Dipole experts advice needed...

[Jonasz]

I'm in the making of a high end dipole system. My problem is what configuration and drivers to use. The crossover will be an active 4th order LR, yet to be designed. I like a really big airy soundstage with and effortless sound with great dynamics.

After years of searching for drivers I've come up with these:

2 Hiquphon OW1
2 Dayton RS28A
2 Seas H1212
2 Seas Excel Crescendo
2 SS 9900
2 XT25
2 Excel W15CY
2 Excel W18EX
4 Excel W22
2 Seas L18 (H1224)
2 Vifa XT18WO
2 SS 8565-00
2 Dayton RS10"
8 Peerless XLS12"

I've almost ( :roll: ) decided that the subs will be Linkwitz Phoenix woofers with 4 XLS per channel. After that I'm open to suggestions...

Some configurations I've thought about:

WWMT with W22, W15 and H1212/RS28/Crescendo (waste of money only using the W22 up to 3-400 hz)
WWMT with RS10 or SS8565, W18/W22 and H1212/RS28/Crescendo
WWMTM with RS10 or SS8565, W22 and H1212/RS28/Crescendo

I can think of more than those...

I know I may have to buy a couple more of the 8565 or RS10 if I go with a WWMT or WWMTM with either of those. What I want to know is wich configuration and drivers you think will meet my goals best and wich crossoverfrequencies to use.

Regards,
Jonas

[Paul H]

I assume you've seen the thread on the dipoles I just finished, but in case you haven't:

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

There's a bunch of information and opinions from various people in there, including of course my selection of preferred drivers

Paul

[JonMarsh]

Paul's suggestion is very good. With all the drivers you have, it becomes a matter of "how big?" and "how complex" do you want to go?

If you want a tweeter like the Hiqaphon on top, due to it's extension and dispersion, you won't be crossing directly to W22 in all likelihood. If you want a "simpler" system to design and build, it's a matter of picking the drivers for the best tradeoffs.

If you plan on H or W baffles, you've got to have drivers that will work to 110-120 Hz or so on the "midrange, to keep the dipole bass units out of their problematic range.

One approach that could work with reasonable SPL capability is an "upper module" like the top section of the Arvo- dual W22 with the H1212. That could be crossed over with the right EQ and network to the Peerless 12's with some care. This would be only a 3 way system, which isn't too bad to manage.

Another approach, throwing simplicity out the door, is the Hiqaphon on top if you like it, the W15CY or W18 on the upper midrange, the RS270's on the lower midrange/upper bass, and the Peerless in separate cabinets on the bottom end. How complex do you want to get?

[Jonasz]

Thanks Paul, read throught the whole thread and it was very interesting reading, got some good (I hope ) ideas I will try to incorporate.

Btw the resulting speakers are veeery nice indeed! :T

[Jonasz]

Jon, I will use the XLS12 w-baffles up to 100-120hz. I like the idea of an Arvostyle topmodule with dual W22's and a H1212, but as Paul said, the twin 10" give a very effortless and clear bassound wich I like very much. I think the 10's will give some nice extra headroom... :twisted:

As much as I love the Hiquphontweeter I don't think the W18 can match the OW1 >2,5khz crossoverneeds. I know Dennis M uses the combination with success, but according to the measurements I've seen the W18 needs to be crossed below 1,8khz to give it's best. I do want a speaker that uses the drivers within their best range.

What do you guys think of this combo, WWMT, with RS10, W22 or W18 and H1212? Crossed at 300 hz and 1,4-1,6 khz depending on the mid. Btw, are H1212 considered the best tweeter of the above listed?

My goal is to do an all active (analog) filtersetup, but again, nothing but the basscabinets are decided so far... :P One major advantage of the active filter is that you can test different tweeters easier.

[Paul H]

Jonas,

Thanks for the compliment.

Disregarding for the moment that the thread title requested experts advice here's a few comments from my bit of experience with my dipoles:

To build a dipole that reaches down to 30-40 Hz at high volumes requires major driver displacement/volume - I have 2-15" per side. Using 4 XLS per side will give you some great capacity for effortless dipole sub-bass. You could build a very good speaker with 2 per side but I'd say if you have the 4 per side use them.

Then if you wanted to stay (relatively) simple you could cross from the XLS's to a pair of W22's per side at say 160-200 Hz, and cross to the tweeter at 1200-1500 Hz. This would result in a three-way something similar to Jon's arvos.

