3-way or 4-way dipoles?

[Mazeroth]

I have these current pieces to begin building my new dipole speakers:

Four Dayton RS12 HiFi subs (2 per speaker)
Four Dayton RS180 mids (2 per speaker)
Two Dayton RS28A unshielded tweeters

My initial plan was to do an MTM up top, inspired by Monte Kay's RS dipole project. After reading about his new dipole project I started thinking of a different project for myself:

http://www.mfk-projects.com/in_the_works.htm

I'm thinking of sticking with the Dayton RS12 HiFi subs, going with dual Dayton RS225 midbass drivers, dual Dayton RS125 mids (possibly RS150s), and a single Dayton RS28A tweeter, in a WMTMW configuration, or possibly a single mid to create a WMTW top section. Monte runs his Excel 4.5" mid from 700hz to 2100hz, which would help alleviate both the RS180s and RS28A from problem areas, if I were to stick with the original plan. I do like massive dynamics and that was the thinking behind dual mids, but I could be swayed to using a single.

For this project I would be using a Behringer DCX2496, which I already own. In the case of building the dipoles with separate woofers and mids I would need another DCX2496, which isn't really a big problem. Two DCX2496s provide 12 channels of crossover. These speakers would require 8, a center channel could take 3, and the infinite baffle sub I'll be using to augment 25-30hz and down will take the last channel. These will be used roughly 80% for music, and the only time I would use the center or IB is for watching movies.

With that, do you guys see any flaws in my plan? I like going overboard when building stuff and I hate to second guess anything I do. Going 4-way active dipoles will leave me with something I don't think I could second guess, as there's not much more you can do! I love Dayton RS drivers, so I'm 100% sold on those.

Thanks for any input you guys can give me!

[Paul H]

A few more or less random thoughts ...

I have a thread here documenting my 4-ways, which has some good ideas and advice from various folks here on dipoles:

http://www.htguide.com/forum/showthread.php4?t=11889&highlight=dipole+arvos

The first thing I would suggest you do is check out Linkwitz's site, specifically the spl chart which will give you a good idea of how low your 12" drivers can reach, and at what volumes. Check out your RS180's capabilities too while you're there.

If you're using 12" drivers for bass you should be able to comfortably cross them to a pair of 7" drivers, skipping the RS225's altogether and going for a three-way. Crossovers could be at say 200 and 1400 for this scheme.

If you wanted to cross the tweeter higher you could roll off one of the mids, (think 2.5 way mmt sitting on top of dual 12's) to prevent lobing between mids at the higher crossover point.

The only reason I went to a 4-way was that I wanted an (almost) full-range dipole for theatre use in a fairly large room, so I started with two 15" on the bottom end, and needed the extra (10" in my case) woofers to reach up to the midranges.

I'm very proud of my diy custom-built active crossover, but if I had to do it again I probably wouldn't. Which is to say, I'd use a behringer like you have. The results are great, but the headaches to build it were large - active crossover design and construction is for electrical engineers ..

I think you already own the makings of a great set of speakers. Your subwoofer will likely need to play a little a higher than the 25-30 Hz that you mentioned, especially if you're looking for big dynamics.

Paul

[Paul W]

Either a 3-way or 4-way will work, but here are some "20/20 hindsight" changes I would make in a 4-way similar to the one you are considering.



I wanted a very clean bass driver with very high breakup and ended up with the W26...fortunately it is very forgiving of overexcursion because I did overdrive them on occasion. Had the RS315HF been available at the time, I would have used it...and a larger baffle. Even with the RS315, 35Hz may be optimistic...up to 50Hz xo to an IB will sound fine, even with music.

The W22 is my favorite driver...so no change. RS225 should work well.

Instead of the D76, my do-over would use a "soft cone" upper-mid like an AT Flex or Seas M15CH001 crossed about 3k to a larger ribbon like the AC G1. Another alternative would be small stiff cones like the RS125 or Seas 12CY001 crossed at 2k to a very stout dome like the Seas Mill or Revelator. WMTMW provides a lot of dynamic range, so pick a tweeter with stones. I did NOT like broad vertical dispersion (OW1) in an MTM.

Finally, the baffle would have been wider; with roundovers of at least 4". I am convinced large roundovers are a contributor to clear sound.

If you have room for the WMTMW, you'll be rewarded with a very "real" soundstage.

Test Mules

[Mazeroth]

Thanks for the input!

Paul, I was beating myself up looking for your 4-way dipoles over at diyaudio.com. I literally spent 30 minutes searching and searching for them, when they were on this site the whole time! I really need to spend some more time over here!

I get what you're saying about keeping it simple (can I say that about a dipole?) and sticking with a 3-way. I just wanted to make sure the drivers were distorting as little as possible. I initially thought that because the RS180s are only handling 100-120hz and up that shouldn't be so bad. However, after plugging in some numbers in SL's spl_max spreadsheet, I'm not so confident.

