that (similar to your speaker) for a long time, except with the mids in spheres. I haven't gotten around to finding/building suitable spheres though... I had also been pondering using small - medium sized sonotube (is that what you're using?) and building some kind of baffle insert. I was thinking of doing an MTM that way and wrapping the tube in veneer.

I might try something like that with a couple of my W5-704Ds, the Aurasound NT1 seems like a good choice in tweeter, you can have it perfectly time alighned, except if you move up/down the angle will change...

I was thinking about angling the speakers "out" a bit towards the listener, and maybe placing them closer together, forming a wedge shaped space between the mids with the tweeter middle similar to yours.

I've also been trying to figure out a good reflector design for an omni speaker. I saw one commercial design where they put a metal sphere above the speaker. This speaker also looks interesting, though not omni: http://www.soundandvisionmag.com/spe...5-speaker.html

I went to listen to them and was not impressed at all, however

[dlneubec]

Cool! The last time I worked on a speaker design that got finished (the Micro Walsh at Ohm), most of the work was put into designing the cabinet, but, you'd have to cut the speaker open and look on the inside, if you really wanted to see what was done.

I'd been thinking about doing a design like that (similar to your speaker) for a long time, except with the mids in spheres. I haven't gotten around to finding/building suitable spheres though... I had also been pondering using small - medium sized sonotube (is that what you're using?) and building some kind of baffle insert. I was thinking of doing an MTM that way and wrapping the tube in veneer.

I might try something like that with a couple of my W5-704Ds, the Aurasound NT1 seems like a good choice in tweeter, you can have it perfectly time alighned, except if you move up/down the angle will change...

I was thinking about angling the speakers "out" a bit towards the listener, and maybe placing them closer together, forming a wedge shaped space between the mids with the tweeter middle similar to yours.

I've also been trying to figure out a good reflector design for an omni speaker. I saw one commercial design where they put a metal sphere above the speaker. This speaker also looks interesting, though not omni: http://www.soundandvisionmag.com/spe...5-speaker.html

I went to listen to them and was not impressed at all, however

These are 6" diameter schedule 40 PVC pipe, lined with 3/4", 3layer Sonic Barrier and then stuffed with Acousta-stuf.

I've thought about angling the tweeter toward the listener as well, but so far have just toed them in where needed.

The distance between the mids is critical in terms of interaction with one another and their respective baffles and how it effects the frequency response, all of which is dependent on the crossover point. I don't recommend designing an omni project without full measurement capability. It's simply too hard to predict the FR of various combinations.

This is my 4th omni project overall and I've done a lot of experimenting with diffuser shapes, sizes, orientations, etc. and IMO, the only ones that work are the concave cones, like Duevel uses. If you search on omnidirectional here at HTGuide, you can read about what I found in 3 other projects, one with omnidirectional in the title, one called the Mentor II, and one called HOSS.

[peter_m]

Your speakers look stunning. Must make for an interesting conversation piece when people see them...

Do you have any suggestions for an inexpensive omni-directional project for people like me that are curious about the benefit of such a design? The lower the cost, the easier for someone curious to try it out and hear it for them selfs...

Regards,
Peter

PS: Without wanting to sound arrogant...According to this definition http://en.wikipedia.org/wiki/Convex_mirror, don't you mean convex?

[dlneubec]

Thanks!

I don't currently have any suggestions for a simple DIY omni, though I've been thinking about doing such a design, maybe using some of the Tangband W5-740d's and a neo tweeter. Something that would allow folks to get their feet wet in the omni design area, so to speak. If I come up with something concrete, I'll post it here. I'm not sure how much interest there would be, however.

To me, if I look at a surface it is bowed inwardly, that's a concave surface. If its bowed outwardly, then it is convex. Looking at the back side of a bowl is the convex side, the inside of the bowl is the concave side. In this case, we want a cone shape, but the cone profile is bowed inwardly not outwardly. To me, that's a concave cone.

