The Mentor II: omnidirectional project #2

**Scottg** · 01 March 2007, 16:56 Thursday

Your Mentor 1 was a MUCH better design..

The slanted baffle and in particular the dual radial driver approach is not a good idea. (think of the side combing it will generate.)

To get "around" the problems of a an altered response depending on the driver's axis you will either need to use a small diameter driver OR lower the crossover point of the tweeter (..or both).

My suggestion within your basic parameters:

Radial driver (upward firing single driver) = Dayton RS52AN (small diameter DOME insuring no dispersion problems to as high as 2 kHz), good performance, good price. (..also it should be able to maintain spl quite a bit better than a cone driver AND it has a small acoustic center which means that imaging should be more precise.)

Tweeter.. we want physical offset BUT we have a problem with aesthetics for a tweeter that resides ABOVE the radial driver. Additionally we have a problem with using a tweeter at lower freq.s because of the increased excursion. Solution: Use Zaph's Seas 27TDFC in the modified waveguide below the tweeter. If you want you can simply add another tweeter on the rear of the baffle in the same position to achieve a dipole character (..reversing the phase for that driver of course).

Zaph|Audio

http://www.zaphaudio.com/Waveguidetmm.html

For the midbasses - not terribly critical.. i.e. there are lots of choices here.

The baffle should have some "tilt" to it - say 5-10 degrees.. and for the sake of aesthetics (and non parallel sides), I'd make all the side panels have that same "rake" (rather like a small obelisk but without the pyramid on top). This commercial design has the tweeter (though non-waveguide loaded) at the bottom:

6moons audio reviews: FJ Om Revisited

http://www.6moons.com/audioreviews/fj2/om.html

**joecarrow** · 01 March 2007, 17:06 Thursday

Just a few miscellaneous comments-

The idea of front firing woofers for bass punch- well, most of that bass punch is between 50 and 300 hz. It's all about having good response in that area, and orientation of the woofer will matter most if it's large enough not to be omnidirectional in that range.

Also, believe that the reason that some of the well respected midwoofers have poor off-axis response is because they behave very pistonically. Youll notice that the 830860 has good off-axis performance, and it's larger than the L12. The 830860 is poly, and the L12 is aluminum.

If you want pistonic operation and good dispersion, you just need a small driver. I think that a 2" or 3" dome is probably your best bet. If you had the RS225 firing upward from the end of a pipe or tube, and the RS52 firing downard concentrically with it, you'd just need to position the tweeter carefully to get good interaction and minimize diffraction issues.

You could also have the RS225 front-firing, wit the RS52 firing upward on top of the box. I wouldn't say that this would necessarily have better punch in the bass, but it would greatly simplify the diffraction and cavity resonance issues.

**dlneubec** · 01 March 2007, 18:54 Thursday

Originally posted by Scottg

Your Mentor 1 was a MUCH better design..

The slanted baffle and in particular the dual radial driver approach is not a good idea. (think of the side combing it will generate.)

To get "around" the problems of a an altered response depending on the driver's axis you will either need to use a small diameter driver OR lower the crossover point of the tweeter (..or both).

Hi Scott,

Thanks for the feedback.

I'm not sure I understand why you think the sloped baffle is a bad idea. This is what Mirage is doing on their new OMD-28 speaker.

FWIW, Absolute Sound did a review of the OMD-28 and gave it very high praise. Acording to the review, Mirage claims that the 15º slope directs the rear wave more upward on the back wall, allowing the front output to open more fully toward the listening area and also improves imaging. Accorging to Mirage, this puts about 30% of the drivers output toward the listening area, with about 70% off side and rear wall surfaces, which they also claim is a ratio which closely mimics the characteristics of live music.
Mirage OMD-28

Originally posted by Scottg

My suggestion within your basic parameters:

Radial driver (upward firing single driver) = Dayton RS52AN (small diameter DOME insuring no dispersion problems to as high as 2 kHz), good performance, good price. (..also it should be able to maintain spl quite a bit better than a cone driver AND it has a small acoustic center which means that imaging should be more precise.)

I had thought about the RS52. My concerns were that at 91db, if it had the 6db or so losses I saw with my first omni project, on both the RS180's and the RS28(upward firing), it would perform at about 85db at best on the listening axis. Also, since I want at least as much output as I have with the current design, I had a hard time imagining how 1 RS52 with it's 2" dome could put out the kind of sound that 2 RS-180's are doing now, albeit over a larger range (100hz to 1450hz crossovers). I have an RS52, so if I get to the prototype phase, I will most certainly give a try. The measurements will be the best way to tell. B&O is certainly using an upfiring midrange in its BeoLab omni, but it does have a diffuser/difflector that redirects the sound into a 180º front wave, according to the creators claims.

Originally posted by Scottg

Tweeter.. we want physical offset BUT we have a problem with aesthetics for a tweeter that resides ABOVE the radial driver. Additionally we have a problem with using a tweeter at lower freq.s because of the increased excursion. Solution: Use Zaph's Seas 27TDFC in the modified waveguide below the tweeter. If you want you can simply add another tweeter on the rear of the baffle in the same position to achieve a dipole character (..reversing the phase for that driver of course).

http://www.zaphaudio.com/Waveguidetmm.html

I'm not following what you are suggesting on the wave guide. Can you explain further? Maybe, you talking about a downfiring tweeter with a wave guide? If so, I've been there and done that on the original and diffusers simply did not work in my many attempts.

Originally posted by Scottg

For the midbasses - not terribly critical.. i.e. there are lots of choices here.

The baffle should have some "tilt" to it - say 5-10 degrees.. and for the sake of aesthetics (and non parallel sides), I'd make all the side panels have that same "rake" (rather like a small obelisk but without the pyramid on top). This commercial design has the tweeter (though non-waveguide loaded) at the bottom:
http://www.6moons.com/audioreviews/fj2/om.html

I disagree with the mid not being critical. It seems to me that it determines where the woofers need to cross, what tweeters are possible to use, what crossovers are feasible, etc., not to mention that it carries most of the sound you hear.

Those speakers are interesting looking. How do you suppose they deal with the innefficiency of one up firing woofer? I wanted to avoid needing to incorporate delay in my passive network. With the tweeter on the front like that, you would have to include a delaty network, would you not. This was one of the main problems I could forsee with the two mid idea, especially the front to rear arrangement.

What did you think of the idea of using one high efficiency Pro driver? I believe that is what Duevel does with their omni's, in addition to using a compression driver on the top end that is omnidirectionally horn loaded.

**augerpro** · 01 March 2007, 19:06 Thursday

Dan check out the Fostex drivers at madisound for a pro mid. Very nice for the price.

**dlneubec** · 01 March 2007, 19:27 Thursday

Originally posted by joecarrow

Just a few miscellaneous comments-

The idea of front firing woofers for bass punch- well, most of that bass punch is between 50 and 300 hz. It's all about having good response in that area, and orientation of the woofer will matter most if it's large enough not to be omnidirectional in that range.