The baffle I'm using will be roughly 13" wide and will more than likely have "wings" that go back 2-3". As far as the effective path difference, does this mean from the center of the driver to the center of the rear of the driver? If so, that should mean 6.5" for the front, 6.5" for the rear, and roughly 3" for the wings, for a total of 16" or 406mm. After plugging that info into the spreadsheet it gives me xmax at 100hz with 96db output, at 141hz 105db output, and I'm guessing at 120hz I'll have around 100db output, which is the highest I would probably cross them over. Granted, I'll have two RS180s so add 6db to that. I want great dynamic capabilities with minimal distortion, so do you guys feel these will be able to achieve this? My listening distance will be around 12 feet away, so I'll be losing roughly 15db of output from my distance. At 120hz, and 12 feet away, they'll be reaching xmax at 91db. Hmmm

Keep the info coming, and as always, thanks for helping the new guy out!
;x(

[JonMarsh]

I'd consider slightly larger baffles if you can handle it cosmetically, especially to make things easier on your mids. After agonizing a lot, I settled on 15-16" for the Arvo's, (with a 180 Hz crossover) and will use 16" panels with 3-4" wings angled back like Pauls above, for the new project (Isiris).

Keep us posted on your design and your progress. :T

~Jon

[engr_dave]

Wow Paul W,
Your speakers are along the lines of my thinking. :T I have been enjoying some Dynaudio WMTMWs with 24W100s, D76, and Esotar 330 for a dozen years now. . They don't measure all that flat, but I keep coming back to them because they image so well. These were a modification of the Dynaudio "Myrage" system. That alignment was vented, but not very well executed - and I didn't need the extension (sub, of course) - so I shrunk them & changed to sealed. I should fiddle more with the XO to eliminate some "bloat" in the upper bass/lower midrange, but I'm itching to build something new. You guys have me curious about dipoles. I thought about building Linkwitz' Orion, but I really like the WMTMW configuration.

I'm currently finishing up some little tower Dayton 3's for my dad (I had parts lying around)... and my next project in the queue is a dual opposing RL-p15 sub, sealed with LT... but then I'd be up for an upgrade to my mains. The Exodus Apex looks interesting also - pseudo-dipole with rear-facing fullranager. Should be capable of serious output with low distortion. I occasionally clip my Bryston 4BST (400wpc@4) into those fairly efficient (90dB) Dynaudios 8O - so that gives you some idea of the output capability that I am looking for. Bi-amp or multi-amp is OK, but I'd probably build my own analog active filters. Extra A/D and D/A converters in the signal path scare me. Budget would be <$3K for parts.

Thoughts?

[Paul H]

Thanks for the input!

I get what you're saying about keeping it simple (can I say that about a dipole?) and sticking with a 3-way. I just wanted to make sure the drivers were distorting as little as possible. I initially thought that because the RS180s are only handling 100-120hz and up that shouldn't be so bad. However, after plugging in some numbers in SL's spl_max spreadsheet, I'm not so confident.

The baffle I'm using will be roughly 13" wide and will more than likely have "wings" that go back 2-3". As far as the effective path difference, does this mean from the center of the driver to the center of the rear of the driver? If so, that should mean 6.5" for the front, 6.5" for the rear, and roughly 3" for the wings, for a total of 16" or 406mm. After plugging that info into the spreadsheet it gives me xmax at 100hz with 96db output, at 141hz 105db output, and I'm guessing at 120hz I'll have around 100db output, which is the highest I would probably cross them over. Granted, I'll have two RS180s so add 6db to that. I want great dynamic capabilities with minimal distortion, so do you guys feel these will be able to achieve this? My listening distance will be around 12 feet away, so I'll be losing roughly 15db of output from my distance. At 120hz, and 12 feet away, they'll be reaching xmax at 91db. Hmmm

Keep the info coming, and as always, thanks for helping the new guy out!
;x(

The path difference you're using seems right. Remember that while this is a useful simulation, it's guideline only, as SL notes on his site. You will definitely want to try some different cardboard shapes and sizes once you've settled on an approximate configuration before you cut up the good wood.

My inclination would be to cross those 12's higher in a 3-way - someone who's actually used/tested the RS woofers (I haven't) may chip in here, but I'd be looking at 160-200 Hz as the range where higher frequency distortion from the 12" are balancing with higher output distortion from the 7" drivers.

Of course, I'm really comfortable throwing out crossover suggestions at you, 'cause you can dial in a new one with the behringer anytime for zero cost and little effort

Paul

[Paul W]

I occasionally clip my Bryston 4BST (400wpc@4) into those fairly efficient (90dB) Dynaudios - so that gives you some idea of the output capability that I am looking for.

Yikes Dave! A session of that would set my tinnitus singing!

If you want to do that with dipoles, I'd first look for really good aluminum cone 15"s, shorting rings, low Le, breakup over 1k, silent pole vent, etc. (Actually, I'd like to know what 15" fits that description since I haven't researched them.) W22 lower mid, maybe your Dynaudio 24Ws...dunno. For your SPL, I withdraw the Seas M15 nomination in favor of the AT. No personal experience, but your existing tweeters may be fine. Largish baffle...as Paul H said see SL's website and use his calculator as a guide.
Paul

[ThomasW]

f you want to do that with dipoles, I'd first look for really good aluminum cone 15"s, shorting rings, low Le, breakup over 1k, silent pole vent, etc. (Actually, I'd like to know what 15" fits that description since I haven't researched them.)

Parts Express in coming out with a 15" RS woofer in the not too distant future. Don't have any specifics about it yet...

[Mazeroth]

Again, thanks for all the input! I really appreciate it.

I'm set on doing the 3-way with the RS315s, RS180s, and an RS28. I'll be going with a 14" wide baffle but am not quite sold on doing the angled wings. My woodworking skills are minimal and if I'm going to try crossing the RS315s a little higher (say 160hz) I don't think I'll need the added depth the angled wings can provide. Also, if these don't work out I'll just trash them and build new cabinets, no biggie :B

One question I do have deals with the placement of the MTM. I'd like to place the RS180s fairly close to each other and after some testing I can get their frames 1.5" apart with the RS28 in the middle, sticking out to the side about 2". Now, do I want to line the RS180s up in the center of the baffle, the RS28 in the center of the baffle, or keep the whole "unit" of drivers dead center in the baffle, meaning the frames of the RS180s will be, say, 3" away from the left side and the frame of the RS28 will also be 3" away from the right side?

Thanks!

[Paul H]

One question I do have deals with the placement of the MTM. I'd like to place the RS180s fairly close to each other and after some testing I can get their frames 1.5" apart with the RS28 in the middle, sticking out to the side about 2". Now, do I want to line the RS180s up in the center of the baffle, the RS28 in the center of the baffle, or keep the whole "unit" of drivers dead center in the baffle, meaning the frames of the RS180s will be, say, 3" away from the left side and the frame of the RS28 will also be 3" away from the right side?

Thanks!

That's one of those questions that's best answered "it depends".

I'd suggest that you check out one of the online baffle diffraction simulators. Set it up for your specific baffle and driver sizes, and experiment with different placements to obtain smoothest freq response.

Try this one here:

http://www.pvconsultants.com/audio/diffraction/downloadbds.htm

Paul

[Mazeroth]

I'm almost done drawing out the plans for the dipoles and had two questions for you guys:

1. In regards to the dual RS315 HiFi woofer mounting, as of right now the plans I drew up have them mounted in an H frame simliar to the Orions. I see the Arvo and a few others have the woofers mounted in the front baffle so both cones are visible. Is either one better?

2. The MTM up top. Right now I have it so the RS180s are 1.5" frame to frame and the RS28 is in the middle, protruding almost 2" out to the right (or left). After looking at the Modula and other projects it seems most people leave a large gap between woofers in their MTMs. Because I'm using a 14" wide baffle and can have the tweeter offset more, is there a large benefit to putting the RS180s close to each other? I will be using the DCX2496 so I'll have plenty of filters and EQ to play with.

Thanks a million!

[Paul W]

#1 The Orion woofer arrangement provides some benefit in distortion cancellation. With crossovers above 100Hz, I prefer the "normal" configuration to minimize cavity resonances.

#2 The layout you're considering will have an asymetric polar pattern. I keep the mids and tweeter in a vertical line and then, on narrow baffles, shift the line horizontally. Usually, it only takes 1/2" or so to make quite a difference in diffraction.

Use the BDS simulator at the link Paul H gave...also the polar response tools. You'll be able to simulate driver layouts, wings, roundovers, etc without a bit of MDF dust!

Paul W
(sleepless in Tokyo)

[JonMarsh]

Simulate 20 times, cut once. Oh, and measure a few times in between, too!

~Jon

[oneoldude]

I'd consider slightly larger baffles if you can handle it cosmetically, especially to make things easier on your mids. After agonizing a lot, I settled on 15-16" for the Arvo's, (with a 180 Hz crossover) and will use 16" panels with 3-4" wings angled back like Pauls above, for the new project (Isiris).

~Jon

I know the design is still on a napkin, but what are you thinking about re the woofer baffle. Are you planning to do like Paul and share the same baffle for the mid and woofer or do you plan to extend the baffle for the woofer like in the Arvo.

I guess what I am asking is that since you are describing what SL calls "D" (effective path difference) of about 12" for the mid, what is your napkin plan for "D" for the woofers?

Thanks