Omnidirectional Speaker Project, any interest, help?

[dlneubec]

Hi everyone,

I have an update on my progress.

Last night I mocked up my first choice for the crossover design. It was the one called Bump3 posted above. I figured I might as well start with the simplest one (9 components). I listened to it quite a bit last night and was very pleased. I have only one speaker built to listen to, so that makes it a little harder. I can say that it is incredibly open, detailed and smooth.

This morning, I measured the response. The farfield measurement, with a 5ms gated window is shown in the first pic below. I was a bit concerned about the peak in the 7kHz to 11kHz area so I decided to go ahead try to implement my second choice, the Bump3a shown in the post above. It is essentially the same as bump3, but altered to include a RLC trap in the tweeter net to flatten the response in that area. The designed response is shown in the second photo. The measured response is shown in the third photo. This was a farfield of the entire system merged with a nearfield on the RS180's. I had already messed around a bit with some of the trap values and the R12 resistor from the original design. I wanted to start with a little bit hotter top end, since a number of people consider the RS28 a bit subdued at the top end. I can't be sure yet which version I like better, but I think it is the one with the RLC trap. I'm as pleased as I could be with the sound, given I'm listening to just one speaker. It has a VERY big sound for one speaker. I plan to listen to it for a week or so before considering any further changes.

Since this system will be used for both music, HT and general TV watching, it may take awhile to find the optimum settings. I will use it on one side as the left main and continue to use the NaO Mini as the right main.

One question I have that maybe someone can answer. The impedance drops a bit below 4ohms in the 3kHz to 6kHz area, with the 1ohm resistor I started with at R12. Should I be concerned about this? There doesn't seem to be any problem as far as the amp is concerned. If I up R12 to a 1.2 or 1.5ohm resistor, it brings the impedance up in that area, but of course drops the SPL of the tweeter a bit also.

Anyone see anything else in my FR, impedance or impedance phase plots that I should be concerned with?

[john k...]

(JPK) Congradulations Dan. Measurements look pretty good. The impedance dip should be of no concern. You will probably have to make your final judgement based on listening. Not bad for a first progect with SoundEasy.

[dlneubec]

Hi all,

Here is a moc kup of how I'm thinking of finishing the omni. It is Italian Rosewood, with black trim and baffles. I'm not sure if I'm going to do a gloss black or a truck bed liner, but I'm leaning toward the latter.

Any opinions?

[juventino]

Hi all,

Dan, I'm very impressed with your work!

So you have decided to eliminate the diffusor for the tweeter? If it works for your setup, then by all means, do it.

But the question remains - how do the professionals (Duevel, Mirage, B&O, etc.) manage to diffuse their tweeter and retain response in the high frequencies?

Just using the above-mentioned manufacturers as examples, we know it can be done using several different designs.

Do the shape of the diffusor and the height at which it is suspended have to be just right? Or are these parameters incredibly sensitive to the tweeter used?

It would be very disappointing if we had to give up omni-diffused high frequencies; having an excellent off-axis response from the tweeter is surely one of the most appealing aspects of an omni design?

It seems a shame to compromise on what could be a truly glorious loudspeaker... though compromise often seems to sum up DIYing.

I really need to shed some of this idealism!

Anyway, I'm going to give this omni thing a shot myself. I've had some EJ Jordan JXR6 HD drivers lying around for a while, and aside from the obvious problems of sensitivity and the dubious power handling of the Jordans, I'm going to try diffusing them with a variety of shapes. I've got some egg-shaped pieces of pine lying around of various sizes, and while at a hardware store recently I noticed that those 'hats' you put on top of chimneys to keep the rain out of them would, inverted, be a potentially useful diffusor.

So I'm going to get a hold of some materials and get DIYing! As soon as I get some spare time...

[dlneubec]

So you have decided to eliminate the diffusor for the tweeter? If it works for your setup, then by all means, do it.

But the question remains - how do the professionals (Duevel, Mirage, B&O, etc.) manage to diffuse their tweeter and retain response in the high frequencies?

Just using the above-mentioned manufacturers as examples, we know it can be done using several different designs.

Do the shape of the diffusor and the height at which it is suspended have to be just right? Or are these parameters incredibly sensitive to the tweeter used?