Hi Joe,

From some of the feedback I've had on the original Mentor, all up or downfiring drivers seems to be looked at with a doubtful eye. Also, it is hard to incorporate two bass drivers in an up or downfiring format, not so with front firing. If you go to one driver downfiring, you almost have to go to a sub to get good output and keep the box reasonably sized. My thinking was to increase the omni range in the mids, while decreasing it bass. Still, the bass is already omni in nature up to 100hz or so, so I have decreased the omni range from 100 to 500 or so. My hope was with the midrange extended from 1450 to maybe 2500 or 3000, that would more than offset the omni loss down low. In addition, if the crossover slopes could be less steep, the coverage is actually increased beyond that.

Originally posted by joecarrow

Also, believe that the reason that some of the well respected midwoofers have poor off-axis response is because they behave very pistonically. Youll notice that the 830860 has good off-axis performance, and it's larger than the L12. The 830860 is poly, and the L12 is aluminum.

So, are you saying that the 830860 is a poor solution and that the L12 would be better?

Originally posted by joecarrow

If you want pistonic operation and good dispersion, you just need a small driver. I think that a 2" or 3" dome is probably your best bet. If you had the RS225 firing upward from the end of a pipe or tube, and the RS52 firing downard concentrically with it, you'd just need to position the tweeter carefully to get good interaction and minimize diffraction issues.

Ok, that's a second for trying the RS52. I will definitely test it when the time comes. It would certainly be a simpler solution. What do you say about my concerns expressed above about using it? Can we replace the output of two RS-180's through the midrange with one RS52? I was hoping to avoid the top box concept used in the original project.

Based on my measurements from the first project, I just don't see how I could possibly get enough output from one woofer and one mid, in an up or down firing arrangement, to equal the output I have with the current design. It would seem like you would have to have a down firing RS225 at one end of the tube, and up firing at the other, then a maybe an up and down firing RS52 above the tube with the tweeter in between, kind of like it is in the current design.

How about an up and downfiring RS225, in a 3.5way configuration with the bottom woofer a fill in woofer? That would make them omni their entire range, but there still needs to be a solution to the mid driver issues. Sounds like a very difficult crossover design!

Originally posted by joecarrow

You could also have the RS225 front-firing, wit the RS52 firing upward on top of the box. I wouldn't say that this would necessarily have better punch in the bass, but it would greatly simplify the diffraction and cavity resonance issues.

Yes, it would. I'll be skeptical about the overall efficiency until I can do some testing.

**dlneubec** · 01 March 2007, 19:28 Thursday

Originally posted by augerpro

Dan check out the Fostex drivers at madisound for a pro mid. Very nice for the price.

Ok, I'll take a look at them. Thanks.

**Dennis H** · 01 March 2007, 19:33 Thursday

Hey Dan,

A few thoughts about omni part deux.....

An upfiring mid seems fine as you've had good luck with that. I agree with the earlier poster that 2 mids may create more problems than they solve.

If you want to cross to the woofers at 4-500 (should be fine with a single small-cone mid), I'd fire one woofer forward and one back to keep the omni pattern.

Likewise with the tweeters, I'd experiment with two firing front and back in phase to keep the omni radiation pattern. A dipole tweet will put the tweeter's power response down 5dB compared to the mid and woofers and a horn/waveguide will make it even worse.

**dlneubec** · 01 March 2007, 19:46 Thursday

Originally posted by augerpro

Dan check out the Fostex drivers at madisound for a pro mid. Very nice for the price.

Hey Augerpro,

I ws just looking at them, I didn't see anything over 90bd sensitivity, and I expect to need at least 94db.

BTW, I did see the Audax PR170MO 6.5", which it says has a 100db sensitivity and looks like it might work from 500 to maybe 1800-2000hz, based on the 60º off axis responce. It looks like it would take a steep slope on the top end however.

Anyone know anyting about this driver?

**dlneubec** · 01 March 2007, 20:07 Thursday

Originally posted by Dennis H

Hey Dan,

A few thoughts about omni part deux.....

An upfiring mid seems fine as you've had good luck with that. I agree with the earlier poster that 2 mids may create more problems than they solve.

If you want to cross to the woofers at 4-500 (should be fine with a single small-cone mid), I'd fire one woofer forward and one back to keep the omni pattern.

Likewise with the tweeters, I'd experiment with two firing front and back in phase to keep the omni radiation pattern. A dipole tweet will put the tweeter's power response down 5dB compared to the mid and woofers and a horn/waveguide will make it even worse.

Hi Dennis,

I agree that two mids probably will create more problems than benefit. From the first project, however, I know that I will lose about 6db or more of the mids on axis sensitivity when used upfiring. Most mids are around 88 db max, which puts me down at 82db on the listening axis, not nearly enough. this is what led to the dual RS-180's in the first project. Any suggestions?

I was hopeful for the side by side mids in that they might have nulls parallel to the front plane of the speaker, sort of like the dipole nulls, kind of like tipping a WMTW CC so it faced up with a 15º angle forward. Maybe I could try some measurements with my CC front firing and then in this orientation to see what differences I get.

I can see the logic of both front and rear firing woofers and tweeters. I had done some early sketches like this. Are you suggesting a single front firing tweeter and a single rear firing tweeter, or two of each? I can't be sure from the way you put it.

**Dennis H** · 01 March 2007, 20:12 Thursday

Are you suggesting a single front firing tweeter and a single rear firing tweeter, or two of each?

Yeah, one of each. Sorry for the confusion.

**augerpro** · 01 March 2007, 21:13 Thursday

Originally posted by dlneubec

Hey Augerpro,

I ws just looking at them, I didn't see anything over 90bd sensitivity, and I expect to need at least 94db.

BTW, I did see the Audax PR170MO 6.5", which it says has a 100db sensitivity and looks like it might work from 500 to maybe 1800-2000hz, based on the 60º off axis responce. It looks like it would take a steep slope on the top end however.

Anyone know anyting about this driver?

Here:

http://www.madisound.com/cgi-bin/index.cgi?cart_id=1906936.2032&pid=325

http://www.madisound.com/cgi-bin/index.cgi?cart_id=1906936.2032&pid=324

**Scottg** · 01 March 2007, 21:37 Thursday

Originally posted by dlneubec

Hi Scott,

Thanks for the feedback.

I'm not sure I understand why you think the sloped baffle is a bad idea. This is what Mirage is doing on their new OMD-28 speaker.

FWIW, Absolute Sound did a review of the OMD-28 and gave it very high praise. Acording to the review, Mirage claims that the 15º slope directs the rear wave more upward on the back wall, allowing the front output to open more fully toward the listening area and also improves imaging. Accorging to Mirage, this puts about 30% of the drivers output toward the listening area, with about 70% off side and rear wall surfaces, which they also claim is a ratio which closely mimics the characteristics of live music.
Mirage OMD-28

I had thought about the RS52. My concerns were that at 91db, if it had the 6db or so losses I saw with my first omni project, on both the RS180's and the RS28(upward firing), it would perform at about 85db at best on the listening axis. Also, since I want at least as much output as I have with the current design, I had a hard time imagining how 1 RS52 with it's 2" dome could put out the kind of sound that 2 RS-180's are doing now, albeit over a larger range (100hz to 1450hz crossovers). I have an RS52, so if I get to the prototype phase, I will most certainly give a try. The measurements will be the best way to tell. B&O is certainly using an upfiring midrange in its BeoLab omni, but it does have a diffuser/difflector that redirects the sound into a 180º front wave, according to the creators claims.

