The Mentor II: omnidirectional project #2

[joecarrow]

Well, the RS28a isn't exactly known for soaring top-end extension to begin with... the "28" refers to millimeters- that's 1.10 inches, so it's really not going to have great off-axis response.

I've been a lot of talk and not a lot of stepping up with measurements. I happen to have some of those tweeters, an ECM8000 microphone, and a laptop with what I think is a decent line input. Where can I get some software to measure this? I'd be glad to mount one of these tweeters on a baffle and take a few measurements. I have some clay as well, so maybe I can even come up with a diffuser.

[dlneubec]

I think your prob. with the originals had to do with no "grill cover" and that somewhat "inventive" use of italian rosewood veneer. Make it more monolithic esthetically would prob. improve things a LOT. Funny though, BRILLIANT idea - first project is somewhat "taste challenged" allowing W I D E latitude for future projects.. ;x( :B

The ohms are fairly monolithic AND.. hmm.. who was it that suggested an obelisk shape without the top pyramid?

Your best bet there for dimensional stability is using HVAC filter foam like Wilson Audio uses (..obviously painted gloss black).

Hi Scott,

I never thought about HVAC filter foam. I'll have to look for some.

Here is a look I came up with that is more monolithic as you mentioned. In this case, I was thinking of creating narrow slots around the edges of the box at the openings and using standard speaker fabric. It would be held in place in the slots with the same material used to hold screens in windows. I think it would have to be sewn together in pieces with seams at the corners, rather than just stretched around it. I'd have to figure out what to do with the surface mount tweeter in the back.

It would also look more monolithic if you only covered the woofer section in the middle and left the top open.

[dlneubec]

Well, the RS28a isn't exactly known for soaring top-end extension to begin with... the "28" refers to millimeters- that's 1.10 inches, so it's really not going to have great off-axis response.

I've been a lot of talk and not a lot of stepping up with measurements. I happen to have some of those tweeters, an ECM8000 microphone, and a laptop with what I think is a decent line input. Where can I get some software to measure this? I'd be glad to mount one of these tweeters on a baffle and take a few measurements. I have some clay as well, so maybe I can even come up with a diffuser.

Hi Joe,

I ordered a couple of the RS225's last night, a second RS52 and decided to go ahead and pick up this tweeter as a possible supertweeter in a 4way or as a high crossover from the RS52, if I can figure out how to get the CTC spacing closer to the RS52.
Dayton ND20FB

I'll probably have some mearuements of this tweeter in the next few days, both on and 90º off axis.

I'm using SoundEasy for measurements, crossover design, modalling, etc. I'm not familiar with the others, but I have seen Speaker Workshop (freeware), LspCad, and Arta (also has a free download version, I believe) mentioned quite often as options. There have been a number of threads in this regard that will help guide you if you do some searching. You might also have to invest in an external USB soundcard (~$80). Most internal soundcards in laptops do not work with measuerment software. Here is the one I got that works very well.

http://www.sweetwater.com/store/detail/UA1EX/

I first tried an MAudio Transit, which many seem to favor, but had nothing but problems with getting it installed and working properly. I had none with the Edirol.

[dlneubec]

In the interest of keeping an open mind and considering all options, I tried some measurements of the RS52 with some diffusers above it last night. I tried an 8" and 6" dome/half=sphere diffuser and a Duevel-like diffuser (see the photos below)

In the graph below, the blue line is the best FR I could get with the box/baffle above the RS52 (1" up), the black was the 6" dome diffuser and the red was the duevel-like diffuser.

I was impressed especially by the Duevel-like diffuser in that it has a smooth lower end and upper end roll off. the upper end roll off is pretty smooth all the way out to 9khz.

It got me to thinking about using a combination of an upfiring RS52 with an 8" waveguide firing into a duevel-like diffuser. I know that an 8" WG is not large enough to do much by itself with a 2" dome mid, but combined with this diffuser, not so sure. Attached is a conceptual plan where I do this with a cut down duevel-like diffuser in which I house and upfiring tweeter, and above that firing down is a supertweeter. The thinking would be a 4way with a crossover at around 6khz to the second tweeter.

Hey, I know it's way out there, but I enjoy thinking outside the box!

[nerd of nerds]

I have a somewhat off topic question but still relevant to the conversation at hand...

It is my understanding that the reason we time align drivers is so that the waves at the crossover frequency don't end up cancelling each other out on the listening axis, rather off the vertical axis. We can get around physical time aligning by using a delay circuit on the tweeter, since it is generally farther forward than the woofer.

So, who does the pluto have to have such a (frankly) non WAF friendly design? Yes, its small, but I really am not fond of the look. Couldn't one just build a standard box with the woofer on top, tweeter on the front, and give it a good bit of delay?

[dlneubec]

I have a somewhat off topic question but still relevant to the conversation at hand...

It is my understanding that the reason we time align drivers is so that the waves at the crossover frequency don't end up cancelling each other out on the listening axis, rather off the vertical axis. We can get around physical time aligning by using a delay circuit on the tweeter, since it is generally farther forward than the woofer.

So, who does the pluto have to have such a (frankly) non WAF friendly design? Yes, its small, but I really am not fond of the look. Couldn't one just build a standard box with the woofer on top, tweeter on the front, and give it a good bit of delay?

As I understand it, passive delay networks are very complicated and something you want to avoid where possible. I believe it is much easier with an active digital network approach.

Here is an article on time and phase alignment by John Kreskovsky.

http://www.geocities.com/kreskovs/TimeAligned1.html

[dlneubec]

I have some very interesting results from waveguide and waveguide combined with diffuser experiments.

I was interested to see what the RS52 would do upfiring in a 8" PE waveguide. I got the WG yesterday and after work wet to cutting it back to fit the RS52, just outside of the mesh grill. It turns out that it has 4 bolt heads in that area that complicated things a bit.

First let me explain how I dealt with altering the WG. I came up with an easier way (for me) to cut the WG down. I also used this technique on a 6-1/2" MCM WG for testing with a 27tdfc.

First,I cut a piece of 3/4" mdf to a size just barely larger than the size of the WG dia. , in this case just s smidgen over 6-1/2" and 8". Then I screwed the WG to the board so that it fit just inside the square. Then I used my table saw (10" ryobi), which will extend a blade a little over 3-1/2" high at 90º. Then I adjusted the rip fence to a little bigger than i new it needed to be. The idea is to run the mdf piece vertically along the rip fence so that the blade cuts though the WG perpendicular to the throat. For the 6-1/2" WG, I had to pass it thrhough once, then turn it over and pass it through again. The carbide blade made easy work of the plastic. I checked the hole diameter size and then adjusted the fence and passed through again.

