Naked driver dipole with different Seas tweeters

DEQ+TheEnd

Member
2005-11-11 1:44 am
I’m thinking of unboxing my SEAS Excel speakers to try out naked drivers on a chain like StigErik had going for a while.
My tweeters are the CRESCENDO T29CF-002. I don’t see myself sink this amount of money in another pair of domes as I then rather try out some true dipole tweeters.

Question is. How do you think the cheaper Seas T25CF001 will perform as back tweeter in a naked driver dipole setup with the CRESCENDO? Dome size are 29 vs 25mm.
I understand I have to address the difference in sensitivity. But will it degrade the CRESCENDO to same sound quality as the T25CF001. Or could the difference in performance be “camouflaged” as reflection from my not treated back wall?

http://www.seas.no/images/stories/excel/pdfdataheet/e0040_t29cf002_crecsendo_datasheet.pdf
http://www.seas.no/images/stories/excel/pdfdataheet/e0006_t25cf001_datasheet.pdf
 
Last edited:
In a 'naked' dipole the way to go for the tweeter section is either an AMT (expensive), a Neo3W (cheaper) or a pair of small neodymium domes (also cheap). It's important to minimize the source separation to avoid blooming of the polar pattern, which is impossible with dome tweeters with large magnets.
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
Thank you for input.

I should maybe clarify a little. The quest is to check out if its' possible to implement a DIY dipole solution in our living room in a manner my better half can accept. I already know I will not be able to do any visible room treatment. That's what kept me inside the box until now :)

Budget for the prototype should be as low as possible, but I realize I'm not going to this party for free. If I can make it work I can see myself with some BG Planar in the end.

My current "parts bin" for this prototype holds the following parts:
2 pcs Seas E0040 T29CF-002 - CRESCENDO Tweeter
4 pcs Seas E0018 W18E-001 Mid/Woofer
2 pcs Peerless XXL 12" Sub

3 pair Class-D amp modules from Hypex and Coldamp. All of them different :(
DCX/DEC2496 combo with mic.

Come to think of it there are a couple Esotec D260 tweeters laying around here also. Maybe I can use these for the initial tests?
 
Last edited:
Some things to consider:
If you have no control whatsoever of the front wall situation, it might be a good idea to show a picture (or any helpful illustration) of that wall before starting the project. Placing a dipole too near to a highly reflective wall is no good idea.

The dipole-8 is quite sensitive to SPL differencies between the front and rear source. Anything more than 3 dB difference will be critical. If you see a need to attenuate the rear driver 6 dB (half as loud as the front driver) you have lost the figure 8 almost completely.

Naked drivers are the least forgiving dipole implementation. Are you able/going to do in-room measurements?

Rudolf
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
Thanks Rudolf. I'm currently reading your thread "On the directivity of dipole tweeters". Very informative although I'm not smart enough to grasp it all yet :)

Am not clear if the size of the figure 8 are related to the size and shape of the driver/baffle rather than frequency depended? I've always thought the later, but lost my bearing reading your thread?

Are you able/going to do in-room measurements?

I got a mic and preamp, but missing software and knowledge so any pointer to the modern way of doing it is appreciated.
 
Last edited:
I'm currently reading your thread "On the directivity of dipole tweeters". Very informative although I'm not smart enough to grasp it all yet :)
So we are on the same level: I wasn't smart enough to tell the story in a way that everybody can grasp it. :eek: :)
Am not clear if the size of the figure 8 are related to the size and shape of the driver/baffle rather than frequency depended? I've always thought the later, but lost my bearing reading your thread?
Neither nor - or both. Depends on the way you look at it.
The "figure 8" is related to the radiation pattern of a dipole. This pattern should have maximum SPL to the front and back, and SPL "nulls" to the sides. Ideally the shape of this "8" should be the same for all frequencies. To achieve that, baffles should be small (compared to the wavelength involved). If you mount your tweeters back to back, you create quite a distance between the front dome and the rear dome of that pair. This distance is equivalent to the radius of a baffle, where you would mount a Neo3 (for instance) on, which has no distance between the front and back side of its diaphragm.
That's why you no longer see such bulky tweeters in the newer efforts of johnK or SL.
I got a mic and preamp, but missing software and knowledge so any pointer to the modern way of doing it is appreciated.
Learning loudspeaker measurement on this project is like learning to swim in the breakers. :eek:
Not sure if this is the best way to get ones feet wet.

Rudolf
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
Thanks Rudolf! I will not lay any fault on you sir. It's all mine doing :) BTW, your cook book "Open baffle dipoles - How they work" is great reading. Thank you for sharing.

One thought though. I'm wondering how a dome that fires into a disperser will perform in a dipole? It should eliminate some of the problem with dual sources? But might introduce other problems even harder to work around?
If it's worth testing I'm thinking one can tilt it in different position if needed.
Top mounting might not be the best place?
Experiment with direction. The dome might work better pointing down or to the side depending on the efficiency of the disperser?

[IMGDEAD]http://www.luxury-insider.com/uploads/news/2011/02/trenner-friedl-duke-loudspeakers-3.jpg?width=600[/IMGDEAD]

Regarding room measurement I've only used the internal DEQ sw for guidance to tweak eq. Anyone here experienced with REW? Is it worth investing time to learn this application, or can I do better with my time?
 
A Man of Character marries one woman for life, but fools around with several different speakers :)

YOU MIGHT:
1) Keep your current speakers and just remove the Peerless XXL 12" for dipole experiments. The Seas T29CF and W18E are not "great" candidates for a SOTA dipole like the JohnK Nano Note II or Linkwitz LX521.
2) Admit that you will never "really" be happy until you experiment with a 4-way dipole similar to NoteII or LX521.
3) Design an H-frame for your current pair of Peerless XXL 12" and cut wood for two for future stereo.
--- FIRST BUILD... one H-frame with both Peerless XXL to test if your ears love dipole bass in your room.
4) Design a "paddle" shaped baffle that can be used for (8" or 10" midbass) + (3"-4" mid) + (dual domes or BG3 planar). Cut wood for 2*3=6 baffles which you can later route for different speakers.
5) Add some European engineering flair by developing a hanging baffle design. The LX521 wastes 2" in width for the mid-tweet isolation stand-off frame.
6) Admit that you are in love ... with dipole bass ... and set a budget for midbass+mid+tweet. Admit that you are kinky enough to try a 4-way!
 

Attachments

  • 4way Dipole.jpg
    4way Dipole.jpg
    51.5 KB · Views: 185

DEQ+TheEnd

Member
2005-11-11 1:44 am
A Man of Character marries one woman for life, but fools around with several different speakers :)

--- FIRST BUILD... one H-frame with both Peerless XXL to test if your ears love dipole bass in your room.

Haha :D Well spoken. It's so lyrical someone should make a melody to go with those words.

You are reading my mind regarding H frames. Currently browsing the web in pursue of design rules. I've seen some measurement where frame depth is the same as with/height. Both based on practical size for the given speaker. i.e 2x12" woofers give cabinet size 13"x13"X26". Not sure if this i in the right ballpark?

More on the dome into a disperser. First I was thinking "8" pattern would look very strange as the dome vary in size all the way around. But then I realize same is in some extend true for back backfiring tweeters which also take the shape of a ball when mounted back to back.
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
1) The Seas T29CF and W18E are not "great" candidates for a SOTA dipole like the JohnK Nano Note II or Linkwitz LX521.

Reading JohnK site it seems like I got an answers to my initial question. "Can I mix front and rear tweeters." JohnK does so in the 2012 version of NaO II Rs. Where he use more expensive parts, but also a little different design idea compared to the Note II.

To summarize my understanding so far. When using large dome tweeters for dipoles the trick is to actually use a baffle rather to than hang them up naked.

Another thing I've noticed so far on this journey is that nether JohnK nor Linkwitz use dipole tweeters although both of them seems to know these devices well? Why so, they seems to be a favorite design "rule" among most diy? The budget for NaO II Rs and Orion could fit a planer or ribbon. Is it the often needed EQ that complicate the commercial product that make them stay away?
 
Last edited:
.... your cook book "Open baffle dipoles - How they work" is great reading.
If you dare to try some German, my latest German version of that cook book is more up to date. When I looked into Norwegian loudspeaker forums, I was surprised how much I actually did comprehend. And I found that you guys have a very nice and relaxed sort of humour :D. Maybe it works the other way too. But certainly not for the humour :eek:.

Just to illustrate what tewinkel was writing in post #2, I simulated the geometry of a Neo3 in Boxsim and compared it to a pair of CRESCENDO tweeters mounted back-to-back. Frequency response in infinite baffle was assumed to be ruler flat for both cases. First the fake Neo3 - which I own and have measured. The 0° dip at 7-8 kHz is WAY deeper in reality than shown in the simulation:
comp_Neo3.gif
Now compare with the paired Crescendos:
comp_T29CF.gif
It is absolutely helpless to try those drivers in such a constellation.

LineSource has made some fine comments about ways to go for you. I could not have said it better.

Regarding room measurement I've only used the internal DEQ sw for guidance to tweak eq. Anyone here experienced with REW?
ARTA is known to NOT being easy on the beginner. But it has quite a following at least in Europe. You may look if you find some support for ARTA in your country.

Rudolf
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
I find websites on your native language very informative. And they are plentiful. But I have to get help from translate and the text google present to me rarely let me see the person behind the keyboard. Any jokes are well hidden :)

Those graph are very informative. Thank you. I noticed it flatten out at 60 degree. Could that be explained by the ball theory I had in post #11? At his angle the dome is at it's closest to the oposite dome? The more on axis the longer distance between sources. Or is it just the tweeter front plate isolating the two sources? I forgot CRESCENDO is actually a horn loading device. I guess doubling this device for dipole is even harder than normal domes.
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
Of course not!!! Sorry for rambling.

A dome set flat can in no way compare to two domes back to back. A singe dome set flat will put out energy in phase all the way around. Back to back tweeters are normally connected 180 degree out of faze.
Again, sorry for rambling.

...but then again, it could be interesting to see how much it mess up the sound. I is known that some people with wide baffles prefer shooting both tweeters in faze.
 

DEQ+TheEnd

Member
2005-11-11 1:44 am
Unfold H-baffle

After a few hours fiddling with Edge I think I’m starting to understand how to navigate this software.
If my understanding of Edge are correct W18E-001 aren’t that bad for use naked. 176mm diameter should give me a dipole peak around 1500Hz. Using two of them below this peak with some EQ might have enough output to work well down to meet a couple of 15 inch woofers in H-frames.

One question thought. How to unfold a H-baffle for use as input parameter in Edge?
Is it as simple as width + depth and height + depth?

Thanks!