I'm not following what you are suggesting on the wave guide. Can you explain further? Maybe, you talking about a downfiring tweeter with a wave guide? If so, I've been there and done that on the original and diffusers simply did not work in my many attempts.

I disagree with the mid not being critical. It seems to me that it determines where the woofers need to cross, what tweeters are possible to use, what crossovers are feasible, etc., not to mention that it carries most of the sound you hear.

Those speakers are interesting looking. How do you suppose they deal with the innefficiency of one up firing woofer? I wanted to avoid needing to incorporate delay in my passive network. With the tweeter on the front like that, you would have to include a delaty network, would you not. This was one of the main problems I could forsee with the two mid idea, especially the front to rear arrangement.

What did you think of the idea of using one high efficiency Pro driver? I believe that is what Duevel does with their omni's, in addition to using a compression driver on the top end that is omnidirectionally horn loaded.

Mirage's design is an attempt to improve the problems that upper freq. loss in spl's with both an alteration in off-axis performance (i.e. less than 90 degrees) and with the use of a waveguide. Both are compromises that your design does not need to necessarily endure. Ask yourself what the waveguides were like in your earlier experimentation. Moving it forward simply increases baffle contribution (which is in effect another form of waveguide). Subjectivly this will decrease the depth of the speaker.. still go ahead and try it and see If I'm correct.

The dome mid should not experience that great a reduction in spl - at least above 900 Hz or so (..but a quick measurement on your part will confirm or deny this). This will make greater demands on the midbass's passband, BUT again there are a lot of drivers that are available for use in this range. Of course you will need the second midbass driver to compensate for baffle-loss.

What I meant by the midbass "not being critical" is that there are plenty of options to pursue.. that is NOT the case with radial or the tweeter waveguide. Note the HiVi mid dome is also another option.. but it does have higher 5th order distortion.

What I was trying to suggest with the waveguide tweeter was - picture that FJ Ohm speaker.. but instead of the tweeter as it is, picture the waveguide tweeter in its place. Because of the waveguide - it will have a physical offset. (..note of course that you wouldn't have an upward firing midbass driver in relation to the FJ Ohm either, rather the 2 inch dome.. and additionally you would have 2 midbass drivers on the front baffle.)

I don't think that the FJ Ohm deals with the eff. issue. Their spl rating is prob. an in-room averaged spl.

The one driver midbass isn't going to work well - you have to think of baffle step loss.

**dlneubec** · 02 March 2007, 08:38 Friday

Originally posted by augerpro

Here:

http://www.madisound.com/cgi-bin/index.cgi?cart_id=1906936.2032&pid=325

http://www.madisound.com/cgi-bin/ind...6.2032&pid=324

Hi Brandon,

I completely overlooked the ones with the dual cones. The one thing I noticed is that they seem to roll off pretty quickly at 60º off axis and may ne be able to cross much higher than my current RS180's, though they roll of at a much slower rate and would probably allow shallower slopes in the crossover.

I don't know anything about speakers with whizzer-type cones. I always thought that was kind of a car audio thing. What kind of distortion do these introduce?

**dlneubec** · 02 March 2007, 09:10 Friday

Hi Scott,

Thanks for the follow up. My comments are below.

Originally posted by Scottg

Mirage's design is an attempt to improve the problems that upper freq. loss in spl's with both an alteration in off-axis performance (i.e. less than 90 degrees) and with the use of a waveguide. Both are compromises that your design does not need to necessarily endure. Ask yourself what the waveguides were like in your earlier experimentation. Moving it forward simply increases baffle contribution (which is in effect another form of waveguide). Subjectivly this will decrease the depth of the speaker.. still go ahead and try it and see If I'm correct.

What you refer to as a waveguide is what I have been calling a diffuser or deflector, I believe. I don't plan to use that technique here since it did not work well on the first model. The one exception was the Duevel-like concave cone was effective at extending the roll off of the RS180's on axis by about 300hz and making the roll off much less steep (see below). I may start with a separate mid enclosure that stacks on the bass enclosure(s). That way I can test angled tops and get a better idea of what they do to on and off axis performance.

Originally posted by Scottg

The dome mid should not experience that great a reduction in spl - at least above 900 Hz or so (..but a quick measurement on your part will confirm or deny this). This will make greater demands on the midbass's passband, BUT again there are a lot of drivers that are available for use in this range. Of course you will need the second midbass driver to compensate for baffle-loss.

I remain skeptical. While I did not try a dome mid, the steep drop off was evident in both the RS180's and RS28 when used up or downfiring. I will definitely test the RS52 in an upfiring and angle upfiring setup to see what kind of on axis response I get. If it were to work, it would be a much easier solution than dual mids.

Anybody have some off axis meassurements of the RS52, say out at 45-60º? That might give us a good idea on how well it would work.

Originally posted by Scottg

What I was trying to suggest with the waveguide tweeter was - picture that FJ Ohm speaker.. but instead of the tweeter as it is, picture the waveguide tweeter in its place. Because of the waveguide - it will have a physical offset. (..note of course that you wouldn't have an upward firing midbass driver in relation to the FJ Ohm either, rather the 2 inch dome.. and additionally you would have 2 midbass drivers on the front baffle.)

Ok, I see what you are suggesting. On Zaphs implementation, it sets the tweeter back about 1-1/4". I'll have to measure the RS225, but my guess is that it will need to offset further back than that if we are trying to align their VC's. Also, if you consider that the up firing RS52 has a 5-1/8" dia. faceplate and if it was installed with 7/16" clearance from the baffle edge, that puts it's VC center back 3". Tilting the baffle up actually moves the VC center forward, but not by a lot.

**dlneubec** · 02 March 2007, 09:19 Friday

Ok, here are some of the ideas/suggestions so far.

Keep the front firing RS225's, but use a wave guide tweeter above it and then an up firing RS52 above that. A rear firing tweeter could also be added on the opposite side.

Keep one front firing RS225, move the second to rear firing. Use an up firing mid, perhaps the RS52 or a higher sensitivity pro mid cone, and use a front and rear firing tweeter on a baffle above the up firing mid.

Go back to an up or down firing arrangement for the RS225's to keep them omno, an up (or down) firing mid and have both a front and rear firing tweeter.

Any other ideas?

Question: Is it wise to mix dipole radiation with one driver with omni radiation from another?

**dlneubec** · 02 March 2007, 09:26 Friday

Here is another concept that has occurred to me. Use and up and down firing RS225. The RS225's are crossed low enough so that the wavelengths are pretty long, so it might not have significant diffraction issues if you were to mount a narrow box above it that held a front and rear firing mid and a front and rear firing tweeter.

**joecarrow** · 02 March 2007, 12:27 Friday

Dan, that'd be fine- except the mid is probably playing to a high enough frequency that its external environment will matter a great deal.

Depending on what the RS52 looks like off axis, I think the front+rear firing RS225s plus top firing RS52 plus some undetermined tweeter sounds like the most promising arrangement.

**dlneubec** · 02 March 2007, 14:38 Friday

Ok, here is a couple quick new looks. Version A is pretty much the same, but the woofers have been shifted forward to align with the top mounted RS52 (if that indeed works). The inside cabinet is up to about 65 liters, so one might isolate the top part of the box for a sealed version (50L or so) or use the whole 65L and do a TL, ported or PR system.

