Greg
That's right! I forgot about Piega. I was in switzerland recently and wanted to visit their showroom, but never got around to it.I asked Igor a while back if he ever thought about designing a "Piega" style point source drive unit. His responded by saying that it was an interesting design, but would be prohibitively expensive to produce. He also said compressing and separating the mylar tweeter section in the center of the unit would be sub-optimal for the best midrange perfomance.
But man, I would love to hear that dipole version!
Seth
That's right! I forgot about Piega. I was in switzerland recently and wanted to visit their showroom, but never got around to it.I asked Igor a while back if he ever thought about designing a "Piega" style point source drive unit. His responded by saying that it was an interesting design, but would be prohibitively expensive to produce. He also said compressing and separating the mylar tweeter section in the center of the unit would be sub-optimal for the best midrange perfomance.
But man, I would love to hear that dipole version!
Seth
Last edited:
Here's an interesting article with pictures of the Piega coax in various stages of assembly. The cool Swiss dude is the guy that assembles all the planar magnetic units.
piegaloudspeakers
piegaloudspeakers
Here is another article with the larger coax being assembled..Sorry, It's actually the smaller one..
http://www.piega.ch/download/Fidelity42_Piega.pdf
http://www.piega.ch/download/Fidelity42_Piega.pdf
Last edited:
Very interesting comment. I use 18Sound 6Nd430 which Greg tested and did try it without baffle. I didn't like the sound of this driver baffle-less compared to when it was mounted on a 20 cm wide baffle. This is a bit contrary to what proponents of the 'Constant Directivity' trend advocate. But there might be some things I overlooked for an explanation that you have a clue to.
My system: http://www.diyaudio.com/forums/multi-way/123512-ultimate-ob-gallery-47.html#post2433427 .
/Erling
Erling
This design could work very well if each driver was offset mounted in a tiny rectangular baffle with golden, or at least Fibonacci ratios. For example, lets say you are dealing with a 7" driver. You could offset mount it in a baffle that is 7.5" x 12"(Obviously works the same for cm) The baffles length dimensions would actually be placed horizontally(but far enough away from each other to breath. If they are placed too close together, they effectively become one giant baffle) for closer driver spacing... The ultimate goal for a baffle would be to have every part of the outside edge be at different distances from the acoustic center of the driver.(Impossible? Yes. but that's the goal..) The best possible shape for a baffle would probably be a golden ratio nautilus starting at the edge of the driver, and working it's way out..
Seth
Last edited:
Erling
Sorry about that... The reason that I brought that up is because I assumed(wrong) that if you didn't realize that a round raw driver acted as it's own circular baffle, that maybe you also weren't aware of the effect of baffle shape on driver response. My bad.
Seth
The reason they sound bad? because the defraction effects aren't offset by different distances. It basically adds up to one big off axis difraction ripple at the same frequencies, in all directions..
seth
ErlingSeth,
To ascertain that we talk about the same phenomenon. What you call 'the lower cut off point' I suppose is what commonly is referenced as 'the dipole peak' which is dependent upon baffle size, and the effects of, also of baffle appearance and speaker placement.
/Erling
Using a 7" driver again as an example, the lower acoustic cut off point would start occurring when the frequency length exceeds the diameter of the drive unit. In this case, around 1,900 hz, right in the middle of the ear's most sensitive range. Below that point, on an open baffle, you get a smooth, even 6 db/octave roll off. What happens just above that point? The best analogy for this that you can actually see, is the zaph commentary on surface mounting tweeters with 4" diameter plates. Not exactly the same as an open baffle situation, but it is enough to illustrate the point..
Seth
Seth,
We are talking about the same thing. I thought perhaps you meant something different with your term.
I would rather have called it 'upper cut off point' I think.
/Erling
Erling
Do I have my phraseology screwed up? I see the upper cut off point as where you choose to "electrically" low pass the woofer. The lower cut off point(In this case) is when the 7" drive unit has an inevitable 6db/octave acoustic high pass roll off starting at 1,900 hz due to the driver's rear wave acoustically canceling out the front wave. The only way to compensate for this acoustic roll off is to use 6db/octave diametric equalization below 1,900 hz..Not a good thing.
Seth
Last edited:
Erling
Here is an excellent linkwitz article that illustrates exactly how a round open baffle shape effects the frequency response. Diffraction from baffle edges The key phrase regarding the "transition region" that I was referring to is in the last paragraph of the article..
Seth
Here is an excellent linkwitz article that illustrates exactly how a round open baffle shape effects the frequency response. Diffraction from baffle edges The key phrase regarding the "transition region" that I was referring to is in the last paragraph of the article..
Seth
Last edited:
Seth,
Yes, Linkwitz is a good source of information for OB.
I'll give you another link to Rudolf Finkes's site, Dipolplus. He has some good dipole papers in German and English in the download section: Offene Schallwand (OB) . The German one is covering a little more ground than the English one. We are now just talking past each other with some different terminology.
/Erling
Yes, Linkwitz is a good source of information for OB.
I'll give you another link to Rudolf Finkes's site, Dipolplus. He has some good dipole papers in German and English in the download section: Offene Schallwand (OB) . The German one is covering a little more ground than the English one. We are now just talking past each other with some different terminology.
/Erling
On the topic of Neo10s and naked drivers: http://www.diyaudio.com/forums/multi-way/175745-three-way-nude-swinging-dipole-thread.html
The Neo 10 and 3 are perfect for this type of "nude" speaker and I'm sure(maybe) it sounds as good as the measurements would indicate. The really difficult thing I have with this design though, is that the Neo 10 would begin (In real life) its inevitable acoustic high pass roll off at 2,700 hz due to its 5" width. That is, in my opinion, a very high price to pay for saving money on a router. To each his own...
Something about the measurements troubles me though. It appears that he is compensating for the acoustic roll off of each drive unit with some type of eq. Dsp perhaps? Anyway, there is no way that these drive units(especially the bass driver) could possibly go that low without a huge amount of low end compensation, which comes of course, at the expense of very high non linear distortion. Hmm..
Seth
Last edited:
As a perfect example of what I am talking about..
Jamo Reference R 907 loudspeaker Measurements | Stereophile.com . As you can see, the mid-range driver starts its 6db/octave acoustic high pass roll off at 600 hz (Its not a perfect 6db roll off, because the drive unit's resonant peak is pulling it up a bit..). This is with a 23" wide (Including thickness) baffle!
Seth
Jamo Reference R 907 loudspeaker Measurements | Stereophile.com . As you can see, the mid-range driver starts its 6db/octave acoustic high pass roll off at 600 hz (Its not a perfect 6db roll off, because the drive unit's resonant peak is pulling it up a bit..). This is with a 23" wide (Including thickness) baffle!
Seth
Last edited:
Yeah, running drivers nude like that requires "dipole compensation" to get a linear response from the drivers.....it seems that the Neo10 in that app would be stressing at the low end, but I get the feeling that the designer in that thread is quite meticulous in his work.....
The real reason for doing this is not to save on the router, but to get the baffle as small as possible. The upshot is the driver can be used higher before it begins beaming, but the eq needed to get a flat response can be quite a bit, depending on the driver.
I've got 4-way dipoles set up this way, with the upper 3 way "nude" consisting of an 8" cone, a 3" cone (looking at a Neo8 to replace this), and a Neo3. To me, a 4-way is best here to keep each driver in its optimal passband and to reduce excursion and power demands at the lower end due to the dipole compensation.
Always tradeoffs!
The real reason for doing this is not to save on the router, but to get the baffle as small as possible. The upshot is the driver can be used higher before it begins beaming, but the eq needed to get a flat response can be quite a bit, depending on the driver.
I've got 4-way dipoles set up this way, with the upper 3 way "nude" consisting of an 8" cone, a 3" cone (looking at a Neo8 to replace this), and a Neo3. To me, a 4-way is best here to keep each driver in its optimal passband and to reduce excursion and power demands at the lower end due to the dipole compensation.
Always tradeoffs!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.