All in all, in about 15 -20 minutes, I had both WG's cut down to the exact size I needed. With the 8" WG, I had to pass it through on all four sides to make the cut, but it was no problem to get a good flat cut.

Also, with the 8" WG for the RS52, I had to cut notches to fit around bolt heads on the flange. Attached are some photo's of the RS52 and WG. I simply taped the RS52 to the WG for my testing.

I'll report on the testing in the next post. I'm pretty excited about the results of the WG with the duevel-like diffuser! :T

[dlneubec]

Ok, now for the early results. I started by mounting the RS52 with 8" PE WG in a cardboard box. The WG just set on top of the box, no way to flush mount it. All meaurements are made with the driver firing up and the mic at 90º off axis toward what would be the listening position.

The first graph below shows the RS52 raw without any WG or diffuser in blue. The green line is the best FR I could get by positioning a box/flat baffle above the RS52. The black line is the RS52 mounted in the WG. The pink line is the RS52 in 8" WG with a duevel-like diffuser suspended above it. Wow, look how much more extended the FR is out to 10khz. The WG/Duevel-like diffuser combination increases SPL at 1khz by about 6db over the raw RS52 and then drops only about 8db out to 10khz, where the others drop about 2x that much. With equalization, it looks like I could use the RS52 out much further with this combination.

For reference, the next graph is the same, except it includes the raw 27TDFC upfiring with no diffuser or WG in red. The raw TDFC is a pretty good match to the raw RS52, but as expected, drops off quite severely starting at 6khz, just like the RS28a did back when I tested it.

The next has too many lines on it, sorry, but I'll try to explain. For reference, the black line is the RS52 in 8"WG with duevel diffuser. The rest are of the 27TDFC. Again, the red line is the 27TDFC raw upfiring, now WG or diffuser. The blue line is the tdfc, but now it is mounted in the 6.5" MCM WG. The green line is the TDFC mounted in the MCM WG with a 6" dome diffuser above it. The pink line is the TDFC mounted in the MCM WG with the duvel diffuser above it. Look at how much spl increase there is and how much more extended the TDFC is with the combo of the WG and Duevel diffuser. If one was to equalize the RS52 and tdfc with WG's and duevel diffusers at about 98db, you might have a pretty good FR out to over 10khz. Except for the dip in the TDFC response at about 13khz, it is not that bad all the way out to 17 or 18khz.

Finally I tested the Dayton ND20FB-4 3/4" Neo tweeter, with the idea that maybe a 4-way configuration with this as a sort of super tweeter would help fiull in the top end where the TDFC drops off. The last graph is the same as the last one, except it shows the ND20FB in purple. Note however, that this is in a direct front firing arrangement. It did little if any better than the TDFC when upfiring. My thinking is that it might be useable to fill in the top end, but probably only patially omni, by having it angled forward, somewhere between front firing and upfiring, maybe 45º. I will have to do more tests with it to determine what angle gives the optimum response.

A couple other notes of caution. These measurements were taken on a farily large box. The results are likely to be different with a smaller baffle. Also, all the Duevel diffuser measurements were with the 13" dia. version I had made for testing with the original omni project. I don;t see using something that large, so I will either have to cut them down, or have new ones made that are smaller. I don't know what the effect smaller diffusers will have.

Finally, attached is a pdf of a plan that incorporates the omnidirectional waveguide mid and tweeter, etc.

Any thoughts?

[Paul W]

Dan,
Your idea of using horizontally mounted waveguides for the domes seems like a good starting point. This approach seems to hold much more promise than some of your earlier concepts.

For diffuser shape, holding 360 degree area expansion consistent with any typical 60-90 degree horn will yield a very narrow flare (like the actual Duvell you posted). I've tested the 27TDFC in a waveguide with mixed results and have come to believe that, generally, soft domes are not happy operating in a restricted space (the domes may be deforming). So, I would go with a hard dome tweeter or ring radiator; better yet a compression driver. I'm still not crazy about a supertweeter firing at a "random" angle, so a high sensitivity hard dome or compression tweeter would give you enough output to flatten response in the crossover.

For the diffusers, you could use waveguides for a female mold. Start by adding a hollow slender pointed cone extending back through the throat, add a release agent, then fill with anything from concrete to water putty mixed with horticultural vermiculite. Consider using different size diffusers (keep 'em large) to adjust mouth size while achieving a smooth flare and appropriate height at the "point" of the cone. Keep everything smooth for minimum HF attenuation.

Just thoughts...I haven't tried any of this
Paul

[dlneubec]

Hi Paul,

Dan,
Your idea of using horizontally mounted waveguides for the domes seems like a good starting point. This approach seems to hold much more promise than some of your earlier concepts.

Yes, I agree. I'm starting to get excited about what I'm seeing in my measurements (more on that further down).

For diffuser shape, holding 360 degree area expansion consistent with any typical 60-90 degree horn will yield a very narrow flare (like the actual Duvell you posted). I've tested the 27TDFC in a waveguide with mixed results and have come to believe that, generally, soft domes are not happy operating in a restricted space (the domes may be deforming). So, I would go with a hard dome tweeter or ring radiator; better yet a compression driver. I'm still not crazy about a supertweeter firing at a "random" angle, so a high sensitivity hard dome or compression tweeter would give you enough output to flatten response in the crossover.

Interestingly enough, the 27TDFC and the Dayton ND20FB 3/4" neo, at least initially, seem to give better results out of the WG, but using a 5.5" Duevel-like diffuser. The LAST graph below shows the TDFC upfiring, measured 90º off axis (red). The pink line is the the TDFC in the 6.5" MCM WG. The blue line is the TDFC in the WG with the 5.5" Duevel diffuser mounted above it (best responce I could get from various heights,etc.) The black line is the TDFC upfiring, no WG, but using the Duevel diffuser, with the point of the concave cone even with the back edge of the dome. That responce is probably good enough to use, without the supertweeter. On the other hand, the one with the diffuser and WG is a bit smoother and with some eq to flatten it, may be better, I don't know. I agree, that the supertweeter idea is definately a last resort. What do you think of these plots?