Version B has re-arranged the RS225's so they fire up and down, but at each other, in the center of the bottom box, which puts the output centered around 16.5" from the floor, maybe not too bad from a floor to mid bounce cancellation standpoint. The boxes would need to remain sealed in this version. I'm unsure about the seperation needed between them at this point, but you get the idea. It has a bit more flexibility, since the second woofer does not have to be firing out of the top baffle. It retains the top sloped baffle but shows the possibility of either the RS52 or a 6.5" pro-type driver on the baffle. The tweeter baffle is above with front and rear tweeters, all VC's can be aligned.

I'd be interested to hear what folks think about using a 94db 6.5" Pro diver for the mid from about 500-2.5 or 3khz or so.

Version C uses the bass bin concept from version B, but flattens the top and uses a top box that has front and rear firing mids and tweeters as possibilities. Not quite the omni mid I was originally hoping for, but could be interesting. All VC's are aligned, at lest the front mounted and woofers are.

I'll keep noodling on it to see what else I can come up with. :T I'm having trouble finding a solution that uses the waveguide tweeter on the front and still aligns VC's with the woofer and tweeter.

I haven't attempted a front and rear firing RS225 version as of yet.

Any thoughts or further ideas to explore?

Attached Files

**joecarrow** · 02 March 2007, 15:38 Friday

C seems interesting because you could have switchable dipole/bipole operation.

**Scottg** · 02 March 2007, 16:07 Friday

Originally posted by dlneubec

Ok, I see what you are suggesting. On Zaphs implementation, it sets the tweeter back about 1-1/4". I'll have to measure the RS225, but my guess is that it will need to offset further back than that if we are trying to align their VC's. Also, if you consider that the up firing RS52 has a 5-1/8" dia. faceplate and if it was installed with 7/16" clearance from the baffle edge, that puts it's VC center back 3". Tilting the baffle up actually moves the VC center forward, but not by a lot.

I didn't realize it was that shallow. Perhaps the waveguide from Music Supply Center?:

http://www.musicsupplycenter.com/ProductDetails.asp?ProductCode=810%2D002

Considering the diameter of the waveguide you could start using 12" pro drivers for the midbass AND more than likely use a 1st order electrical for tweeter to radial mid. Moreover you could probably use a more eff. mid like the B&C., Audax, Maxx Fidelity, 18Sound, or at a better price - the Ciare 6.38 NdMR

404 Not Found

http://www.assistanceaudio.com/08_CIARE.html

(Perhaps the Peerless HDS tweeter for the waveguide if utilized with a more eff. radial mid.?)

All of that "cranks" up the price considerably though (not the waveguide - just the "extras").

Also I'm sure a fine blade hand saw can cut that faceplate (for the Dayton mid-dome) to allow you to move it much closer the the front of the baffle.

I would NOT recommend a bipole pattern of any kind (unless very low in freq.), again combing becomes a serious audible issue in that instance.

**nerd of nerds** · 03 March 2007, 14:34 Saturday

Originally posted by dlneubec

Ok, here is a couple quick new looks. Version A is pretty much the same, but the woofers have been shifted forward to align with the top mounted RS52 (if that indeed works). The inside cabinet is up to about 65 liters, so one might isolate the top part of the box for a sealed version (50L or so) or use the whole 65L and do a TL, ported or PR system.

Version B has re-arranged the RS225's so they fire up and down, but at each other, in the center of the bottom box, which puts the output centered around 16.5" from the floor, maybe not too bad from a floor to mid bounce cancellation standpoint. The boxes would need to remain sealed in this version. I'm unsure about the seperation needed between them at this point, but you get the idea. It has a bit more flexibility, since the second woofer does not have to be firing out of the top baffle. It retains the top sloped baffle but shows the possibility of either the RS52 or a 6.5" pro-type driver on the baffle. The tweeter baffle is above with front and rear tweeters, all VC's can be aligned.

I'd be interested to hear what folks think about using a 94db 6.5" Pro diver for the mid from about 500-2.5 or 3khz or so.

Version C uses the bass bin concept from version B, but flattens the top and uses a top box that has front and rear firing mids and tweeters as possibilities. Not quite the omni mid I was originally hoping for, but could be interesting. All VC's are aligned, at lest the front mounted and woofers are.

I'll keep noodling on it to see what else I can come up with. :T I'm having trouble finding a solution that uses the waveguide tweeter on the front and still aligns VC's with the woofer and tweeter.

I haven't attempted a front and rear firing RS225 version as of yet.

Any thoughts or further ideas to explore?

Why don't you put the front tweeter in a waveguide on a thick baffle to get its voice coil alligned with the rear firing tweeter?

**dlneubec** · 03 March 2007, 18:14 Saturday

Originally posted by nerd of nerds

Why don't you put the front tweeter in a waveguide on a thick baffle to get its voice coil alligned with the rear firing tweeter?

That is what Scott is suggesting as well. I'm trying some plans for that, but there are some problems with that approach (see plan attached). Here is what i'm struggling with.

The first problem is that the ideal tweeter location is close to ear height when seated, say 36"-40" or so. If your plan is to put the mid on top, firing up, and the waveguide tweeter on the front below it, you end up with the mid very high, maybe too high. See the attached plan for a first attempt at this. The tweeter is a bit low at 35 1/2" and even with the RS52 moved as close to the front of the top as possible, it ends up at just over 41".

The second problem is getting a waveguide that is deep enough and not too large to fit the cabinet. As the WG gets bigger in dia. the farther it has to be from the mid and that can cause problems, not to mention, it would only raise the mid in even higher.

Thirdly, I've read many designers complain about the size of the flange on the RS52 (5 1/8" dia.) causing problems with CTC spacing with the tweeter and requiring a small flange tweeter. When you put the tweeter in a WG and use the RS52, it seems like you are asking for trouble. If you use an upfiring cone driver instead, perhaps a 6.5" Pro driver to get the effiency to work, that will push the center of the voice coil back even further, making it that much more difficult to come up with a waveguide to work with a dome tweeter. I suppose you would have to go with a compression tweeter at that point.

This is my first 3way and only my second full design, so perhaps others can address what the concerns would be here in more detail.

Attached Files

Omnidirectional-v02d_3-3-07.pdf (66.4 KB, 448 views)

**dlneubec** · 03 March 2007, 18:28 Saturday

Here is another conceptual plan.

It uses the centrally located up and down firing RS225 bass bin concept. In order to get more spl out of the mid (and less omni dispersion as a result) I', showing the tweeter baffle oriented over the RS52 dome with a circular area drilled out, sort of like the BeoLab omni speaker. This should redirect a lot more sound forward and less off the back wall. The tweeter baffle also acts as a partial baffle for the RS52. The rear facing tweeter, when switched on, could help to round out the omni effect at the upper end.

I think I could build the tweeter baffle without too much trouble. I'd start by glueing up the 2-layer tweeter baffle first. Then I would clamp a couple more 3/4 pieces of scarp mdf to it. The I would drill a 2 1/8"-2 1/4" hole at an angle through the bottom of the baffle and scap pieces until th hole exits from the 2 layers on the glued up tweeter baffle.

What kind of problems might occur from a diffraction or resonce standpoint I can't predict until I try it and take measurements.