A four-way is considerably more complicated and places serious reliance on the crossovers being very well done, but it allows the drivers to cruise comfortably only in their best frequency range.

If you're seriously looking at a 4-way the W18 as midrange will comfortably go down to 300Hz, and will be slightly less directional and easier to cross over at the upper end than the W22.

What do you listen to? How loud? What size room?

What do you have/plan to have for testing/measuring/designing crossovers?

Paul

[sfdoddsy]

Btw, are H1212 considered the best tweeter of the above listed?

I'm not sure that 'best' is entirely appropriate. 'Best value' might be closer. I've still heard anything to beat the Seas Millenium, but then again I've been too busy enjoying for much fiddling lately.

Steve

[Dennis H]

Zaph's measurements seem to agree pretty well with Mark's. The Mill has more higher order distortion products than the 1212.

Seas Millennium:



Seas 27TBFCG (H1212):

[JonMarsh]

Well, another "Ultra Arvo II" idea was a WMTMW or MTMWW with dual RS265HF, and dipole subs like the Beethoven grand (Peerless XLS, TC2+, or RS315HF). But it do get a bit more complex. Actually, I've been playing with that idea in the back of my head, using the 883 as the mids, and not having a U baffle or H baffle on the 10's.

I'm planning a measurement "shoot-out" which won't so much be any new knowledge for the "community" as it will be a chance for me to exercise myself with both Praxis and CLIO- up for eval will be the RS225 vs W22 vs 830884, the 830883 vs RS180 vs 850439 vs Eton 370 (cause I have some), and TC2+ vs RS315HF, and RS270 vs RS265HF. I'm not buying anymore stinking drivers until I finish BUILDING something.

And at this time, I would like to designate Paul H our official forum "expert" on dipoles, based on the workmanship, effort, and completion rate of his current design! :T :T

~Jon

[Victor]

Jonasz,

Well..., my system uses two W-frame dipole woofer cabinets with a total of four XLS drivers per side, just like you are planning to build. Each woofer cabinet is exactly as Linkwitz described them on his site. See here I am using a stacked arrangement with 2 woofers cabinets facing in opposite directions. Here are some of my observations that might be useful to you.

I must say that having 4 XLS drivers per side results in a quite remarkable and I would even say grandiose bass. Above 40 Hz the stacked dipole woofer system is good to more then 100 dB, perhaps around 105 dB or even a little more. I did some very cursory measurements with SPL meter a while ago. The quality of bass is absolutely first rate and I would not want to go with any other woofer system at this time. Down side, - it is big. Here is the picture

I must however caution you that the SPL does drop below the 40 Hz mark, as PaulH correctly outlined in his response, although it is still very respectable at about somewhat around 95 dB at 20 Hz. Again, those numbers are from memory. For many people this would be sufficient, but I like organ music at very realistic levels and also my system doubles for the HT use. With this in mind, I must say that having 4 XLS woofers per side is not sufficient for my application.

At this time I am looking at augmenting the below 20 to 40 Hz frequency range with a sealed subwoofer, - one per side.

Smile comes to my face every time I look in the corner of my storage area where unbeknownst to my wife I am hiding a 15-inch Avalanche driver! One has arrived and one is en route! Once I place those Ave15s in 4 Ft3 boxes, one of the two things will happen, - either I will be permanently exiled back to Siberia or I will finally achieve what I would call realistic levels of low frequency sound, regardless of the CD or the DVD that I might throw at my system. Time will tell, but in all likelihood I’ll be back with questions regarding the construction of the sealed sub.

I think the considerations as I described them are worth looking into before you invest time and expense of building a stacked dipole W-frame woofer system.

Another point, - my system in its current incarnation uses a totally digital cross-over electronics, including a volume control. However, the W-frame woofers based on XLS drivers still required me to use an analog circuit to accomplish the EQ. I found that the commonly available digital boxes such as DCX2496, that I use, do not have the necessary headroom. If you are planning a digital approach to crossover design, perhaps a different choice of drivers that do not need as much EQ as XLS might be of benefit to you. Personally I’ll be retiring my DCX2496 and the DEQ2496, once I get a stand-alone DSP-based circuit running with FIR filters sometime next year, at which time I am hoping to fix the headroom problems for the XLS woofers.

Regards,
Victor

[Jonasz]

Paul, our newly promoted dipole expert: I will follow you advice and go for a threeway. I understand that a fourway may still be a little out of my reach, I better begin with some easier stuff and learn along the way.

I think that an Arvo style topmodule may be a good start. But I do want to have active filters so I easier can test different drivers. I used to own a pair of Phoenix speakers and I still have the crossoverboards to play with. When I'm gonna use w-baffles I think that 4th order LR at 120 hz is the highest crossoverpoint I can use without getting into trouble. In the other end maybe 4th order LR is a good obtion too? Or is there some way to transform these puppies into those magic CE 48 db filters Jon plays around with? :P

I do have LspCAD 5.25 wich I've never used... ops: Maybe this is the time? No idea how to though... :B

Victor: Great speakers you have there! Interesting to hear your opinions. I do think 4 xls per side will be enough, at least to start with. If not, I will have to add closed woofers just like you are doing... Btw, I showed my girlfriend the pic of your speakersystem and she didn't even think they were big enough! 8O I guess I'm lucky... :B I think I will disappoint her, mine won't be half the size of that... 8O :B

[Paul H]

Jonas and Jon,

My primary area of expertise in speakers is in selecting projects to design/build that are large, complex and way beyond my knowledge and capabilities when I start them

Nonetheless, a few more thoughts and food for thoughts:

The three-way is probably a good way to go, and as Victor noted even that will likely turn into a 4-way including a separate subwoofer for the 20-40 Hz octave.

I was targeting 105 dB theoretical, and 2-15" woofers will do that down to about 40 Hz. Four 15" woofers only go to the low 30 Hz range, and 8 of them would be needed to hit that volume at 24 Hz. Before you know it you're building speakers with a dozen drivers (oops sorry Thomas :B )

You could cross the XLS's a little higher than 120 Hz and take some strain off the 8" - but then you should look at a U or shallow H-frame.

Active is certainly the way to go for the low end, as equalization will be required for the woofers, and the cost/size of passive crossovers gets prohibitive.

If you're interested in an active cauer elliptic filter it appears difficult - see my other referenced thread for a few comments on that. You could consider a passive mid/tweeter xover combined with an active lower crossover. I used a 4th order (active) electrical filter on the mid and tweeter, along with a notch filter on the mid to tame the famous resonances.

I started with textbook 4th order LR's, and then modified the component values in LSPCad to get the response I wanted.

BTW, Yes, you do want to start up and use LSPCad ..



Paul

[JonMarsh]

But upgrade to 6 Standard as soon as you can... it's well worth it.

~Jon

[Jman]

Victor-

your system looks very familiar

Your 2nd av15 still hasn't arrived!? My 4 av18's arrived in a 7-14 day window via UPS non-priority (had to phone those AA goofs twice to get a refund as I paid for air-priority).

Jesse

[JoshK]

Can you still get the AV15? I thought it was all sold out. I'd love another to match the one that I have. I didn't think ahead and bought only one.

[Jman]

No they sold out months ago, hence why it's so ludicrous that Victor hasn't received his yet. I waited nearly 2 months from phone order to receiving them. 1 would still make a nice sealed or ported sub.

[mikec]

Zaph's measurements seem to agree pretty well with Mark's. The Mill has more higher order distortion products than the 1212

Curious if you've actually listened to both.

[JonMarsh]

An interesting point which I plan to investigate further around Xmas- I have a speaker I'm finishing up for the bedroom which I'm going to leave the crossovers external and try both. Actaully, I've got a couple of other tweeters I'm looking at for it, too, but some aren't commerically available anymore.

There's a theory about that some folks like the sound of tweeters wtih more high order harmonic distoriton becuase it lends more "air" to the sound- this occurs with many ribbon designs. I'm pondering getting a moderately high end ADC unit just so I can do my own "live versus Memorex" tests, as I have access to a drummer I can record live, and own a few reference microphones.

[Davey]

Jon,

It's the old cliche' (that I've heard you use a time or two).....One mans distortion is another mans "air." Or vice versa.
How all the relative objective measurements add up to audible differences is unclear because the evaluations are subjective.....Which is basically the opposite of objective. It's kind of a Catch-22.

Cheers,

Davey.

[JonMarsh]

I'm sure Rossarian would agree with you completely- and since freedom of choice is our country's guiding principle, to each his own, right?

OTOH, I'm the type of curmudgeon that mainly reads reviews that have some accompanying measurements (like Stereophile), becuase in my heart of hears, I hope/believe that they can be correlated to some significant degree- if we're doing the right measurements.

~Jon