The next graph shows the TDFC (black) without the WG, with the 5.5" concave cone diffuser. Then pink line is the TDFC configured the same, except 90º off axis from the listening position. It's not perfect omni dispersion, I assume because the diifuser is not centered on the dome, but just behind it, but it's not bad at all. The red plot, is the 3/4" Dayton ND20fb neo tweeter, upfiring, on the listening axis, with the diffuser centered on the back edge of the dome (like it was with the TDFC). What a nice response, +-1db from 2khz to about 18khz, upfiring! That looks very useable to me (but what do I know!). The blue line is the ND20fb at 90º to the listening axis, not bad out to 13-14khz.

The final graph shows the TDCF (black) the ND20fb (red) both without WG and with the diffuser just at the back of the dome. It also includes the RS52 in PE 8" WG with a 13" Duevel diffuser above it (blue) and two RS225's firing at each other 4.5" apart. These are with 4ms gated window, FYI.

For the diffusers, you could use waveguides for a female mold. Start by adding a hollow slender pointed cone extending back through the throat, add a release agent, then fill with anything from concrete to water putty mixed with horticultural vermiculite. Consider using different size diffusers (keep 'em large) to adjust mouth size while achieving a smooth flare and appropriate height at the "point" of the cone. Keep everything smooth for minimum HF attenuation.

That's a great idea to use the WG's as forms for diffusers. My son in law is an artist and does a lot of forming for sculpture, etc. I bet he could help me come up with a good way to create the diffusers. I would assume there is an optimum shape for each driver, and WG, but I sure don't know how to figure out what it is. However, the ones I have are actually working quite well.

Attached are a couple photo's of the ND20fb with the 5.5" concave cone diffuser and one with the diffuser mounted over the MCM WG with the 27TDFC.

Just thoughts...I haven't tried any of this
Paul

Hey, I appreciate the thoughts, whether you've tried any of it or not. Anything that might help me keep moving in the right direction I'm interested in hearing. Any other thoughts you have, don't hesitate. :T

At this point I'm leaning toward the two RS225's, the single RS52 in the 8" PE WG with the concave cone diffuser and either the ND20fb, 27TDFC or another yet to be named tweeter. I don't know much about the ND20fb, but the FR sure looks encouraging in this application. The price sure is right, but I don't know about distortion, etc.

[dlneubec]

I had the graphs mixed up in the decription above. It is fixed in the description now.

If I can work it out to use one tweeter, with no waveguide, but a concave cone diffuser instead, I can initially see a couple options for how to design it.

Those two options are shown in the attached pdf's. The second one is closer to the way Duevel did it in their 2way omni.

In one the tweeter is closer physically to the mid, but the point at which there output comes together is somewaht the same. Does anyone a guess how you would apply the CTC distance in this situation? Will there be significant problems with how far the mid and tweeters are apart, or is it more imporatant where and how there output is combined?

[Paul W]

Dan,
Gotta admit I'm still confused about which plot is which, but a couple more thoughts...

If this were mine, I would center the tweeter diffuser on the dome and then tweak the crossover for flat response. This would result in the same response for 360 degrees which may sound a bit more natural. It will probably require only a cap in parallel with a series resistor to accomplish this.

To explain why I'm encouraging large diameter diffusers...at relatively low frequencies/longer wavelengths, driver radiation may tend to bypass or "wrap" around the diffusers. If this is actually a problem, I would first suspect the low end of the RS52. After you get the final layout flat on axis, try extended sessions with pink noise listening for midrange coloration...and use the RTA in Soundeasy with the microphone at a dozen or more locations in the room.

YMMV and all that stuff :W
Paul

[dlneubec]

Sorry for the confusion. Maybe these will clear things up a bit.

As I recall, the response did not look good with the diffuser centered on the domes and didn't look like something that was easy to correct electrically. I'll take some measurments and post them. Maybe with the optimum shaped diffuser and WG it the FR would be better.

[dlneubec]

Here's a a bit about the Dayton neo tweeter in my measurements. It is the ND20FB-4, see:

Dayton ND20FB-4

It has a rising responce at the top end, which I think works well in the upfiring application since it causes a natural roll off at the high end. It is also rated at 93db. I have yet to measure it in the MCM WG, but will, though the throat size of the one I cut down is too big at 1.5" dia. It prbably needs to be closer to 1" for the neo. I may have to cut another one down to give a fair look and try to give it a better throat size match.

Zaph has is in his tweeter reviews below, though he calls it the ND20TB-4. It looks useable up above 3.5khz or so, which might be fine with the RS52.

Zaph Tweeter MishMash

[dlneubec]

I've done some MCM 6.5" Waveguide tests with the Dayton ND20fb 3/4" neo tweeter. The graph below shows the ND20 in the WG (black) no diffuser. The red plot is with the 5.5" concave cone diffuser centered above it.

Looking good, no? Yes!

Also, since the diffuser is centered, as Paul was suggesting, the 360º response should be identical. Unless I'm missing something, it looks like I have a candidate for a tweeter for a crossover from the RS52 from about 3.5khz or above.

I'm also going to run a quick test of a 22TAF/G, which also has a rising response, but I'd be surprised if it looked this good. :T

One qualifier, this is in the WG in a cardboard box, and the WG is not flush mounted.

Feedback anyone?

[dlneubec]

Ok, here is the 22tafg for comparison. The blue line is the 22tafg in WG w/o diffuser. The green line is the 22tafg in WG with diffuser. For reference, the black line is the ND20 in WG w/o diffuser and the red line is the ND20 in WG with diffuser.

The ND20 is a soft dome, cheap (>$5), a little flatter, a little higher SPL and a little better top end. The 22tafg a metal dome, still a good deal at >$27, has a little more extended bottom end and probably has better distortion numbers. BTW, the 22tafg has the hexagrid and my WG needs to be cut to a little larger hole than 1.5" to fit flush. With a better WG to tweeter throat match, maybe it will look a little flatter as well.

Any thoughts?

[Paul W]

Looks like you're ready to make MDF dust :T

[dlneubec]

I think I'm ready to take a partial plunge. :T I still have to determine the optimum spacing between the RS225's, however, which is why I say "partial plunge".

I will probably build the bottom half for a test mule from the plan attached. Then maybe I can cut just the baffle for the section above, mounting it on a cardboard box or something temporary so I can test that spacing. I hope it can be closer than 4.25" shown in the current plans. I think that much spacing is detrimental from an aesthetic point of view. It would also allow the mid and tweeter to be lower and the total hieght to reduce some, all of which would probably be beneficial.

It's kind of a catch 22. I don't want to cut out the pieces for the top box because their dimensions will change should I alter the spacing between the two boxes. However, without a box for the top woofer, it's hard to do a final test. Hopefully, I can get it close enough with a baffle of the correct size and something to close off the back wave of the top woofer. I might also want to test some sort of diffuser between the two RS225's to see if it improves FR.

I'm also in dire need of some thoughts on CTC spacing between the mid and tweeter in this application. I don't know if I should measure flang to flange, which is 3-7/8", or top of WG to top of WG, which is 3-3/16", since this is where the actual output could combine betwen the two. The former suggesting a crossover as high as 3500hz versus 4250hz for the latter (if I figure that correctly :?? ).

Anyone have any thoughts on this question? anyone see other issue I should be concerned about before I start making dust?

Thanks!

[joecarrow]

I think the spacing from mid to tweet *could* work. I'm sorry there's not really a rule of thumb on it, but it looks like they're less than a wavelength (approaching half a wavelength) at crossover. Definitely definitely the right idea- but how can we know it's right till you give it a shot?

One thing I do feel pretty sure of- as long as the tweeter doesn't start getting pushed low enough to start distorting, it should help you acoustically to push the crossover as low as you can.

At 600 hz, I really think that diffusers will have limited effects on the woofer. I could be wrong, but how's that jive with your experience thus far?

Also, depending on how nice you finished them- I'd kind of like the look of those top diffusers exposed. Maybe it's too much character. When I was selecting what to build, my girlfriend said she didn't care how large the boxes were as long as they didn't look "ugly"... and after showing her a lot of projects, it became clear that most of the more creative projects were too much.

[dlneubec]

I think the spacing from mid to tweet *could* work. I'm sorry there's not really a rule of thumb on it, but it looks like they're less than a wavelength (approaching half a wavelength) at crossover. Definitely definitely the right idea- but how can we know it's right till you give it a shot?

Hi Joe,

That may be the only way to know, but a little speculation might be beneficial in case there is something I can do to help the situation with the design.

One thing I do feel pretty sure of- as long as the tweeter doesn't start getting pushed low enough to start distorting, it should help you acoustically to push the crossover as low as you can.

If I use the ND20fb, I'm not sure I can go much lower than 3.5khz with it, since it gets into a much higher distortion area, according to zaph's measurements. The 22tdfc might work, but it would be more difficult to fit inside the RS52's diffuser since it is a lot bigger, has the mesh grill issue and would probably necessitate a taller diffuser. Another one that is a little smaller might be the Vifa XT25SC50 ring raditator. I have read it suggested that ring radtiators should work better with waveguides. The latter two tweeters could be pushed down to maybe 2.5khz LR4, 3khz LR2, according to Zaph. Maybe I should pick up the Vifa to test out. Any thoughts on those options?

At 600 hz, I really think that diffusers will have limited effects on the woofer. I could be wrong, but how's that jive with your experience thus far?

With what I've done so far, I'm not certain any differences will be at all signficant, but it will be easier to tell with the actual baffle size used above and below. What I would probably do is use spacers and test the baffles ar about 3/4" increments apart to zone in on the area that works best and then fine tune. I might also throw a diffuser in there just to see what it does.

Also, depending on how nice you finished them- I'd kind of like the look of those top diffusers exposed. Maybe it's too much character. When I was selecting what to build, my girlfriend said she didn't care how large the boxes were as long as they didn't look "ugly"... and after showing her a lot of projects, it became clear that most of the more creative projects were too much.

Yea, I think I might like it with the top uncovered as well. I don't knowo how I'd build that cowl top and not mess up the sound quality. At this point I'm thinking contrasting woods, maye something light, like Maple on the body with a darker solid hardwood over MDF on the baffles and trim and perhaps on the diffusers or maybe back to solid maple on them. I just haven't thought about it enough yet.

[dlneubec]

Hey Joe,

How about something like this?

[joecarrow]

If I use the ND20fb, I'm not sure I can go much lower than 3.5khz with it, since it gets into a much higher distortion area, according to zaph's measurements.

Agreed. One thing to check out- http://zaphaudio.com/blog.html

Zaph's blog is back! He also just added a measurement for a $9 Aura tweeter that's supposed to be really similar to the $5 dayton neo, but with a titanium dome, slightly better response, and low distortion down to 3khz. This is surely worth a look.

Also, regarding the aesthetics- I think you're on to something... but I'm an engineer, not an artist.

I think you're really onto something now; and I can't help but mention that it's starting to resemble that Beolab 5 again 8) . It has a tweeter with diffuser above a dome with diffuser... I only keep bringing that one up because I was really impressed with its sound the time I heard it.

The other part of your design that still worries me is the distance between the woofers and the mid. As always, closer is better- but the distance seems to be safely less than a wavelength at crossover. It should be fine- it's not that far, and 600 hz is kind of on the low side.

I wish you the best of luck.

[dlneubec]

Agreed. One thing to check out- http://zaphaudio.com/blog.html

Zaph's blog is back! He also just added a measurement for a $9 Aura tweeter that's supposed to be really similar to the $5 dayton neo, but with a titanium dome, slightly better response, and low distortion down to 3khz. This is surely worth a look.

Also, regarding the aesthetics- I think you're on to something... but I'm an engineer, not an artist.

I think you're really onto something now; and I can't help but mention that it's starting to resemble that Beolab 5 again 8) . It has a tweeter with diffuser above a dome with diffuser... I only keep bringing that one up because I was really impressed with its sound the time I heard it.

The other part of your design that still worries me is the distance between the woofers and the mid. As always, closer is better- but the distance seems to be safely less than a wavelength at crossover. It should be fine- it's not that far, and 600 hz is kind of on the low side.

I wish you the best of luck.

I did see that Aura driver. The concern I have with it is that it has a decreasing responce from 10k-20khz, rather than an inclining responce, which seems to be needed in this upfiring format to get a flat FR. The ND20 and 22tafg have that, not many others I've found do. It's really hard to say what it would look like at 90º off axis without trying it and $9 is well worth the risk for the small form factor, lower crossover and distortion, so I'll definitely give it a shot.

Hey, even engineers have taste (I have a couple partners who are civil engineers and others who work for us). :T Something about form following function..... :W

It's sort of ending up like a blend of the concepts of the Beloab 5 and the Duevel Jupiter. I hope I can work it all out.

If I can reduce the woofer to woofer distance, that can only help with the mid to woofer distance, I would think. As the top box moves down, it becomes wider to match the slope of the bottom box, so it therefore it can get shorter and maintain the same volume. The result is a closer mid to woofer distance from both the top woofer and bottom woofer. All positives I would think.

What would you guess would be the CTC distance I should target, the one from VC to VC or from baffle/waveguide edge to baffle/waveguide edge? With the mid to woofer distance, that makes about a 2" difference in distance.

[joecarrow]

I wouldn't worry about finding the precise CTC distance; it's not like you can really account for it in the crossover or anywhere else. You just want to minimize it.

Have you considered, or ever used, aperiodic enclosures? The Dayton Reference drivers require a large box, and since you're looking at getting your deep bass from a subwoofer, an aperiodic enclosure might allow you to use a smaller box and more easily reduce your woofer to mid distance.

[dlneubec]

I wouldn't worry about finding the precise CTC distance; it's not like you can really account for it in the crossover or anywhere else. You just want to minimize it.

Have you considered, or ever used, aperiodic enclosures? The Dayton Reference drivers require a large box, and since you're looking at getting your deep bass from a subwoofer, an aperiodic enclosure might allow you to use a smaller box and more easily reduce your woofer to mid distance.

I have not considered aperiodic enclosures. I don't know a thing about them, how they work or if there are any problems with them.

BTW, I did order two more neo tweeters to try out.
Aurasound NT1-204-8D 3/4" Titanium
and
Seas 27AFNC/G (H1397) 1" aluminum

[stinems]

Dan,
I've been watching this thread with great interest.

A speculation on the CTC distance issue: in this scenario, could it be the vertical dispersion of the diffusor/WG that determines proper CTC distances?

I just remember reading in the Sausalito (www.sawonline.com) white papers that his acoustic lens (the Beolab 5 diffusor) exhibits a much narrower dispersion field vertically off axis as a result of greatly widening the response horizontally off axis. Remember this is the exact tradeoff that was desired.

Perhaps I don't understand the true nature of what makes drivers sound "integrated", but my intuition I says that if the vertical dispersion fields between drivers significantly overlap, you'll get two drivers that pretty much sound like a single point source.

I suppose there are a couple ways you can manipulate this verticle dispersion field overlap: CTC , size of the horizontal plane baffles (ie, "top" mid/tweet baffles), positioning of the duevel-esque diffusor...

I'm no expert, so its likely that my intuition is all wet. Am I making any sense? Any thoughts?

~Sam

[dlneubec]

Dan,
I've been watching this thread with great interest.

A speculation on the CTC distance issue: in this scenario, could it be the vertical dispersion of the diffusor/WG that determines proper CTC distances?

I just remember reading in the Sausalito (www.sawonline.com) white papers that his acoustic lens (the Beolab 5 diffusor) exhibits a much narrower dispersion field vertically off axis as a result of greatly widening the response horizontally off axis. Remember this is the exact tradeoff that was desired.

Perhaps I don't understand the true nature of what makes drivers sound "integrated", but my intuition I says that if the vertical dispersion fields between drivers significantly overlap, you'll get two drivers that pretty much sound like a single point source.

I suppose there are a couple ways you can manipulate this verticle dispersion field overlap: CTC , size of the horizontal plane baffles (ie, "top" mid/tweet baffles), positioning of the duevel-esque diffusor...

I'm no expert, so its likely that my intuition is all wet. Am I making any sense? Any thoughts?

~Sam

Hi Sam,

I wish I understood more about what makes drivers sound integrated in relation to CTC distance, dispersion, etc. That's why I've been trying to solicit advice and/or speculation, and getting little responce, I must say.

Besides the wavelentgh issue, there is normaly a power response issue where one driver is sort of beaming and the driver it crosses over to is more omni in nature because of the longer wavelengths in the lower frequecny range of it output. In this case, since the drivers are being used omnidirectionally throughout their range, I would think power response would the same from driver to driver, so it makes sense to me that this would help make drivers sound more integrated. Could this make it more forgiving from a CTC standpoint?

I agree that the WG/diffuser combnation will definitely have a strong impact on vertical dispersion in favor of a cylindrical shaped wave front that moves out horizontally from the speaker. I have no idea if that hurts or helps the integration of the drivers from a CTC standpoint. My hope would also be that by limiting the vertical dispersion, there would be less concern with floor and ceiling bounce issues.

I'll have to do some vertical measurements with the mid and tweeter with WG's and diffusers. I should take a look at the white paper. Maybe there are some answers there.

Thanks for the feedback. I wish I could get more participation from the resident "experts". :cry: I guess this is just not an intersting project for them.

[Dennis H]

Thanks for the feedback. I wish I could get more participation from the resident "experts". I guess this is just not an intersting project for them.

Well, I'm no expert but I find your project very interesting. I just don't feel like I have anything to contribute because I don't have any hands-on experience with what you're doing so, in a sense, you are the resident expert and we're all learning from your experiments. Keep up the good work.

[dlneubec]

so, in a sense, you are the resident expert and we're all learning from your experiments. Keep up the good work.

Now that's a scary prospect! :#

Thanks for your help Dennis. You've been a welcome contributor on this and the first omni project. Ok so here's a job for you. Every so often just make a post saying you think I'm headed in the right direction. It will make me feel better! :W

BTW, In looking at those SAW white papers, they do make me feel better about what I'm trying to accomplish. One of the biggest differences I see is that they were focused on creating an acoustic lens that would focus the output into approximately a 180º horizontal dispersion (30º vertical) from the speaker forward, rather than in 360º horizontal as I'm trying to do here. I wonder if the 180º dispersion was truly the original design goal or if the goals are tailored somewhat to the solution they came up with? Especially since in one of the papers, they talk about the ultimate speaker being a pulsating sphere that radiates equally in all directions.

Being in a design profession myself, I know that sometimes goals stated after the design is complete are not always the same as the ones you started with. :B

[joecarrow]

Something to consider- center-to-center spacing is about more than driver integration and coherence of the soundstage. I think the thing that would worry me most about excessive center to center spacing is the constructive and destructive interference patterns causing an uneven frequency response that's dependent on listening position. It's why why can't have two tweeters on the front face- not because your ear can pinpoint within an inch exactly where the sound is coming from- but because they cause comb filtering and crazy lobing.

Also, I totally hear what you're saying about SAW changing their design goals after the fact. You're not supposed to do it, but if you don't sit down to write the report until after the fact, who's really going to know? I think it's within the bounds of reasonable fudging, and I wouldn't really fault them for it if they believed that their end product was of high quality.

Supposing that they made a conscious decision to have that radiation pattern, I can see a big reason for the choice. By restricting themselves to the most audible portion of the spherical radiation, they've increased system efficiency.

One thing I noticed too late- that Aurasound titanium tweeter I suggested actually has significantly lower sensitivity than the cloth dome Dayton. Sorry I didn't catch that sooner- it may or may not be an issue for you.

[Paul W]

Dan,
I like the new styling really well except for the straight back. The look might be more balanced with the same slope on all four cabinet sides. You could still maintain the drivers in an arc relative to listening position.

As far as the mid and tweeter are concerned, the apparent acoustic source in the listening plane is likely the diffuser...so I believe the proposed mid/tweet "CTC" is fine for any reasonable xo frequency. Since the diffuser signal actually originates at the driver, I'll hypothesize a virtual acoustic source an inch or two behind the diffuser, regardless of direction in the 360 listening plane. (That sentence should generate some comments...hopefully useful to you :lol: ) It may be worthwhile to take some vertical on/off-axis measurements because the midrange diffuser is at/under 1/4 wavelength radius on the low end...possibly causing power response discontinuity.

For the woofers, you could change to a 3.5 with one of the 8" just below the RS52, the .5 woofer downfiring. That should clear any midrange CTC questions, but would push the very low end further from the mid. Maybe look at a floorbounce calculator before you decide.
Paul

[dlneubec]

Something to consider- center-to-center spacing is about more than driver integration and coherence of the soundstage. I think the thing that would worry me most about excessive center to center spacing is the constructive and destructive interference patterns causing an uneven frequency response that's dependent on listening position. It's why why can't have two tweeters on the front face- not because your ear can pinpoint within an inch exactly where the sound is coming from- but because they cause comb filtering and crazy lobing.

Hi Joe,

Yup, that's the thing I'm worried about. However, if you look at the scale of that dome mid on the Beloab (assuming it is at least a 2") the tweeter has to be at least 5" above that, just scaling the photo, so my application is much closer than theirs either VC to VC or WG edge to WG edge. I forget where they said they are crossing the mid to tweeter at. Also, I get the impression that since they claim it to be essentially flat to 15khz or so, that it has the top end roll off problem that seems to be somewhat corrected with the use of the WG/Diffuser combo. These things encourage me a little.

The Duevel's have their woofer/mid and mid/tweeter, very far from one another, but where the WG/diffuser output comes togeher is immediuately adjacent. I could mimc this design by putting the tweeter and WG on top with the driver firing down into the diffuser, but at this point I prefer to keep the drivers closer together.
Duevel Bella Luna

The problem will come when I have to choose crossover targets. Remember, I'e only done one two way crossover design and no 3ways, so when I get to that stage, I will need lots of help or there will be an awful lot of trial and error.

One thing I noticed too late- that Aurasound titanium tweeter I suggested actually has significantly lower sensitivity than the cloth dome Dayton. Sorry I didn't catch that sooner- it may or may not be an issue for you.

No problem. I noticed that, but decided that for $9, I'd give it a try. It would be nice to have some tweeter options that could be crossed lower, if need be.

[dlneubec]

Dan,
I like the new styling really well except for the straight back. The look might be more balanced with the same slope on all four cabinet sides. You could still maintain the drivers in an arc relative to listening position.

Hi Paul,

I agree withg you from an aesthetics standpoint. However, to maintain the RS225 enclosure volume at about 25L, for each box, I would have to make the base/footprint that much wider , or raise the height of the mid and tweeter. The straight back was my compromize solution to add box volume, while controlling footprint size, front width and mid and tweeter height. If I were making them with RS180's or something that I could comfortably do a much smaller box with, I would probably have tapered all sides equally, or perhaps made them all parallel.

BTW, as a side note, I've been working on the plans for the construction of the bottom box so I can do some better woofer testing and woofer spacing testing between the upper and lower boxes. These angles are going to be a royal pain in the behind. For example, in front elevation view, the sides are drawn at a 5º slope and the bottom box height is 12". Well that is based on a vertical dimension only. Since the front is sloped 10º, the actual length of that piece is more like 12 5/16". That also means that I can't cut that piece at a 5º slope, since it is covering 5/16" more length than the drawn slope and that the back, sides and front are all at different angles and lengths!. Anyway, having one straight side, at least this time around may be the only thing that keep me sane. :W

As far as the mid and tweeter are concerned, the apparent acoustic source in the listening plane is likely the diffuser...so I believe the proposed mid/tweet "CTC" is fine for any reasonable xo frequency. Since the diffuser signal actually originates at the driver, I'll hypothesize a virtual acoustic source an inch or two behind the diffuser, regardless of direction in the 360 listening plane. (That sentence should generate some comments...hopefully useful to you :lol: ) It may be worthwhile to take some vertical on/off-axis measurements because the midrange diffuser is at/under 1/4 wavelength radius on the low end...possibly causing power response discontinuity.

I'm glad to hear you hypothesize this. It does seem like the normal driver center to criver center standard might not apply here. I will definitely take some off axis vertical measurements in the process, one thing I neglected to do on the first omni project.

Thanks for climbing out on the limb with me. :B

For the woofers, you could change to a 3.5 with one of the 8" just below the RS52, the .5 woofer downfiring. That should clear any midrange CTC questions, but would push the very low end further from the mid. Maybe look at a floorbounce calculator before you decide.
Paul

I'm not sure if I'm following you here. Is there a way to do what you are suggesting and maintain the 25l box size for each woofer?

BTW, here are a couple of BoxyCad2 screens showing the mid to woofer floor bounce effect and recommended crossover point. The woofer heights are assumed to be directly in between the two woofers. The first plot is a plan with 4.25" between the woofer baffles and the next is with 3" between the woofer baffles. The midrange height is adjusted down accordingly in the second one. They both range in the low 600hz range for recommended crossover point, which is right in the area I have been thinking about, 600hz-700hz.

Any reactions based on these?

[stinems]

Something to consider- center-to-center spacing is about more than driver integration and coherence of the soundstage. I think the thing that would worry me most about excessive center to center spacing is the constructive and destructive interference patterns causing an uneven frequency response that's dependent on listening position. It's why why can't have two tweeters on the front face- not because your ear can pinpoint within an inch exactly where the sound is coming from- but because they cause comb filtering and crazy lobing.

I thought the whole point with these horizontal diffusors is that you don't get (or get very little) lobing due to interference in the first place. "...uneven frequency response that's dependent upon listening position." Let me clarify, there will still be lobing with these diffusors, but only if your listening position moves vertically, ie, closer to the floor or ceiling. Moving from the middle of your listening couch to the far end should impart no change in sound due to lobing/interference (moving horizontally).