If anyone has any thoughts on this, I"d sure love to hear them!

Attached Files

Omnidirectional-v02f_3-3-07.pdf (55.2 KB, 431 views)

**nerd of nerds** · 03 March 2007, 19:27 Saturday

Originally posted by dlneubec

Here is another conceptual plan.

It uses the centrally located up and down firing RS225 bass bin concept. In order to get more spl out of the mid (and less omni dispersion as a result) I', showing the tweeter baffle oriented over the RS52 dome with a circular area drilled out, sort of like the BeoLab omni speaker. This should redirect a lot more sound forward and less off the back wall. The tweeter baffle also acts as a partial baffle for the RS52. The rear facing tweeter, when switched on, could help to round out the omni effect at the upper end.

I think I could build the tweeter baffle without too much trouble. I'd start by glueing up the 2-layer tweeter baffle first. Then I would clamp a couple more 3/4 pieces of scarp mdf to it. The I would drill a 2 1/8"-2 1/4" hole at an angle through the bottom of the baffle and scap pieces until th hole exits from the 2 layers on the glued up tweeter baffle.

What kind of problems might occur from a diffraction or resonce standpoint I can't predict until I try it and take measurements.

If anyone has any thoughts on this, I"d sure love to hear them!

Instead of an optional dipole tweeter why not just use an open back planar like the B&G PDR3?

It performed well in john's "battle of the non-domes" test, better than teh D26 in some respects. You would have to remeasure it without the back cup but this would provide a cheaper dipole tweeter solution.

Furthermore, this thread has gotten me interested in a small (read: Tiny) omni speaker again. I was thinking about using a small 3" TB or HiVi full ranger at the end of a 3" piece of pvc (the hi vi 3" i have fits perfectly in fact) with a similar design to the Pluto+, ie, a small subwoofer crossed over higher to avoid the low end distortion of the hivi woofer. Perhaps even a small desktop/nearfield design...

EDIT: As far as the dayton dome having a large flang...Why not chop some of it off? Its been done to get the C2C spacing on tweeters down before and i see no reason it isn't possible to slice a bit of the mid away.
this would keep you from that somewhat goofy lookin baffle on the 8" drivers and you wouldn't have to have such a deep waveguide on it.

**dlneubec** · 04 March 2007, 09:39 Sunday

Originally posted by nerd of nerds

Instead of an optional dipole tweeter why not just use an open back planar like the B&G PDR3?

It performed well in john's "battle of the non-domes" test, better than teh D26 in some respects. You would have to remeasure it without the back cup but this would provide a cheaper dipole tweeter solution.

It is one I am considering, however, everything I have read suggests to me that it does not perform well in dipole mode (with no back cup), inlcuding Zaph's review of it. I would not be much cheaper than a couple of the many $30 tweeters, by the time you purchase the face plate.

Originally posted by nerd of nerds

EDIT: As far as the dayton dome having a large flang...Why not chop some of it off? Its been done to get the C2C spacing on tweeters down before and i see no reason it isn't possible to slice a bit of the mid away.
this would keep you from that somewhat goofy lookin baffle on the 8" drivers and you wouldn't have to have such a deep waveguide on it.

It would be hard to move it much closer to the front of the box than I already have in the latest plan (1/4" back from the front edge). This is for two reasons. First it has a 1/8" aluminum flange, not an easy thing to chop off and do a good job of it. Second the outside dia is 5 1/8", but the body dia. is close to 4 1/8" (which it the cutout size given). That means that you have only a 1/2" overlap from the body to the edge of the flange. If you put it 1/4" in from the edge, that means the 4 1/8" cutout is only 3/4" from the edge. I would not want to move the cutout much closer than 1/2" from the edge, which only gains another 1/4", not a lot of gain for the work of cutting off the flange and probably not enough to allow the Zaph-type WG impementation to work. The Zaph-type implementation only sets the tweeter back about 1.25". Even I went to within 1/4" of the edge of the baffle with the cutout, I would still have the center of the RS52 at least 2 5/16" back, so have have over an inch to make up yet.

**dlneubec** · 04 March 2007, 10:03 Sunday

I think I may need to step back and try to determine what will drive other decisions on this project. Perhaps it is to make every driver as close to omni performance as possible, or perhaps it is to utilize a waveguide tweeter for better power responce and crossover transition with the mid, or perhaps it is time (VC) alignment. This might make other decisions clearer. Right now I think there are too many variables I'm trying to juggle up in the air. :E

Alternatively I might want to start with one section, say the bass section. If I were to study it further and determine what type of aligment I wanted to use, it would help make further decisions on the mid(s) and tweeter more clear. Candidates might be up and down firing (omni) away from each other, up and down firing (omni) toward each other, driect firing front, front and rear firing (sort of omni or dipole), etc.

For example, let's say I go back to the up and downfiring woofer idea and use the model in the plans I have already done with them firing at each other. I can then study whether they should be offset in the baffle left to right or centered and front to back or centered. Once this is done and say I want to align VC's, that would allow only certain solutions for the mid(s) and tweeter(s).

Does that kind of approach make sense?

**nerd of nerds** · 04 March 2007, 13:35 Sunday

Originally posted by dlneubec

I think I may need to step back and try to determine what will drive other decisions on this project. Perhaps it is to make every driver as close to omni performance as possible, or perhaps it is to utilize a waveguide tweeter for better power responce and crossover transition with the mid, or perhaps it is time (VC) alignment. This might make other decisions clearer. Right now I think there are too many variables I'm trying to juggle up in the air. :E

Alternatively I might want to start with one section, say the bass section. If I were to study it further and determine what type of aligment I wanted to use, it would help make further decisions on the mid(s) and tweeter more clear. Candidates might be up and down firing (omni) away from each other, up and down firing (omni) toward each other, driect firing front, front and rear firing (sort of omni or dipole), etc.

For example, let's say I go back to the up and downfiring woofer idea and use the model in the plans I have already done with them firing at each other. I can then study whether they should be offset in the baffle left to right or centered and front to back or centered. Once this is done and say I want to align VC's, that would allow only certain solutions for the mid(s) and tweeter(s).

Does that kind of approach make sense?

As cheesy as this may sound, maybe you should figure out a mission statement for this project? Of course you want to make an omnidirectional speaker, but what is the overall goal? Price? Performance? Simple crossover?

Based on the performance of the 5" TB titanium driver it might be worth it to buy a couple of the little 3" titaniums to use as mids for this. It would help with the alignment quite a bit. The 3" bamboo driver performed well, however, nowhere near the performance of the dayton dome.

And yet another idea comes to mind. Creative sound solutions has a truncated frame version of the WR125S. You'd want to cut away a bit of the front baffle to get it close enough to the front, but its going to be in a seperate enclosure from the 8s right?

**dlneubec** · 05 March 2007, 10:38 Monday

What I want to do is create an omni that is an improvment on the first design. The first priority is sound quality improvement (that's oringinal

) and secondly in appearance/size. I'd like it to be a 3way without a sub built in. The other goals I listed at the start of the thread were intended to get the above things accomplished. I've sort of backed off the one about front firing woofers, though.