Concave

[critofur]

What's the speaker in your "Avatar" photo? Looks kind of like a Radialstrahler or
German Physiks?

[dlneubec]

That's my last hybrid omni project, called HOSS (Hybrid Omnidirectional Speaker System). Look here:

HOSS

[Jim Holtz]

Hi Dan,

I'm a little late to the party but I did want to say that this looks like a really neat project. I really enjoyed the HOSS speakers you brought to Iowa. They were very easy on the ears and the eyes. You did a great job on them.

I'm looking forward to hearing them this fall if we can get you back to Iowa. :T

Jim

[dlneubec]

Here are some follow up results on the SoundRounds, but this time measured results. Below is the measured impedance/phase, the frequency response at 2.5meters with a 5ms gated window and horizontal off axis FR's at 0º, 15º, 30º and 45º. FR, all the way out to 45º is about +-1db out to about 4khz, widens to +-2 out to 9khz and at 20khz the level is down 12db at 45º. The results are accurate only down to about 500hz or so, due to the short gated window, so ignore everything below that.

I was hoping for a flat FR at 15º off axis, and other than the bump in the tweeter response at about 17khz, it far exceeds my expectations.

[dlneubec]

Hi Dan,

I'm a little late to the party but I did want to say that this looks like a really neat project. I really enjoyed the HOSS speakers you brought to Iowa. They were very easy on the ears and the eyes. You did a great job on them.

I'm looking forward to hearing them this fall if we can get you back to Iowa. :T

Jim

Thanks Jim, I hope to be there. I can't imagine the competition being tougher or with closer results than we had in 2007.

FYI, I'm in the process of changing the HOSS over to all passive 3way, no plate amps. They sound cleaner and clearly better integrated between the bass and the mids than they were with the plate amps, plus it knocks about $200 off the price. Now.......... what to do with the two extra 240w plate amps I have........... :yesnod:

[---k---]

Dan,
That FR looks amazing. Now I'm sitting here thinking that it is a darn shame that Iowa is such a long time away.

[JonP]

Thanks Jim, I hope to be there. I can't imagine the competition being tougher or with closer results than we had in 2007.

FYI, I'm in the process of changing the HOSS over to all passive 3way, no plate amps. They sound cleaner and clearly better integrated between the bass and the mids than they were with the plate amps, plus it knocks about $200 off the price. Now.......... what to do with the two extra 240w plate amps I have........... :yesnod:

Hmmmm.... now I'm thinking of a few things to do with another 240W...

Great results, Dan... I wish those drivers were still available. Anyone heard rumor of them coming back into production? On another thought, how close does a B3N come to one of those as a replacement?

[dlneubec]

Hi Jon,

I haven't researched it, but I bet that HiVi or Tangband have a 3" driver that would do the job, probably not without some crossover reworking, however.

I have heard a rumor that PE may still release this driver (the RS90). I really hope they do, because it is a great midrange, IMO. It's biggest drawback as a driver is low sensitivity, but if you can use two in the design, it's not too bad. I'd kinda like to do a center channel with it in a WMTMW arrangement, the MTM vertical, with a neo tweet.

BTW, I did some listening to it early on with no crossover, no tweeter and just the HT reciever crossing over to a sub at 100hz or so and it sounded surprisingly good, especially since it's a metal driver.

[dlneubec]

Dan,
That FR looks amazing. Now I'm sitting here thinking that it is a darn shame that Iowa is such a long time away.

Thanks, Ryan.

I'm hopeful that Shawn will have the InDIYana event this year in April. Last I heard he is still planning to, but no formal announcements have been made yet. Last year it was in Richmond, IN. That might not be too bad for you to come to.

[dlneubec]

I forgot to mention in my introduction that the RS225 is in a somewhat undersized box (45L) with a slot port tuned to about 26hz and with an F3 around 31-32hz or so, IIRC. I think it gets considerable boundary reinforcement from the floor in this configuration, regardles of room placement, which extends the response quite well. The strength of the bass from a single RS225 was a very pleasant surprise to me in this design. I've heard folks say there is something about slot ports that is hard to beat. Maybe they are right.