It would be very disappointing if we had to give up omni-diffused high frequencies; having an excellent off-axis response from the tweeter is surely one of the most appealing aspects of an omni design?

It seems a shame to compromise on what could be a truly glorious loudspeaker... though compromise often seems to sum up DIYing.

I really need to shed some of this idealism!

Anyway, I'm going to give this omni thing a shot myself. I've had some EJ Jordan JXR6 HD drivers lying around for a while, and aside from the obvious problems of sensitivity and the dubious power handling of the Jordans, I'm going to try diffusing them with a variety of shapes. I've got some egg-shaped pieces of pine lying around of various sizes, and while at a hardware store recently I noticed that those 'hats' you put on top of chimneys to keep the rain out of them would, inverted, be a potentially useful diffusor.

So I'm going to get a hold of some materials and get DIYing! As soon as I get some spare time...

Hi,

Yes, I have given up on the omnidirection tweeter configuration. I can't tell you how the pro's are doing it, but I can sepcualate. I think Duevel does it with their higher end systems by using compression drivers with very high sensitiviry and spl and a true horn, not just a diffuser. I think B&O is really a more limited omnidirectional pattern. They push all of the tweeter and mid output forward with the design of their deflector/diffuser. As for Mirage, I have no idea, unless they are using a very high output tweeter, and attenuating the output in the lower range to offset the drop in response at higher ranges.

I can tell you that I tried ball sizes from 1"dia to 5" dia and all at various heights from 0" up to several inches above the tweeter, as well as egg shapes, concave cone shapes, etc. I even tried them at various front to back positions so the diffusers were not always centered over the tweeter. I tried a Duevel Venus-like tweeter diffuser(but not the BellaLuna horn type) as well. In every case, the result fit the same pattern. While some where better than others, I got a good response up to the 4-5kHz range, as I recall, and then a deep drop off as frequencies go higher and thus got more directional. The best I acheived with diffuser configrations was a drop off starting about 5kHz, dropping about 8-10db by 8kHz and then pretty flat out to about 18-19kHz. John Kreskovsky has been contributing to my education on this and pretty much predicted the results almost exactly. Here is how he explained it:

"(JPK) I posted a comment over at PE but let me expand upon it here. The diffusers/reflectors are mostly cosmetic.

Consider a driver mounted on an infinite baffle. To have omni direction response in the front hemisphere it must radiate uniform SPL in all directions. Now consider a typical midrange driver normally mounted at low frequency it is omni directional. If it were enclose by a hemisphere, such that the driver was at the origin, then anywhere on the surface of the hemisphere the SPL would be the same at low frequency. As the frequency rises the driver becomes directional and the region over which the SPL is constant would become focused around the driver’s axis. The radiated power into the front hemisphere is

P = C x 2 Pi R^2 (1-cos(theta))

where theta is the off axis angle over which the SPL remains approximately constant and equal to C. At low frequency theta would be 90 degrees, thus the power would be P = C x 2 Pi R^2.

At high frequency the power would be significantly reduces. Say the SPL were constant over 30 degrees and then, for the sake of discussion, drops to 0.0. The power would be

P = C x 2 Pi R^2 x (1 - 0.866) =C 0.268 Pi R^2.

Now if we were to redirect all that power so that it was spread uniformly over the region between theta = 45 and theta = 90 (that is along the surface of the infinite baffle and out to an angle of 45 degrees off the baffle) the power would be spread over an area of 1.414 Pi R^2 and the power would be

P = C' x 1.414 Pi R^2 = C x 0.268 Pi R^2.

Thus,
C' = 0.268/1.414 = 0.18 or -7.22 dB.

As the driver become more directional these things get worse fast. For example, with a 15 degree spread on axis, a perfect redirection over the same horizontal region would down 13dB. For the most part the reflectors do little more that scatter the higher frequencies so that there isn't a hot spot above the speaker (when only a single up firing driver is used.

Another really simple way to look at this is to make an analogy to painting a wall. At low frequency the total power radiated is like a gallon of paint. At higher frequency it radiates a ½ pint. No matter how you spread it, for a given wall area that ½ pint just isn’t going to be able to provide as thick a coat of paint as the gallon. It’s volume of paint divided by wall area equals thickness. And thickness = SPL.