I'm skeptical that any single driver below 94db will provide enough output in an up or down firing scenario. That includes the RS52, though I have one I will test at some point to see what it is capable of at 90º off axis. I believe the traditional cone drivers used in most DIY products for midranges will have to be doubled up in an omni, which presents a whole lot of challenges and is the reason for the top box in the original design. I'm still hoping to avoid the top box scenario.

Potential answers I see are:

1) High effeciency Pro drivers
2) Deflection/redirection of some of the omni energy toward the listening area
3) Some creative arrangement of dual drivers

To that end, and it's probably a crazy idea, but what about two RS52's mounted face to face, something like the attached plans?

Also attached are some ripple tank studies at mid-low, mid, and mid-high frequencies. Unfortunately, I can't tell from the site what those frequencies are. They simply have sliding scale to set them.

Attached Files

**dlneubec** · 07 March 2007, 11:22 Wednesday

Ok, I decided to take a step back and do some rethinking on what my goals are. Here is my current thinking:

1) An omnidirectional, direct firing hybrid like the original with an extended omni midrange, maybe with a switchable dipole tweeter implementation.
2) A little more attractive design aesthetically, with higher SAF, a smaller front footprint.
3) No subwoofer included this time, instead with sealed woofers rolling off around 50hz.
4) Consider a crossover with lower order slopes and crossover points that don't push the drivers too hard. Entirely passive crossover.
5) Extending the omni coverage up higher in frequency by higher mid to tweeter crossover and possibly lower order slopes in a 3way design.
6) Equivalent or increased efficiency, but possibly higher output without strain on drivers.
7) Ability to include two drivers of each type if needed for efficient output
8 ) Possible use of a waveguide on the tweeter (and a rear firing tweeter opposite) for better crossover transition and controlled directivity though the crossover.
9) Time alignment of the drivers at the listening point (aprox. 11'-12').

I have gone back to the drawing board with these goals in mind. I realized that I was in error thinking the drivers needed to be veritcally alinged when they actually need to be on a curve relative to the listening distance and ear height. It looks like I have to keep the top box concept of the original to have the flexibility to meet goal 7. Attached is a new concept plan. The midrange drivers are at ear height, putting the tweeters 6" above. I'm initially thinking that the dome midrange (RS52) might work better in this configuration, because they might not introduce the additional complication of the driver cavities themselves when drivers firing at one another. I show a dome tweeter in the plans, but I'm thinking a ring radiator might be better with the waveguide, based on what I've read recently.

I'm looking for feedback. Not so much on the aesthetics, but on the design issues. What am I doing wrong? What am I doing right? What am I missing? Anything else I should consider?

Thanks! :T

Attached Files

Omnidirectional-v02g_3-7-07.pdf (76.7 KB, 440 views)

**joecarrow** · 07 March 2007, 12:01 Wednesday

I think a really important thing at this point is to just do that measurement of the RS52 at 90 degrees. It would answer a lot of questions.

Also, taking a step back is a really good idea- I often find that it can rescue a design that's gone down the wrong path or been painted into a corner. Just keep in mind that in your new research you may quickly learn things that will change your assumptions and priorities- those are the things you want to learn first.

It's probably a good idea to figure out the acoustic side of this before worrying too much about aesthetics.

**dlneubec** · 07 March 2007, 12:13 Wednesday

Thanks Joe,

I'm going to see if I can find a stiff cardboard box with an end about this baffle size that I can mount the RS52 in to take some 90º off axis measurements of. I could not only test its off axis behavior, but I could experient with mounting a baffle facing it at various heights to see what effect it has. I assume that at some point, as the frequency goes up and wavelengths get shorter, that diffraction effects will also become more evident. Unfortunately, I only have one RS52, so unless they are back in stock, I can't test with two of them at the moment.

I agree that my assumptions and goals are very likely to change or evolve as I do this type of testing. They sure did the last time! I was hoping for some valuable feedback from some of the experts on the forum who might see some things that are immediately questionable. That might save me a lot of time when it comes time to build a prototype.

**dlneubec** · 08 March 2007, 22:08 Thursday

I ran some prelim measurements tonight. I used a cardboard box and mounted the RS52 on the end. The measurements were taken with a 4ms gated window, so the FR is only good down to about 500-600hz or so. Measurements were at 1m distance at the driver height. The first graph below includes the following:

Fine Lines (all meaurements on the listening axis):

Red- baffle facing straight up
Black- baffle tilted forward 15º
Blue- baffle tilted forward 30º
Green- baffle tilted forward 45º
Purple- baffle tilted 90º (normal front firing)

Heavy Lines (all measurements 90º off axis)

Red- baffle facing straight up
Black- baffle tilted forward 15º
Blue- baffle tilted forward 30º
*Green- baffle tilted forward 45º

*The green was taken first and after looking at the roughness, I thought I might be getting some reflection off an adjacent TV, so I moved the box a bit further away for the other. The rest of them tracked very closely to one another and the rougness is probably diffraction off the square edges of the box.

I'm not sure what to make of these, except a few things. There is a definite drop off on FR of around 8db from direct firing to full up firing. That seems to indicate that two RS52's would be needed in the upfiring position to get similar response to a single direct firing RS52.

You can also see how the direct firing RS52 is pretty flat from 600hz out to 10khz. The 90º upfiring FR is pretty flat from about 500 to 2500, where it gently rolls off about 10db at 8khz. The 15º, 30º and 45º forward sloped baffles generally follow the same pattern as the 90º up firing, but with higher spl. The direct firing and 45º sloped baffle track fairly closely out to about 2.5Khz, where the 45º begins to fall off and the direct firing remains flat.

The heavy plots, all 90º to the listening axis track very closely to one another regardless of the baffle slope, so off axis performance may not be effected much be the baffle slope. It seems to me that this indicates that you can slope the baffle forward from upfiring and not effect the off axis response, at least out to 90º. I will have take some measurements behind with from behind the baffle to see what reduction in FR there would be off the wall behind the omni speaker as the baffle is sloped forward from upfiring. I assume the amount of slope with reduce the omni effect as it is increased.

I'm guessing I will need to crossover to the tweeter in the 2-3khz range and to the woofers in the 700-800hz range.

Here is the surprise. The second graph is a shows the FR plots of 2 different single RS180s at 90º upfiring. The surprise is that a single RS180 firing up is about 5db below a single RS52 firing up. Unless I'm missing something, that would suggest that a single RS52 firing up would perform at about the same spl as two RS180's firing up. The single RS52, sloped 15º forward is about 8db above a single RS180. I did not include the RS52 on the 30º sloped baffle in this plot, but if you compare to the other graph (blue line), you can see that the 30º sloped baffle is about 11db higher than a single RS180.

My assumption at this point is that 1 RS52, firing up is going to give a similar spl as two RS180's but extend the upper end another 1-1.5khz. and allow a flatter crossover slope between it and the tweeter. If I use a forward sloped baffle in the 15º to 30º range, I might be able to even pick up some several db over two RS180's. That said, it stall has to be matched with the two RS225's and the direct firing tweeter. I'm more encouraged about the possibility of a single RS52 doing the job.

I'd sure appreciate anyone else's take on what these measurements indicate to them. Anyone willing to share their insights?

**dlneubec** · 09 March 2007, 14:08 Friday

Ok, I've reworked a plan with the idea of using a single RS52 on a sloped baffle (15º to 30º). The one shown is 30º. I will know which works better with more testing.