[dlneubec]

Hmmmm.... now I'm thinking of a few things to do with another 240W...

Great results, Dan... I wish those drivers were still available. Anyone heard rumor of them coming back into production? On another thought, how close does a B3N come to one of those as a replacement?

Jon,

I did a little looking and the B3N, B3S or M3N look like they might be reasonable replacements, with very similar specs, except they are all stamped steel framed. The biggest difference I see is in specs is that the B3A-B (RS90) has a very flat impedance curve, where the others have a more significant rise.

The B3S is here: Zaphs Small Driver Comparison



Here is the B3A-B:

Image not available

[JonP]

Jon,

I did a little looking and the B3N, B3S or M3N look like they might be reasonable replacements, with very similar specs, except they are all stamped steel framed. The biggest difference I see is in specs is that the B3A-B (RS90) has a very flat impedance curve, where the others have a more significant rise.

Whoah!! Now that's an easy to read curve... :W

It would seem that the B3A-B has some copper in the motor? Could be. I haven't got around to comparing at Zaphs yet. I guess that would mean that the B3A likely is somewhat lower in distortion. OTOH, the others are available, and probably workable... maybe need a zobel to flatten things out. I have some B3S's and M4N's available to measure. Well, there's a bit more difference in the last one...

I'll go look over Zaph's T/S and impedence curves, maybe make a few for the M4N's now that I've got SE up and running again.

[dlneubec]

I finally have an update on this project! :T

I have a pair built and up and running (see the pics). The final crossover points are around 400 Hz and 3700 Hz. I've only had a few short bit of auditioning, but they appear to exceed my expectations when I first started to pencil out this design. It's hard to believe there is under $300 in drivers and crossover parts to the pair. For music only, they rival and may even better my HOSS omni's, which have around $700 in crossover parts and drivers, though the SoundRounds are a couple db's less sensitive and can't be pushed as hard. The upper bass and midrange is more clear than any speaker I've built and compete with the best I've heard. The treble is a bit splashier with the Aura NT1 than the more subdued RS28's in the HOSS, though I'm sure the RS28's are more accurate and lower distortion, especially at high output levels. The bass seems very solid, deep and clean, with much more output than one would expect from a single RS225, though maybe the slot port loading at the floor level and boundary reinforcement from the floor have something to do with that. There is certainly no lack for bass, which is modeled with an F3 of round 31hz or so, IIRC.

Still, just for kicks, I set them up to cross over to the stereo subs you see in the pic at 60hz. The subs are sealed, but with active LT circuits (designed and built by John K. for my NaO Mini's) to extend the low end and the plate amp crossovers were set to 50 Hz. All I can say is WOW, :E what a great combination that made. For those who say you don't need a sub for music with 3way speakers, I'd have to argue with you now and suggest that maybe you haven't heard a well blended implementation. :W

Clearly they are unfinished. I have some waterfall Bubinga veneer to add to the MDF parts, but i want to wait to do that until a bit warmer weather, maybe about May. I was going to paint the PVC tubing a dark blue or dark burgundy to match the decor, but both my wife and I think they might look very nice left as natural PVC white. The PVC is sanded with 180 grit, so they take on a sort of soft, almost suede material appearance (you can't really see this in the pics). The bubinga will darken them up overall significantly as it is and will hopeful add a nice contrast and richness to the look. I will probably use a linseed oil on the solid maple portions to bring out its colors and richness.

[Hdale85]

Do you have any info on how you built the round enclosures for the mids?

[servicetech]

I got to say those speakers look freakin' cool. If you are married how does the spouse like them? As i see more of the omni designs I'm starting to become convinced that is what my next build will be, and at under $300 I can probably sell the spouse on it too.

Is it just photo or does your TV have the notorious "purple blob" in the center? (That model set "50V500" is known for that problem)

[dlneubec]

They are made of 6" schedule 40 PVC tubing. A 6" x 1.5" flat section was cut out and the maple baffles added (backed by 1/4" hdf). the end cap is maple also backed with 1/4" hdf. There is also a 3/4" mdf circular indow brace just behind the cutout area. Screws from the back side of that brace hold that end of the baffle in place. The rear cap is recessed inside the tube a bit and held by the 1/4" dowels you see in the pictures. When the finish has been added, the dowels will be cut flush. They are lined with 3/4" PE sonic barrier and stuffed with acousta-stuff. I made them a lot more complex than was necessarty, so I can take them apart to paint the tubes and finish the wood. I can post some plans if that is helpful, though I have made changes as I built them.

[Hdale85]

Don't worry about it for now. Was just looking at these and those other new ones your working on thinking my bro in law might like them for his speakers. He wants to build a set of some sort .

[dlneubec]

I got to say those speakers look freakin' cool. If you are married how does the spouse like them? As i see more of the omni designs I'm starting to become convinced that is what my next build will be, and at under $300 I can probably sell the spouse on it too.

Is it just photo or does your TV have the notorious "purple blob" in the center? (That model set "50V500" is known for that problem)

Thanks for the nice comments. My wife seems to really like them, but we're are both a little more oriented toward contemporary type designs in our furniture, etc. What is nice is about them is that even though they are fairly tall, at about 48", the upper part is quite small, with the spine up the back, so they don't feel as massive as a typical 48" tall box can look. They are also lighter than a typical mdf box of that size.

Yup, it's starting to get the purple blob in the center. I was going to try and see if I could do anything about it by entering the service menu and seeing what might be available to help, but I'm afraid it will have to be replaced soon and it's only about 4-5 years old or so. Do you know of a fix for the purple blob, or is it time to move this it to a less used location?

[Hdale85]

For the price of 1080p sets right now I'd just grab one of them Can get a 50" 1080p set for 1500 or less relatively easy. I love my new Sony I got. I would doubt there is much you can do about the purple blob.

[Paul Ebert]

Dan,

I haven't been paying attention to your designs (other than to notice that your HOSS avatar looks very cool) ... until now. These are very ingenious.

I wonder if it would work to change the cylinders for the mids such that their axis is vertical rather than pointing out towards the listener. The drivers would be mounted in the ends, facing each other. This might allow for more midrange options, since it could accommodate larger drivers - perhaps the RS100 (if and when it is available) or the RS52.

Thoughts?

[dlneubec]

Hi Paul,

I think that concept would work! In fact, after I built the first prototype I had the same idea and have a sketch of it somewhere, though never took it to drawings. I can see some benefits and drawbacks to it. It would probably be easier to build. Cutting the sections out of those tubes and fashioning baffles to fit in them was a bit of a challenge. The tubes might have to be larger in diameter to get some volume in them, since they would most likely have to be limited in vertical height, due to driver spacing, etc. However, in that orientation, especially with a larger tube, you might be able to use as large as 5"-6" mids or so. Of course, with a circular baffle, their could be diffraction questions to deal with that the horizontal orientation perhaps avoids better. On the other hand, diffraction from the tweeter interacting with the mid baffles might be reduced over the horizontal tube concept, but I'm not certain of that.

All in all, I think it is an excellent idea and one worth further consideration. :T When I get some time, maybe I'll draw something up.

Edit: I've considered and wanted to test how two midrange dome drivers, like the RS52, would work in a hybrid omni concept, which is another reason to test the idea further.

[dlneubec]

Paul,

Turns out I did draw something up, only it was still using the smaller drivers. Check out the attached pdf.

[Paul Ebert]

...The tubes might have to be larger in diameter to get some volume in them, since they would most likely have to be limited in vertical height, due to driver spacing, etc.

Given the closed back of the RS52, this wouldn't be an issue. In fact, any tube would, basically, be cosmetic.

I've got a set of drivers (D26, RS52 and RS225) and most of the enclosure pieces for a M8ta Neo type design, but this is intriguing enough to make me consider changing directions...

Given the efficiency of the RS52 would you really need two of them? That is to say, does the MTM arrangement provide benefits?

This leads me to a general question. How is the vertical response on the SoundRounds? Is this a 'seated only' design?

Thanks!

[dlneubec]

Paul,

I agree that if you are going to use the RS52, there is not much need for the tube, other than aesthetic. In fact a truncated sphere would probably work best, but would certainly complicate the build, unless you are good at turning bowls on a lathe. I was thinking of larger cone mids when I posted the comments about size and volume.

My experience is that you will lose around 6db or so at the listening position when comparing a direct firing version of a driver to an up firing version. So it is a matter of what sensitivity you can live with. Doubling the mid drivers will typically bring the sensitivity back up to about what 1 direct firing driver would produce. Of course, all that energy is not lost, it is just redirected into the room in other directions, so I'm only talking the unreflected sound at the listening position. Of course you have to figure in the sensivity of the woofer and whether you plan to employ it in an omni orientation or a direct firing orientation. If omni, then there is probably little to benefit from 2 RS52's other than they could be padded down to the extent that you might be able to cross them lower without worrying about exceeding driver excursion or thermal limits. If the woofers are direct firing, then you have the potential to use more than one to better match the sensitivity of two mid domes.

There is one other interesting thing about omni design where you are listening off axis. By nature the drivers are rolling off sharply and typically any break-up modes are a good 15db down or so (even in metal cones) as compared to what they are on axis, so this makes your crossover design more flexible in terms of dealing with them, so some degree. Also, I think that driver distortion will similarly be down in the off axis presentation, which might mean generally cleaner sound.

There certainly could be issues with two dome mids firing at each other that I have not anticipated, since I haven't really tried it, however, I don't see any big reason why it would not work. In many ways, it seems like it could work better as it will have more dispersion of off axis sound than a cone might, so potentally less driver to driver interaction. There is less physical driver area that might be effected by the other drivers output and the drivers don't have the depressed cone shape which might cause some cavity resonances when they are firing at each other. On the surface, all things seem to point to domes being potentially better.

I have not measured the Sound Rounds vertically yet, but I don't hear anything that suggests they have a narrow vertical dispersion as I stand up and move around the room. My thinking is that as you move above or past 90º verically off axis of the down firing mid, you are moving under 90º vertically off axis on the up firing mid, so they essentially compensate for one another's losses as you change vertical positions, though you would eventually begin to get phase cancellations or comb filtering, I guess. Again, I have no measurements at this point to support this theory however.

Up and down firing mids do pose challenges in that you have to experiment with the distance between the baffles in order to get the best natural response from the mids. The tweeter baffle shape, size and position, both vertically and front to back have an effect on this as well, particularly as you start to cross over higher. A 1500hz crossover is a lot less sensitive to tweeter baffle size and position than a 3500hz one. That is one of the major reasons why there is essentially no tweeter baffle on the SoundRounds. The less mass between the mids, the more natural the mid response and roll off was.

Basically, it takes some testing to determine the optimum spacing of the mid baffles and what the best size and position for the tweeter baffle is. You do have the same issues with a single mid/woofer and tweeter design, however. I had the same issues to deal with in my Duo design (see the thread below, if you have not checked it out) and it has a single up firing woofer and a front firing tweeter in a simpler two-way design.

I hope that answers your questions.

The Dou project

[dlneubec]

I have been very busy trying to put the veneer and finish on my SoundRound project (and then the Duo project) in an effort to get them ready for the Dayton DIY in July.

I wanted to post this mainly to warn folks not to make the mistake I did yesterday. I had one of the bass bins on the SoundRounds complete. I went with a high gloss finish on Bubinga veneer (over mdf) and a semi-gloss on the solid Maple. See the photo's below. The Bubinga turned out looking very nice, though it was a learning process with the raw, unbacked veneer, Heat-lock glue, veneer softener and the iron on method. I took the photo's below mid-day yesterday. I do this work in my garage and it faces west.

I had the box a few feet back from the garage door and forgot about it when I went in for a couple hours. As the sun got lower in the sky it shined directly on the veneer on one side of the box and it got very hot. The veneer, though it was completely sealed under multiple layers of dewaxed shellac and oil based poly, cracked severely. I don't have a picture of the cracked veneer yet, but will post one. There must be a 7-8, 3"-5" long cracks, perhaps as wide as 1/16" on the side that was left facing the sun. When I saw that, I was devasted. I took it inside to cool off in the hope the cracks might shrink a little, but they have not. I'm not sure how I'm going to repair it at this point. I have no matching veneer pieces left of a size that would allow me to cover the entire side. I could possibly fill the cracks and refinish, but it will not be what I had been hoping for.

Anyway, be forwarned, take care to not let your veneered boxes sit in the sunlight for any length of time, at least when using unbacked veneer over mdf. I'm still sick about it. Here are the before photo's of the veneer.

[Dennis H]

Oh man, that sucks, Dan.

[dlneubec]

Well, other than trying to fix the bad side with the checking due to the sun, I have one box complete and the second about 90% complete and should be done next weekend. The comination is bubinga veneer and solid Maple. Now I just have to figure out how to fix the one "checked" side. Then I have to figure out how to get these, the HOSS and the Duo's all to Dayton. Not sure it is possible to do in a car and I lost my SUV access.

Here are some photo's of the one completed speaker, in my garage.

[augerpro]

Do those really look like mirrors in person?! Dan you must share your method...

[Hdale85]

Those look excellent!

[dlneubec]

Do those really look like mirrors in person?! Dan you must share your method...

Thanks guys! It is not quite mirror, but the picture does reflect what they look like.

It is not that hard to do, it just takes patience and time. In this case here is how I did it. After the veneer was applied, I sanded with 240 grit to a smooth finish. Then I applied one coat of Zinsers Sealcoat or dewaxed shellac to seal it. After drying, sanded with 320 grit, wipe clean and add a second coat of sealcoat. Dry and sand with 400grit to smooth. You essentially want to remove all the "shiny areas". There will still be grain left showing, however. Now I apply one thick coat of Old Masters High gloss poly to one flat surface at a time. I generally tape the adjacent edges with painters tape to keep from dripping over or running onto the adjacent surface. I leave that side horizontal while the poly sets up to avoid any drips or sags, maybe a couple hours. This was all applied with a brush, btw, not sprayed on. The horizontal drying and thick poly coat pretty much settles out any brush marks. You can turn it and do an adjacent side after a couple hours, if you have it prepared already, though you can't yet tape over the first side of poly. I usually let the poly dry for about 2 days. Then sand it with 400 grit. You should now have almost a glass smooth surface and there should be very little, if any grain left showing. Now apply a second coat of Old Masters, just like the first and after a couple days of drying, you now sand with 1500 grit to smooth out any imperfections. Then I polish it buffing disk in my hand drill. Of course, I always clean the surface carefully before each application.

That is all that was done to these surfaces. It will now feel pretty much like glass. It might be improved even further with rubbing compound and wax, but I've never felt the need to go that far.

[dlneubec]

I realized I never posted a final update on this design. Here is a picture of this design as completed, followed by a measured FR at 0-15-30-45º off axis. The graph pretty clearly shows how these are omnipole up through about 4khz and really don't drop off much off axis above that due to the nearly non-existant tweeter baffle. I should create an image of the final xover design to post also, however, I doubt any one tries to build this, since it is rather unique looking and difficult to build and the HiVi B3-AB (Dayton RS90) driver is no longer available. These are available for sale if anyone is interested.

These speakers won their category at DaytonDIY 2008 (under $200 driver cost category) and at IowaDIY 2008 (budget category, under $350 crossover parts and drivers). They received several votes of 8 (0-8 range) for the soundstage/imaging category of judging at Iowa.