In this scenario, my mind's eye envisions two drivers in phase at the crossover point, aligned on their conventional axes (which would be pointing floor to ceiling), both sounding into the same "cylinder" of diffusion. The listener is looking normal to that conventional axis and radially to the "cylinder of diffusion". Since the drivers are in-phase and the speed of sound is constant for all frequencies for a given medium, it won't matter where you sit in the horizontal plane, there should be minimal lobing. This is to say, there is no single horizontal listening axis here, only one big happy horizontal listening "plane".


~Sam

[dlneubec]

Hi Sam,

I think what you are saying with respect to the lack off horizontal lobing is supported by the claims that SAW makes in regards to it's acoustic lens and the BeoLab 5, with one main difference being that they were talking about over 180º rather than 360º.

As you suggest, vertical lobing on the other hand is more of a question mark. Personally, if the narrowed vertical dispersion reduces floor and ceiling bounce, that's a good thing. If, as a result, it ends up with a fairly narrow vertical sweet spot, that is not of concern to me as long as the sweet spot ends up centered at the normal seated listening postion. :T

[Paul W]

Thanks for climbing out on the limb with me.

No problem, I'm hoping my somewhat loopy statement about a virtual source will generate some conversation that might lead to the "correct" shape for the diffusers.

I'm not sure if I'm following you here. Is there a way to do what you are suggesting and maintain the 25l box size for each woofer?

What I'm suggesting is to combine the boxes into a single 50 liter, with one woofer upfiring into a diffuser on the backside of the RS52, the second woofer (the .5) downfiring into the floor. (Might also facilitate the equal angled side aesthetic.)

[dlneubec]

No problem, I'm hoping my somewhat loopy statement about a virtual source will generate some conversation that might lead to the "correct" shape for the diffusers.

It looks like they aren't listening

What I'm suggesting is to combine the boxes into a single 50 liter, with one woofer upfiring into a diffuser on the backside of the RS52, the second woofer (the .5) downfiring into the floor. (Might also facilitate the equal angled side aesthetic.)

Ok, I see what you mean. I may take a look at that later, since I already started on the bottom box section last night. I have most of the cutting done, but want to add some cleats that will allow me to temporarily screw it together for testing, rather than doing any final glue up at this point.

It looks like I did get the angles pretty well figured out as it seems to fit together fairly well. I hope to get it together this weekend in enough time to do some woofer testing, etc.

I should be getting the Aura and Seas tweeters tonight so I can test those with the waveguide/diffuser setup to see how they compare with the ND20 and 22tafg.

[dlneubec]

I ran a few more tweeter tests this evening. First I tested the Aura neo and 27tafnc/g neo tweeters linked previously. They were first tested in the 6.5" WG, with the 5.5" diffuser from the previous tests. The FR with the Aura in this application was bad enough that I did not save any of the plots.

The 27tafnc/g wasn't too bad, close to as good as the 22tafg. The first plot below compares the latter (in blue) to the former (in black). The pink is the 27tafnc/g in the WG only, not diffuser. It might be usable, if I was forced to cross lower than the ND20 requires.

The next plot is only the ND20. The pink line is the ND20 using the 6.5"WG and 5.5" diffuser. The black line is the same WG, but with an 8" dia diffuser. The larger radius diffuser appears to flatten and smooth the response, raise it by a couple db and even pushed the low end response a bit lower.

Now I wanted to test the ND20/wg/8"diffuser combo off axis. The next plot is the same as the last, but adds a 90º axis plot to the mix (in blue). How's that for a 90º off axis response? :T

I've come to the conclusion that the ND20 is the right tweeter for this project , unless I find I need to cross lower than the 3khz to 4khz range from the RS52. Anyone have any argument against this selection?

Finally, there is a plot of the ND20/WG/8"diffuser combo (blue) with the plot of the RS52/8"WG/diffuser combo (black) for comparison of the two. There is a lot of overlap here, with the RS52 running about 3db below the ND20 in the 3-4.5khz range.

What do you all think of crossing these two in that range range? Any suggestions as to what to target initially? BTW, I'm thinking the the RS225x2 to the RS52 will be crossed in the 600-700hz range. Also, as noted before, floor bounce between the mid and woofers seems to indicate the optimum crossover to be in the low to mid 600's. If I go with 600hz, then 3.6khz is 2.5 octaves for the mid between crossover points. Is that enough for a 3way? I've read that 3 octaves is a good target. Does that include some of the response above and below the crossover points or is that 3 octave between the actual crossover points?

Of course final decisions will have to wait until I get better woofer measurements and get a box built where all the drivers can be more accurately measured in their planned configuration, but I'd like to have a set of targets in mind.

[dlneubec]

Hi All,

I got started in a prototype last night and had all the parts for the lower box cut out. Then at the last minute I decided to take the section out of the bottom that was intended to mount the crossover in. This allowed me to make the overall box smaller and I changed from a 5deg slope on the sides and 10 on the front to a 4/8 degree combo. As a result, I had to redesign and cut down all the parts this morning (and recut a bezilliion angles!), etc. :E

I sort of started at the RS52 in WG since that demands a certain width and adjusted down from there. It made the footprint smaller and I was still able to keep over 25L in the boxes.

I wanted to build just the basic bottom box (no bracing, no trim, etc.) to test the woofer spacing and FR. I cut the wood baffle for the top box and routed it out for the RS225, but intended to stop there, because I was afraid a change of heights would cause me to recut all the top stuff over again. More about that later.

Attached is a shot of the bottom box with the RS225 in it, bottom box baffle has been chamfered. The uprights are in place and that is the upper box baffle (un-chamfered) sitting next to it. The next pic is with the top box baffle sitting on top. Those are 1" dowels for spacers. I decided to make them larger than the 3/4" I used in the first omni to give them a more substantial look.

Later I'll post some more pics and measurements for review.

[dlneubec]

After the basic bottom box was done, I needed to do some woofer measurements. For the top box, I mounted the second RS225 in the top box baffle I cut out and then taped a cardboard box on top of it, just to try and get some idea of the issues.

I had a ton of problems getting good measurements. First, with only 3" of spacing between the baffle there was a definite cavity resonance, I assume from the two cones. There was a very definite peak in nearfield responce that attenuated as the baffle were separated more. Also, I could not get a very good ff measurement with the box on the floor. I tried at 1m with the mic at 15" (centered between the woofers) and had a very deep valley, I assume from floor bounce. I tried putting the mic on the floor and that was much better. Finally I decided to raise the box up off the floor and take a normal 1m gated farfield measurement and then doa nearfield at that height and combine them. The first picture attached shows the box up off the floor, with the mic height at 41" (this was the nearfield measurement).

When I got the boxes so that the baffles were spaced about 4.25" apart, the cavity resonance was much better under control. Then I decided to try adding a baffle between them to keep the interaction between the two cones at even more a minimum. You can see that in the first pic and in more detail in the second. It is two pieces of 3/4" mdf, about 7"x8.5", chamfered about 3" in from all sides. This really improved the FR.

The first graph below is the FR of the two RS225's with the baffle separated by 4.25" and the baffle mentioned in between. This is a nearfield and farfield merged at about 280hz. BTW, for the farfield, I was actually able to increase the gated window to 8ms in order to make the response good down a little lower. The separation baffle is just sitting on 2 little blocks of MDF. The top baffle is not chamfered, and the top RS225 is in a cardboard box, not sealed. The bottom box is not sealed either at this point. Still, the responce looks pretty good and it should only get better in well sealed boxes with both baffle chamfered (and no little blocks in the way). Having the woofers firing into a baffle actually increases their SPL, just like it did with the RS52 and ND20 firing into their diffusers.

The second graph shows the RS225 with the RS52 and ND20. There looks to be lots of overlap and the spl's of the various drivers match up pretty well.

This looks like enough to go ahead and build the top box with this spacing so I can get some better tests on the woofers.

Question: Does anyone see any problems with the way I measured the woofers (up about 41", ff @ 1m and NF @ about 6", which is the front edge of the upper and lower box baffles?

Any thoughts on crossover points in regards to these results? I'm starting to think I should maybe go a little higher, like 700hz and 4800hz. This should allow me to push the RS52 a bit and play them a little louder before it reaches xmax.

Feedback, please! :sos:

[dlneubec]

I have some additional progress to report. :T

I finished up the rough out of the top box. Attached are some photo's of the process. The first one shows the top baffle construction housing the RS52 and PE 8" waveguide. There is an outer baffle that the WG is screwed into. This baffle will remain removable. There is an inner backer to the baffle the is supported by cleats in the inside of the box. Then there is a circular piece that the RS52 is mounted in. It has been recessed to allow for proper mating with the WG. I will install some speaker gasket to seal around the RS52 from the inside of the box. I might put some around the flange of the RS52 under the WG for a good seal here also. The second photo shows them staked up and the third with it flipped over.

The next photo is of the assembled top box from the bottom, where the second RS225 (downfiring one) will be installed. You can see the shelf braces and the RS52 on the other end. The next photo is with the RS225 added. The holes in the RS225 baffle line up with the dowels in the bottom box so they can be stacked together. The length of the dowels determine the spacing. This allows them to easily come apart and be moved around.

[dlneubec]

Here is two more photos. The first is of the top box assembled from the top side. The next is one of the boxes stacked. I did not line up the dowels for this shot, I just set the top box on top (you can see one dowel is a bit askew ).

I have not cut any of the chamfers on the RS225 baffle or the RS52/WG baffle as of yet. I will do that next and mount the little diffuser between the two RS225's. Then I need to start working on the top diffuser with ND20 and waveguide set inside of it. I'm not sure yet how easy that is going to be to do.

[dlneubec]

I have a little more progress to report in the form of photo's.

the first photo shows the wave guide(WG)/diffuser stack including the RS52 in the PE 8"WG, with the concave cone diffuser over it. Inside the concave cone diffuser is the ND20 3/4" neo tweeter inside the MCM 6.5"WG. The height of the Diffuser/WG/Tweeter sectino is not set until more measurements are taken, but you can get the idea of how it will look with the aluminum tubing shown on the bottom section as spacers.

The next photo shows a concave cone diffuser added over the ND20 in WG. The aluminum tubing will be added to this section also, when final height is determined.

The final photo is an overall view. You can see the baffle added between the woofers as discussed previously. It is my plan to enclose the open middle section of the speaker with some sort of speaker cloth/wire/screen mesh around at least 3 sides. I still have to complete the design of the base for the speaker, which is up in the air a bit at the moment.

This weekend, I may go ahead and start gluing some of the permanent parts together, so it does not have to be held by clamps. :B Or, I might do a little more FR testing to determine diffuser heights, etc. I may need to go ahead and add some binding posts to the back panel so I can get to each of the woofers for a good sealed test.

I'm still unsure of the best method to employ for upcoming design measurements. More on that later.

Feedback is welcome! :T

[JonP]

Very cool indeed, Dan!

If you make a square frame of grill cloth to hide the diffusors, (assuming you'ld want to) they would bear a rather strong resemblance to the classic Ohm speakers.

Kind of a roundabout, >20yr ago ommnis to a new design with a similar look today...

Will be interesting to see more measurments... I'd guess you'd have to get a ways back for the three segments to combine and give a decent full range response. And, of course, the room will enter the picture in the low end. Nearfields or sorta nearfields, with the mic right at the diffusor mouths would be easy. Would be interesting but tedious, to map vertical polar response of each section alone... and see what kind of pattern they throw.

Great stuff... lots of progress!! I've been hardly able to get on here and post, not to mention make some progress on projects...

[dlneubec]

Hi Jon,

Good to hear from you. :T

I'm undecided about covering the top diffusers. I'm leaning toward not, but that could change. They would certainly look like the old omni Ohm's with them covered. I'll have some photo's to post later this morning that shows more of what I'm thinking to finish the top section, plus some construction photo's.

I do intend to try and cover the central open section. I will be testing a round diffuser between the woofers as well to see how that compares to the rectangular ones. The round ones would look better and fit with the top section aesthetically.

I'm not sure how I need to do the measurements. The design is setup so that the drivers VC's and acoustic centers of all the drivers will be coincident at the listening position, but that is about 11.5' @ 37". Measuring from that distance won't work and I will need to meausre them closer than that. If I were to se the typical 1m gated measurement, that will mean that the drivers will not be coincident during measurements, even thought they are at the listneing position. Also, the woofers, being about 15"' from the floor, show significant cancellation due to floor bounce at 1m at tweeter height of 38". The best results I got with the woofers so far, were to lift them up to 38" and measure at 1m. I don't know if this is valid or not, however. I tried a ground plane at 1m, which was second best to the 38" up measurement. So far, all measurments have been done inside.

I do plan to do some vertical polar measurements. I just have to figure out how to set that up.

Maybe nearfield is enough on the woofers. I could do a gated farfield of some sort to establish the spl level and phase for matching correctly with the mid and tweeter, and simply merge in the nearfield which would cover most of the useable FR. Since the drivers are all playing in 4pi space over their entire range, as I understand it, there is no need to do any baffle step compensation.