I'm thinking of crossover points in the 600-700hz range and the 2400-2600hz range. A smaller form factor tweeter may help since it might allow a closer CTC distance. BTW, BoxyCad2 suggest that the optimum crossover point for this plan would be 570hz from a floor bounce perspective.

I envision covering the gap where the RS225's are with speaker cloth, but will have to figure out exactly how to do it and make it removeable so the boxes can be separated when necessary.

This weekend I plan to do some more meaurements to try and determine what slope baffle to use. Also, I want to try and mock up a tweeter baffle similar to this (probably out of cardboard) to put over the RS52 to see what that does to it's responce.

Any thoughts?

Attached Files

Omnidirectional-v02h_3-9-07.pdf (68.1 KB, 423 views)

**Scottg** · 09 March 2007, 17:33 Friday

Originally posted by dlneubec

Ok, I've reworked a plan with the idea of using a single RS52 on a sloped baffle (15º to 30º). The one shown is 30º. I will know which works better with more testing.

I'm thinking of crossover points in the 600-700hz range and the 2400-2600hz range. A smaller form factor tweeter may help since it might allow a closer CTC distance. BTW, BoxyCad2 suggest that the optimum crossover point for this plan would be 570hz from a floor bounce perspective.

I envision covering the gap where the RS225's are with speaker cloth, but will have to figure out exactly how to do it and make it removeable so the boxes can be separated when necessary.

This weekend I plan to do some more meaurements to try and determine what slope baffle to use. Also, I want to try and mock up a tweeter baffle similar to this (probably out of cardboard) to put over the RS52 to see what that does to it's responce.

Any thoughts?

Several.

Why use a sloped baffle?

Consider that one of your constraints is a small form factor. IF the height is pretty low (say less than 33 inches in height), then most listening (assuming a listening position that isn't too low) should NOT be at 90 or even 80 degrees off-axis. (..more like 60-75 depending on the distance from the speaker). And there is nothing to say you couldn't make the speaker shorter than that. I believe that FJ Ohm speaker was about 29 inches in height.

..and of course you could also combine the two elements.. i.e. short AND with some slope. Remember though, the more slope the more the baffle will effect the sound (..however, 30 degrees doesn't sound to bad).

Note however that if you DON'T use an up and down config. like your previous speaker, that you will need a variable attenuator/transformer for your tweeter section to adjust for listening distance and height. (I'd suggest looking through Parts Express for their 8 ohm variable transformers - even if only stereo rather than mono. They are usually cheaper AND sound better than a normal resistive padding attenuator.)

For the driver I'd skip the 2X RS225's. Instead I'd get one of these:

Page not found

http://www.partsexpress.com/pe/showdetl.cfm?&Partnumber=264-852

with an EBS vented "alignment". It's a good price and it has enough eff. to allow for a shelving circuit for the baffle step compensation.

Note: play with the "Edge" baffle simulator to see how to smooth out some of that baffle induced ripple. Also remember that you can cut off a good bit of that faceplate of the mid in order to get it closer to the baffle's edge.

To keep things cheap I'd still go for the MCM waveguide and use a HiVi K1 tweeter instead (perhaps X2 for dipole radiation or maybe even bipole radiation as long as the waveguide keeps any side combing well down in spl's).

Again, consider the formula I mentioned previously with the waveguide BELOW the omni mid. Perhaps with a shape similar to this:

Page Not Found - Dagogo

http://www.dagogo.com/VerityTaminoX2.html

What you have proposed via PDF - will have FAR less spousal approval than your previous design. Additionally, its also a LOT more complex, in ALL respects (..particularly the build process).

**dlneubec** · 11 March 2007, 21:36 Sunday

I did some more testing of the RS52 this weekend. I first tried it with the 30º sloped baffle with various vertical tweeter baffle configurations, like shown in the last plan I posted. The results were not encouraging. There were too many interactions with the tweeter baffle.

So, I decided to go back to the drawing board again and try a box/baffle placed directly above the RS52 as shown in the plan attached. I figured that it worked in my original design, so why abandon it. It seems that the FR is directly affected by how high above the driver the box/baffle is placed. I assume the FR is strongly affected by the distance the baffle is away from the driver. The FR is actually boosted some by a baffle placed above it and the closer the baffle, the more the boost. Also, the closer the baffle the higher in frequency that the driver is extended before significant anomalies show up(see the first graph below). The black is the up firing RS52 with no baffle above it. The red is with the baffle placed above the RS52, about 1.25”. You can see that the baffle provided a 3-4db boost in FR which starts to depart at about 600hz up to about 800hz out to about 6khz. I tested heights from 1” to 2” in 1/8” to ¼” increments in order to understand what effects baffle height had on the FR. The second plot shows the FR of various heights from 1” to 1.75”. I also tested the baffle at various positions front to back over the RS52 and side to side. The front to back positioning made a significant difference, but the side to side made little difference when measuring on axis.

Up to this point I had been using a board about 6”x8” placed above the RS52. After an initial round of tests, I decided to rough out a partial box similar to the one in the design below and retested. The results led me to the combination of having the tweeter baffle edge lined up with the front edged of the hexagrid on the RS52 and 1” above it. This seemed to provide the best overall FR, with the most extension and/or the smoother roll off with fewer oddities high in the FR.

I also tested that position at 0, 30, 45, 60, 90, and 180 off axis (see the third graph). The red line is the up firing RS52 with no baffle and the black line the on axis with tweeter baffle above for reference. The off axis plots tracked with one another quite well, except for the one at 180º from the listening axis. All but the on axis have a 10db or so dip at around 6khz, but I would think that would be comfortably attenuated by the crossover, which I expect to be in the 1900-2100hz range.

Finally, I tested the baffle at various mike heights. I did not measure the angles, but simply raised the mike in 4” increments from 20” through 36”, with the baffle height being at 28” (I’m using a solid cardboard box for the testing for now). In graph 4, you can see the up firing RS52 with no baffle in black for reference. The red plot was at 28” and the others 4” and 8” above or below that. The spl increased as you went higher and decreased as you went lower, but the responses tracked on another quite well and the box has sharp edges, so they could be better with a chamfered or rounded over baffle.

I think this design concept has possibilities and I may pursue it further, subject to feedback, if I get any. I’m thinking of crossovers in the range of 600-700hz and 1900-2100hz. The tweeter is undecided at this time. I’m going to stick with the RS225’s for the woofers and the single RS52 for the mid. I will include a plot of a single RS225 in my next post. I hope to use to use a less steep filter between the woofer and the mid, but expect to need a very steep filter on the top end of the RS52, due to the higher CTC I have between it and the wave-guided tweeter.

I hope this design should have a higher SAF than the others I’ve come up with. It ends up at least 5” shorter than the original omni project and much narrower throughout from the front. The sloped sides and front will make for a difficult build, though, since so many parts have their own unique dimensions and angles. The drivers VC’s are all time aligned on an arc from the listening position to maintain time alignment.

Some questions I still have are what tweeter? Would a ring radiator be better with the WG as reported in one of the articles linked to the WG thread? If so, which is better, a 19mm or 25mm, with the 19mm having a much wider dispersion than the 25mm? Is wider dispersion a good thing for a WG or not?

What is the optimum distance between the 2 RS225’s? I may have to build test boxes and take measurements to determine this. This may not be critical at all since I’m not going to push it too high.