Another simply analogy might be to consider a driver at low frequency like a floor light and at high frequency like a spot light. If the light is the same intensity on axis, then when you disperse the light from the spot light over the same area as the flood light, it will no longer be as bright."

For the RS180's I found that they have a steep natural roll off starting around 1.5kHz when performing in 4pi space and listening 90º off axis and up to that point are omnidirectional where diffusers provide very little, if any help. The test I did with the Duevel-like diffusers and the RS180's showed that they would help extend the natural drop off to a bit higher in frequency, maybe 300-400Hz, as I recall, and would probably allow a higher crossover point. I decided however, that too use them would cause more severe problems than it would help, since the tweeter could not be mounted between the RS180's and would have to be set on top of the box or mounted to the front, each having particular problems. I think if one were to use a single, very high efficiency woofer/mid, say 98db or so, you could use one Duevel-like diffuser with it and mount the tweeter on top centered over the woofer and front firing and have very good results. Maybe if you used a tweeter up to its natural omni drop off and then a super tweeter to fill in the high end, you might be able to accomplish a good omni pattern at high frequencies.

BTW, it may be that having the system omnidirectional only for the bass and part of the midrange (up to the 1400Hz-1500Hz crossover) may make for the best of both worlds. Many would suggest that omni's have diffuse imaging. Perhaps this hybrid of omni low-mid and front firing mid-high will provide the big soundstage of the omni, excellent off axis performance for a good part of the range, but also much improved imaging due to the front firing element. Also, if you look at my off axis plots of the tweeter in this application, you will see that it performs quite well horizonatlly off axis.

[john k...]

Hi Dan,

I was wondering what you final verdict was since I haven't heard from you for a while. I would like to add something about some of these diffusers, in particular the B&O Beo5 type, acoustical lens. If you read the B&O literature what they say about the acoustical lens is that it results in a more uniform dispersion of power over 180 degrees horizontally but limited vertically. What it doesn't say is if the frequency response is flat. The point I am getting at is that there are two considerations, uniform dispersion of the radiated power at a given frequency and the variation of the radiated power with frequency. Assuming the AL does redirect the radiated power so as to provide uniform dispersion then, as I indicated, the frequency response at a given position in space would roll off. However, that does not preclude applying equalization to flatten the response. The B&O Beo5 is a digitally, active speaker and I suspect that such equalization is applied.

Given that a typical conventional dome driver mounted on a flat baffle will have a power response that rolls off at 6dB/octave with increasing frequency for flat on axis response, due to the increasing directivity of the dome, then if that power were redirected so that the directivity became constant 6dB/octave increasing total power into the driver would be required to retain flat response with frequency. That's pretty easy to do actively while retaining good efficiency, but passively it means throwing a lot of power away in the crossover.

Anyway, you did the experiments and saw the results for the ball and hyberbolic diffucers.

[dlneubec]

Hi John,

I was hoping you might chime in on this one and clear up my attempt at an explanation of what is happening. :B

BTW, you got mail. I sent you a crossover or two to take a look at a couple days back that you might have missed. I just resent it. I'm attempting to raise the crossover point some and reduce the contribution of the RS28AS below the crossover point. The one I have sounds great, but it might be even better if I don't push the RS28 quite so low.

[dlneubec]

The Mentor omnidirectional project:

Here is an update for those interested. I’ve decided to call it the Mentor omnidirectional project in honor of those who have mentored me through about a half dozen DIY projects, since I got into it a little over a year ago, culminating in my first complete design. Without folks like John Kreskovsky, Roman Bednarek and a few others, I would never have been far enough along to attempt a project like this.

This has been sort of a seat of the pants process. I’ve been learning about crossover design and software (SoundEasy), while trying to learn and understand the complexities of an omnidirectional application. This was a lot for a novice to bite off in one chunk! It was originally planned as fully omni, but ended up a hybrid omni-front firing combo.

I’ve been spending the last month or so measuring, testing, listening, changing, re-measuring, etc. and believe I’m very close to the final design. I’ve now two complete speakers, except for the finish on them. The crossovers are not finalized in terms of soldering, mounting etc. but are very close to a final design.

The drivers used are as follows:

1-DAYTON DVC310-88 12" DVC SERIES SUBWOOFER:
2-DAYTON RS180S-8 7" REFERENCE SERIES SHIELDED WOOFER
1-DAYTON RS28AS-4 SHIELDED 1-1/8" ALUMINUM DOME TWEETER

I had all these drivers on hand from previous DIY projects. The subwoofers were used with my NaO Mini’s and a previous project (DrK MTM’s). Attached are some photo’s of the project (unfinished). Also attached are the crossover design from SoundEasy and the measured FR taken (farfield only shown, RS180x2, RS28, no sub). The subs and RS180s are in sealed cabinets. The subs have active analog networks, designed by John K. for the NaO mini, for a Qtc of .5 and to extend response to 20hz. They roll off on the top end at 100hz, where the RS180’s take over up to about a 1420hz crossover to the RS28.

I have built and tested probably a dozen tweeter baffle shapes and sizes to fine tune the design for as flat a response as possible. This includes the felt added to the top and bottom bevels on the tweeter baffle and to the front of the top box and various height adjustments in 1/8" increments. One final tweak was to tilt the tweeter baffle back, so it is firing about 5º above horizontal, which helps to flatten the on axis response of the tweeter. The combination of smaller tweeter baffle, felt and up tilt removed the broad 3db peak from about 6-10kHz which I had previously used the notch filter to suppress. I have added a notch filter to the woofer net in order to suppress the breakup of the RS180’s even further, which is now over well over -40db.

The boxes are modular and can be completely disassembled in a matter of minutes all the way down to the base. They will be finished in a 22.2mil Italian Rosewood veneer sides combined with painted (probably black truck bed liner) baffles and trim. The crossovers will eventually be placed inside the sub cabinet. I’m currently using 300w MCM plate amps to power the subs. MCM plate amp
They are currently sitting on the floor next to the speakers. I will build wood frames for them and hang/bolt them on the back of the speakers, so they can be removed I favor of a single amp, whenever I get the funds to upgrade. BTW, the RS180-RS28 combo with my crossover measure with an average impedance of around 5ohms, a bit higher than the design modeling indicates.

Now comes the hard part for me, describing how they sound. First let me tell you that my previous DIY projects is limited and include the DrK MTM, then the same MTM, but with a RJB 2.5way crossover in place of the DrK. Crossover. I built the two stereo subs as stands for the DrK’s. I then built RJB’s Microbe SE’s for surrounds and Microbe SE 2.5way MTM for a CC. Next, with the help of John K., I replaced the modified DrK’s with the NaO Mini’s, still using the stereo subs for stands and adding the active control to the subs. These speakers are make up my recent listening experience, since there are no high end hifi stores in my area, only BestBuy and those types. I also briefly owned some Polk speakers shortly before I got into DIY. So, I’m not and experienced audiophile or listener. Also, my equipment is very modest. I use a Harmon Kardon AVR240 HT receiver; an Adcom GFA-5400 amp, MCM 300w plate amps, a Panasonic DVD player for CD’s via inexpensive digital optical cables and the rest are Dayton RCA interconnects and a woven standard 14g speaker wire for the speakers. The components in my crossovers are Jantzen air cores, mostly Dayton 1% metalized poly caps, with some Dayton film and foil bypass caps, and Mills and Dayton resistors.

That said, I’m thrilled with the sound of these speakers. They seem very smooth and extended. The bass goes as deep as I can ever want, and have that tight sealed bass sound. It even seems great for HT with enough deep bass to shake the walls and floors and furniture. The subs and RS180’s play omnidirectional due to their up and downfiring orientation, which covers up to the crossover with the RS28’s. I believe this helps give these speakers the biggest, fullest sound I’ve heard. The measurements I’ve taken indicate that off axis performance is virtually identical (out to the 60º I checked). The RS28, perhaps due to the small, chamfered baffle also has very good off axis response, yet, I think they contribute to very good imaging because they remain front firing. The midrange is the most clear, clean and open I have heard. It has a “real” quality, particularly noticeable in vocals, that I have not heard in my previous experience. The soundstage is very expansive in both width and depth. They also seem to play very loud, cleanly. The only thing I’m still playing with is the R12 resistor level on the tweeter. I previously had it at 1ohm when listening through my Adcom, but it now needs repair to one channel (don’t ask! ops: ) so I switched to powering these with just my HK receiver, which handles them with no problem. The HK actually seems a little smoother and little more laid back sounding than the Adcom(which I think I may end up preferring), so with it I have gone to .6ohm on the tweeter net for now. One thing I lost with going to the HK, was that previously I had controlled the roll off of the RS180’s to the subs at 100hz with a little line level passive 1st order filter using a small cap (.033uf, as I recall) in line between the preamp and the amp, which I can no longer do, so the integration with the sub is not quite as smooth with the HK alone, though the natural roll off of the sealed RS180 is very similar to the first order.

[Dennis H]

Lookin' good Dan! :T

[dlneubec]

BTW, here is a shot of my crossover, in jig form. I got the idea for this jig from Roman B. and I think it is a great idea. For the price of a few cheap binding posts, single and dual banana plugs, you can mock just about anything easy and fast. It also avoids issues that might result from a lot of small alligator clips and leads.

I'm sure others have better methods, but this one works very well.

[dlneubec]

Attached is a comparison of the measured neafield plus farfield FR of the my first Mentor omni box against the second one I built. The FR's include just the RS180's and the RS28, not the built in subs. I can't remember for sure which one is which on the colors in the plot, but one of them has well broken in drivers, from an older MTM project from a year ago (the red plot, as I recall) and one has brand new drivers with just a few hours worth of time on them. Also attached is an impedance plot of the two overlaid. Most of the caps are Dayton +-1% metalized poly caps, but there are a few standards (+-5%) in there also (what I already had on hand). I'm using a few Dayton F&F poly bypass caps in the tweeter circuits also. The inductors are all Jantzen air cores(+-3%) and the resistors are Mills (+-1%) and Dayton(+-2% in the woofer notch filter).

I have a few questions for the experts:

To my ear, they sound absolutley fabulous and I was thinking that the match was pretty good since one set of drivers was new and just purchased and the others were over a year old. However, are the differences between the two speakers significant enough to warrant further adjustments? If so, is it best to wait awhile to let the new drivers break in and then do a new measurement before any further tweaking?

I don't have an LCR meter to measure the components, so I either buy one or rely on the tolerances of the parts. I was thinking of using all Dayton 1% caps (except the F&F bypass caps) in the final build to ensure a closer cap match and calling it good. Is that a good approach, or should I sweat the inductor match as well? If so, what is a good enough, but cheap meter I can get to check them?

When attempting to match a pair of speakers response what is the best approach? Measure each component to ensure they match, or mess with the individual crossovers while measuring changes to get them closer, or both?

One other factor. My room is not symmetrical and there is a large doorway on one side leading to a dining room, so the speaker on that side will never have the same in room response as the other, unless there is a major room rearrangement (a definite SAF no-no), so perhaps none of this matters and what I have is close enough.

Any thoughts?

[cjd]

I would venture to suggest that you've got things close enough.

I note that the black response lacks the dip the red has - what caused this?

[ThomasW]

I doubt that you're going to hear a difference in these speakers. And unless the room placement is identical the differences aren't significant enough for concern.

[dlneubec]

I would venture to suggest that you've got things close enough.

I note that the black response lacks the dip the red has - what caused this?

Are you talking about the dip at 3khz? Look again, they both have it. If you look at the speaker pics posted above, you will probably be able to figure out why it's there. It has been in the raw measurements from day 1. I believe it to be diffraction of woofer cavity related problems due to how the tweeter is mounted and set back from the front of the baffle.

Felt on the top baffle, tweaking of the shape and size of the tweeter baffle, vertical positioning ajdustments of the tweeter baffle and a slight upwards angle on the tweeter baffle actually have reduced it significantly from what it was originally.

[dlneubec]

I doubt that you're going to hear a difference in these speakers. And unless the room placement is identical the differences aren't significant enough for concern.

Thanks, cjd and Thomas, I'm glad to hear that you think this is close enough! :T

[dlneubec]

I would venture to suggest that you've got things close enough.

I note that the black response lacks the dip the red has - what caused this?

Ok, I see what you are talking about? The dip at the low end?

I'm not sure. FYI, these measurments, with SoundEasy, were in room, the farfield at with an 8ms window (used a 4ms for design) at 1m from the front of the speaker plane and the nearfileds are at 100ms but taken at the front of the speaker plane with the mic centered between the woofers and the tweeter baffle in place.

I have noticed that the bottom end of my in house measurements can sometimes vary quite a lot, maybe ambient noise from a refrigerator or furnace, etc.?

[cjd]

Yes, the bass response dip. Something is doing that, and I bet the minimal difference you have in the "flat" zone of response is related to this dip. So they may be closer than they appear.

C

[dlneubec]

That could be. I do know for sure that the dip was from the nearfield response of the 2 RS180's in that speaker, which was merged with the farfield response at 300hz, as I recall. Wouldn't that suggest that the only section it could effect is up to 300hz?

Still, it could explain the section of the response where the greatest difference between the two speakers occurs. Above 300hz, they seem pretty close.

[Rudolf]

As a (small) design alternative in the "front firing tweeter, up firing woofer" category I would like to add a project I just found on the net. It is from Germany, but the pictures will tell most everything. X-over design is not published. You have to buy the drivers at http://www.lautsprechershop.de/kunde...ugel/index.htm to get the plan.

This is the design: http://www.flo-horns.de/PallaColonna/PallaColonna.htm

[dlneubec]

That is an interesting looking design!

I tested something similar. It was a circular front firing baffle for my tweeter orientation that used a dome shaped diffuser on the rear behind the baffle, all centered over the up-firing and down-firing woofers. It was a 6" diameter dome. I tested it in various vertical positions between the two baffles. I also tested one with the tweeter offset in the baffle rather than centered. Neither gave as flat a response as the top and bottom chamfered rectangle in my final design. The round baffle with the offset tweeter measured flatter overall than the one that was centered, however. I assume this was due to less diffraction issues.

I don't have a photo of it when I had in my current setup with both up and down-firing woofers, but here is one back when I had just one up-firing woofer.

[dlneubec]

BTW, I got most of the veneer on one speaker this weekend. It is 22.2mil paper back BFV in Italian Rosewood from Oakwoodveneers.com. It will take about 4'x5' of a 4'x8' sheet to do both speakers. I cut roughly 5 12" strips, 4' long to start with an exacto and straight edge. Then I cut all the individual 4.5" width pieces with one of those large lever blade paper cutters(24 pieces per speaker). It worked extremely well.

Since they were small pieces and easy to work with, I used contact cement. Because I was doing it inside in my living room, I used water based. I covered the mdf and veneer with two coats and then used two pieces of wax paper over the area I was applying it to. That way I could easily position the piece, hold it down, pull out one piece of wax paper and press down in the center of that side. Then I pulled the other piece of wax paper out and pressed the center down. From that point it was simply using a j-roller from the center out. It worked like a charm.

It starts to make it look a little more like a real speaker, don't you think? At this point my plan is to finish the veneer high gloss, perhaps with urethane oil (have to test it first to see if the oil causes any problems) and paint the trim with truck bed liner, but it doesn't look bad with just the natural MDF color either.

[AJINFLA]

Cool. Baby monkey has his own little coffin above . I've pointed out Jack's main speakers before, but I think if I ever get around to doing an omnipole other than a Pluto it might be something like this http://jgbouska.tripod.com/audio/plan9.pdf
Might even throw in a couple flex pipe arms and play the sound track to Lost in Space..

Think his friend might have had an Orion/Bob type dipole in mind?


cheers,

AJ

p.s. check out his site - if you haven't already http://jgbouska.tripod.com/audio/

[dlneubec]

Hey AJ,

That is very cool project! I might have to try something like that sometime, when my skills are a little stronger.

Interestingly, his description of how those speakers sound fits the way I would describe my omni's, except perhaps for the deep bass performance.

[dlneubec]

Things are moving slowly due to the weather. Here is a photo of the veneer after urethane oil was added. I think I will be putting on a high gloss poly spit coat over this once it has dried in a few days (72 hours minimum according to Watco).

Once the veneer has been finished on both boxes, then I will have the task of taking them all apart, masking the veneer and painting all the trim (still thinking black, probably truck bed liner).

[dlneubec]

Hi All,

The finishing is coming along on one speaker, but I have had one big setback and change of plan. I've finished adding urethane oil and pretty much have the spit coat poly finish complete. I wanted a relatively high gloss finish while still retaining the indentations of the grain so it looked like real wood. Attached are three pics of the results so far. The second two show some closeups the box taken back apart. You can see the 12" DVC sub in the base.

The setback has to do with the black, truck bed liner finish I was planning for the trim and baffles. I'm not happy with it at all. The reason is that in my mock up I looked at it as if it would be under full front lighting, but in my room, the speakers are back lit by windows and patio door. You can see it on the box to the left in the last two photo's. I painted just the top box and baffle and it just sort of disappears in front of the back lighting. So, now I have to figure out how to take off, sand, somehow cover up the truck bed liner and cover it with another paint color. Probably a light to medium tan, much like the natural MDF color. I'm thinking of keeping it a flat sort of finish as a contrast to the gloss veneer.

Anyone have any ideas on how to cover up the truck bed liner? Any thought on what type of paint to cover it with? Can I sand down the TBL and use it as a sort of primer? anybody have some different ideas on color?

Thanks.

[Dennis H]

You might try Rustoleum Textured Enamel on a scrap and see if it will stick to the truck bed liner and see how it looks. They have a light tan (sandstone).

http://www.rustoleum.com/product.asp?frm_product_id=20&SBL=1

[dlneubec]

Thanks, Dennis, that looks like it could work! I'll take a look at in person and maybe buy a can to test with.

[dlneubec]

I have just about completed the finish on one speaker. It is Italian Rosewood Veneer (22.2 mil paper backed, Oakwoodveneers.com), with a couple Urethane oil coats and then a wipe on poly/naptha, 1:1, spit coat finish.

I first tested the trim and baffles in black (truck bed liner) and didn't like the way it looked, see note and picture above. I decided to go light and tried several textured finishes such as Rustoleum sand, a tan suede-look and I ended up liking the stone-textured look, which is what us shown in the photo below. It's kind of hard to see the texture in the photo, but it is distinct in person. The tweeter baffle and base is done in a textured black metallic from Duplicolor.

I like the contrast between the light heavy textured trim and baffles and the high gloss veneer. I still have to solder up the final crossover and then have most of the other speaker to add the finish to. It's hard to get this done when you've had a several of single digit to low 20's temps. It's kind of cold in my uninsulated garage and though the kerosene heater warms it up pretty good, it tends to stink up the house! :blink:

[critofur]

interesting comment about the transition to point source radiation. Worth discussing abit i think

That's what current Ohm Acoustics speakers do...

[mike gagne]

I've been thinking that there is probably something ready made to use for the deflectors, lie some kind of large funnel. Somebody just has to stumble upon it.

On the other hand, there are wood shops everywhere that should be able to turn these out relatively quickly and I would think, not too expensively, depending on the material and if they do any finish sanding/etc., especially if you were to get a multiple order of them together. Ideally you could make them modular so they could be stacked back to back to make a two sided lens or use them single sided, which would open up more design possibilities.

I hope there is enough interest to keep this thread going.

As a professional woodworker with a lathe I can tell you that the cones are an easy turning project. The trick will be to get large enough pieces of wood that are dense enough to turn without shredding into fiber. Medium to high density materials (like oak, ash, beech, etc) may work but might distort after turning. Moisture content is key. The Duevals look to be turned from glued up layers of plywood. This is very doable just make sure you use hardwood ply (Finply, Appleply etc) and not fir core which turns with some difficulty and would likely have knots and voids. I wonder if spun copper or aluminum vessels would work as deflectors. These spun shapes can be made from sheet stock with a simple device that is common in certain types of metal shops. I am going to make a set of these after I finish my high efficiency towers. Thanks for all the great ideas in here. aloha