What are the optimum crossover points, slopes and types for this configuration, given the FR of the RS225 and RS52 and the WG and rear firing tweeters?

Is there any benefit to the rear tweeter being WG loaded as well? Should they be wired in phase or not?

What else am I missing and need to consider? Anyone have any thoughts?

Attached Files

Omnidirectional-v02i_3-11-07.pdf (58.4 KB, 295 views)

**dlneubec** · 11 March 2007, 21:43 Sunday

Here is the FR plot of the single upfiring RS225. The RS225 is in green. Note I used a 4ms gated window, so the response is not much good below 500-600hz (same with the rest of these measurements). The black is the favored FR from the RS52 with tweeter baffle above.

You can see that if you add in the second Rs225, the response should be close to matching the RS52, maybe a little below it. However, there may be some boost from the close baffles on the RS225, I don't know yet.

**Scottg** · 12 March 2007, 01:38 Monday

Its very complex.. and it prob. won't be as pleasing to any given spouse (in comparison to the original), BUT I think it looks great and the basic design gets a :T

**joecarrow** · 12 March 2007, 02:37 Monday

Very interesting. Doesn't a waveguide tweeter have more controlled directivity in the lower octaves, compared to a regular tweeter? I think this is a good thing for matching the directivity of a tweeter to a larger midwoofer, since the midwoofer would probably be beaming a bit by the crossover frequency.

In your case, won't this make a sharp change from omnidirectional mid to directional tweeter?

**dlneubec** · 12 March 2007, 07:57 Monday

Originally posted by Scottg

Its very complex.. and it prob. won't be as pleasing to any given spouse (in comparison to the original), BUT I think it looks great and the basic design gets a :T

Probably true about the general SAF, however mine likes it better! :W She's not that crazy about the original, preferring my NaO Mini's instead, from an aesthetics standpoint, that is.

This is also preliminary. It could end up with speaker fabric covering the middle gap if I can figure out a goo way to do it and have it removeable. I want the whole thing to divide into 3 parts for easy moving. I've got a bad lower back (had back surgery back in July) and I no longer try to haul heavy stuff around, so portability is something I'm sensitive to. Heck, maybe I can come up with a hood to go over the entire top kind of like the old Ohm's! That would probably up the SAF.

**dlneubec** · 12 March 2007, 08:17 Monday

Originally posted by joecarrow

Very interesting. Doesn't a waveguide tweeter have more controlled directivity in the lower octaves, compared to a regular tweeter? I think this is a good thing for matching the directivity of a tweeter to a larger midwoofer, since the midwoofer would probably be beaming a bit by the crossover frequency.

In your case, won't this make a sharp change from omnidirectional mid to directional tweeter?

You may be right. That's why I was looking for feedback from the experts on the combination of omni and WG/rear firing tweeter.

My thinking was by having the rear firing tweeter, you might not have that sharp transition, but rather, it could continue a more omni-like sound, the degree of which seems to be dependent upon if they are bipole or dipole tweeters. I'd like to test both. Maybe they could be made switchable, I don't know. :blink:

From what little I understand about bipole's, they do approximate an omnidirectional soundfield (I've never heard one), so theoretically it seems that would be a good match in power response. Dipole, on the other hand has a similar enveloping soundfield, but with the nulls at the sides, which reduce side wall reflections (may be a good thing?). I'm unsure how omni and dipole radiation will match up in power response.

I'm also not sure if the WG helps or hurts in the bipole scenario. I hope some of the Wave Guide experts read this and can comment. It does help in the design by allowing all the drivers VC's to be time aligned while accomodating the optimum placement of the baffle above the RS52 for it's best FR, at least from the tests I've done so far.

What do you think?

**joecarrow** · 12 March 2007, 14:53 Monday

I think a fast and easy (cheap) test would be to get one of these tweeters- http://www.partsexpress.com/pe/showd...number=275-030 - and see how it does firing up. It's a 3/4 inch neodymium tweeter, 6 ohms with 92 db sensitivity. It should have really decent off-axis performance; and it is supposed to work quite well above 3.5 khz. That makes it difficult to transition to the dome with your current baffle setup, but a little experimentation could make for a smooth transition.

**dlneubec** · 12 March 2007, 15:16 Monday

Originally posted by joecarrow

I think a fast and easy (cheap) test would be to get one of these tweeters- http://www.partsexpress.com/pe/showd...number=275-030 - and see how it does firing up. It's a 3/4 inch neodymium tweeter, 6 ohms with 92 db sensitivity. It should have really decent off-axis performance; and it is supposed to work quite well above 3.5 khz. That makes it difficult to transition to the dome with your current baffle setup, but a little experimentation could make for a smooth transition.

Hi Joe,

Thanks for the suggestion.

If you are suggesting I can use an upfiring tweeter, I'm a little skeptical. I did a lot of testing of the RS28 upfiring with the original omni project. I tried it without any diffuser/deflectors and with bunches of different shapes and sizes, heights, etc. No matter what I tried, it had significant drop off in FR, which all driver do off axis as frquency increases. It wasn't bad out to about 6khz, but then dropped steeply about 10db to about 8-9khz, sort of flattened out to about 13khz and then dropped off very steeply. Attached is a plot of the RS28.

Maybe if you could find the perfect tweeter, say a 19mm ring radiator with really good off axis response out to 90º, a rising response on axis, and a high efficiency, maybe it could work with a limited high end. Or maybe you'd have to go 4way and introduce a supertweeter. For example, from the RS28 plts below, use it from 2khz to 6khz crossing to a ST to cover from 6khz to 18khz?

Hmm.., I wonder if that would work? Oh, no, I've not even done a 3way yet and I'm talking about a 4way! :E

**dlneubec** · 12 March 2007, 15:41 Monday

Here is an interesting tidbit. I had some Duevel-like concave cone diffusers made for the original omniproject. I very roughly tested one with the RS52 and with an RS150 by just propping them up rounghly over the drivers.

While this diffuser did not do much with the RS180 in the original project, it is interesting to see what it does to the FR of the RS52 and RS150.

The plot below shows the RS52 with the best response I could get with the box above it (red) and the FR with the duevel-like diffuser above it (light blue). Even though it has a notch at 4.5k, look how much flatter and extended the RS52 is with the diffuser than without it. The dark blue line is a single RS150 upfiring into the same diffuser. Unfortunately, I did not save a plot of the RS150 upfiring without the diffuser. As I recall, it had a much earlier, steeper drop off.

**Scottg** · 13 March 2007, 02:32 Tuesday

Originally posted by dlneubec

Probably true about the general SAF, however mine likes it better! :W She's not that crazy about the original, preferring my NaO Mini's instead, from an aesthetics standpoint, that is.

This is also preliminary. It could end up with speaker fabric covering the middle gap if I can figure out a goo way to do it and have it removeable. I want the whole thing to divide into 3 parts for easy moving. I've got a bad lower back (had back surgery back in July) and I no longer try to haul heavy stuff around, so portability is something I'm sensitive to. Heck, maybe I can come up with a hood to go over the entire top kind of like the old Ohm's! That would probably up the SAF.

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).

The Mentor II: omnidirectional project #2

The Mentor II: omnidirectional project #2

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment