Something stirs in the Undergrowth
-
- User
- Posts: 141
- Joined: Tue Dec 17, 2013 3:48 am
- Location: Northern Colorado, USA
#526
* Pimmboxes, or perhaps more accurately, Open-Back Boxes, or OBB. As mentioned above, the source of the baffle peak and the 6 dB/octave LF cancellation below the baffle peak comes down to the vector phase angle (at the listening position) between the first-arrival from the front of the driver and the rear wave diffracting around the baffle (or edge of the speaker basket). The problem is the magnitude of the two waves are nearly equal over the working frequency range, so the rotating phase angle between the two creates a comb-filter effect, with peaks and nulls.
Reducing the magnitude of the rear wave by even a small amount, say 2 to 6 dB, greatly reduces the depth of the nulls and height of the peaks, since the two waves no longer have equal magnitudes. The rear wave is now always smaller. Simple example: at zero frequency, with no rear attenuation, the two waves exactly match, resulting in total cancellation. With a modest 6 dB attenuation of the rear wave, the vectors, although they have a phase relation of 0 and 180 degrees, are magnitude-mismatched by 6 dB (1.0 and 0.5 in voltage terms), so even at zero frequency, the level at the listening position is only -6 dB down.
This is how the OBB works. Fairly modest absorption of the rear wave reduces the magnitude of the baffle peak and the slope of the LF rolloff. This is not the same as increasing the size of an OB, which simply moves the baffle peak(s) and LF rolloff to lower frequencies, while leaving the magnitudes of the peaks and dips alone (although room interaction would change).
If you had a perfect acoustic resistor, a cardioid radiation pattern would be possible, but with real-world absorbers like long-fiber wool and recycled cotton fiber, the radiation pattern changes with frequency, although I would not expect the omnidirectional pattern of a closed-box nor the perfect dipole of an OB. With real absorbers, there's probably a small rear lobe that varies with frequency.
Downsides: in conversations with Alexander of RAAL, he is convinced there are no linear acoustic resistors. He may have a point. I think the reality is in-between; some absorbers are more linear than others. Subjectively, foam is one of the worst; the Ariels are audibly degraded if even a few cubic inches of foam are used in the transmission line. Recycled cotton like UltraTouch or long-fiber wool (that's been moth-proofed) have much better subjective properties. (It's also important to keep the filling at least several inches away from the cone, otherwise the filling starts to couple to the cone, and efficiency and dynamics are noticeably degraded.)
I think the real tradeoff is tolerance of the inherent comb-filtering of a true OB versus potential coloration of a OBB. Commercial OB systems frequently have brickwall filters above the baffle peak, along with 6 dB/octave boost EQ below the peak; Linkwitz's systems are good examples of this approach. A more "natural" approach are high-Q speakers combined with an OB size that complements room reflections; I'd the OB to be fairly sensitive to rotation angle, since that affects the phase of the direct sound versus the 7 phantom sources.
Reducing the magnitude of the rear wave by even a small amount, say 2 to 6 dB, greatly reduces the depth of the nulls and height of the peaks, since the two waves no longer have equal magnitudes. The rear wave is now always smaller. Simple example: at zero frequency, with no rear attenuation, the two waves exactly match, resulting in total cancellation. With a modest 6 dB attenuation of the rear wave, the vectors, although they have a phase relation of 0 and 180 degrees, are magnitude-mismatched by 6 dB (1.0 and 0.5 in voltage terms), so even at zero frequency, the level at the listening position is only -6 dB down.
This is how the OBB works. Fairly modest absorption of the rear wave reduces the magnitude of the baffle peak and the slope of the LF rolloff. This is not the same as increasing the size of an OB, which simply moves the baffle peak(s) and LF rolloff to lower frequencies, while leaving the magnitudes of the peaks and dips alone (although room interaction would change).
If you had a perfect acoustic resistor, a cardioid radiation pattern would be possible, but with real-world absorbers like long-fiber wool and recycled cotton fiber, the radiation pattern changes with frequency, although I would not expect the omnidirectional pattern of a closed-box nor the perfect dipole of an OB. With real absorbers, there's probably a small rear lobe that varies with frequency.
Downsides: in conversations with Alexander of RAAL, he is convinced there are no linear acoustic resistors. He may have a point. I think the reality is in-between; some absorbers are more linear than others. Subjectively, foam is one of the worst; the Ariels are audibly degraded if even a few cubic inches of foam are used in the transmission line. Recycled cotton like UltraTouch or long-fiber wool (that's been moth-proofed) have much better subjective properties. (It's also important to keep the filling at least several inches away from the cone, otherwise the filling starts to couple to the cone, and efficiency and dynamics are noticeably degraded.)
I think the real tradeoff is tolerance of the inherent comb-filtering of a true OB versus potential coloration of a OBB. Commercial OB systems frequently have brickwall filters above the baffle peak, along with 6 dB/octave boost EQ below the peak; Linkwitz's systems are good examples of this approach. A more "natural" approach are high-Q speakers combined with an OB size that complements room reflections; I'd the OB to be fairly sensitive to rotation angle, since that affects the phase of the direct sound versus the 7 phantom sources.
#527
I used a drill press, and made successively larger holes until the desired diameter was reached. The material chips fairly easily, but I didn't have major problems. I made the holes a little bit oversized to allow a bit of room for adjustment.IslandPink wrote: did you have any difficulty re-drilling the Azura flange for the 4 x 1/4" bolt pattern of the Radian drivers ?
Don't remember if I asked Martin for advice before proceeding, but would definitely do so next time.
So many tubes, so little time...
#528
I built my version of the "beyond the Ariel" project three years ago, actually. After getting as far as I could with optimizing the crossover with Lynn's help (emailing measurements and talking on the phone), I stopped making revisions. My plan was to wait for Lynn's prototypes to be built, so my project could benefit from his experiences.Lynn Olson wrote: Gary Dahl is reviving a project set aside about a year ago, this time with Altec/GPA Alnico-magnet 416 and 515 drivers, as possible alternates to the Acoustic Elegance TD15M's. Gary's approach is to have either a TD15M or Altec/GPA 515 (Alnico) cover the range from 700 Hz to somewhere around 100 to 200 Hz, and a proper LF driver like the AE TD15H with passive radiators for the 30 Hz to 100~200 Hz range.
This is the configuration that I listened to for the past 3 years:
Upper: Azurahorn AH-425, Radian 745-PB. Third-order passive high pass at 700 Hz, impedance correction. Autoformer attenuator by Dave Slagle, with shunt.
Middle: Acoustic Elegance TD15M in 85-liter sealed cabinet, -3 dB at 75 Hz. Second-order passive low pass at 700 Hz, impedance correction.
Lower: Acoustic Elegance TD15H in 141-liter cabinet with two Acoustic Elegance PR-700 passive radiators, one on each side. Fb is 23 Hz, F3 is 43 Hz. Driven by plate amp with integral crossover.
The top end was occasionally augmented with Aurum Cantus G3 ribbons.
Lynn eventually worked on a first prototype with a friend in Texas, which was a 2-way system. He found the NEO version of the Radian 745 to be significantly better than the ceramic-magnet version. He didn't use the TD15M, but was very impressed with the GPA 416 (alnico version). Lynn also got better results using a fourth-order high pass and third-order low pass.
I decided to try NEO drivers, but waited till the Beryllium versions became available. I also wanted to give the GPA 416 alnico a try, but ended up ordering the 515's as well.
My first experiment was to try the new Radian 745NEO/Be drivers. These made a tremendous improvement. Tone colors are much more natural, and the background is blacker. Top end rolls off though, which brings tweeters back into the picture. I put the G3 ribbons back in, and found them to work much better with the Be drivers (due to their greatly reduced HF breakup and resonance).
My next step is to re-cut the woofer openings in the upper boxes to accept the larger flanges of the GPA drivers. Once I am able to settle on a winner between the TD15A, GPA 416 and GPA-515, the new crossovers can be developed.
Finally, the frequency extremes. The plate amp powering the bass boxes will be replaced by a QSC DSP-30 and a Parasound HCA-1500a, and the Aurum Cantus G-3's will step aside to make room for the Raal Lazy Ribbons.
Tried to post a picture of the speakers as they have appeared for the past 3 years, but the forum won't allow this until I cross the 5 posts/3 days threshold.
Gary Dahl
So many tubes, so little time...
- IslandPink
- Amstrad Tower of Power
- Posts: 10041
- Joined: Tue May 29, 2007 7:01 pm
- Location: Denbigh, N.Wales
#529
Yowsa !
More info than I can shake a stick at ... in the short term
Thanks guys .
Gary - I'll get my machinist friend to do the drilling on the Azura flanges, thanks for the advice . I'm sold on the 745 Neo/Be's - just a matter of selling some gear and saving some money !
On the Baffle/ OBB issues, from a quick run-through, all I can usefully offer is that (a) The 6dB baffle peak is (typically) substantially mitigated by having a rectangular baffle plus the floor reinforcement . Maybe if you have a limited size for the upper-bass baffle, this may not hold .
(b) the comb-filtering of the OBB sides can definitely be 'blended' and improved by slanted sides , as per the Quasar .
More info than I can shake a stick at ... in the short term
Thanks guys .
Gary - I'll get my machinist friend to do the drilling on the Azura flanges, thanks for the advice . I'm sold on the 745 Neo/Be's - just a matter of selling some gear and saving some money !
On the Baffle/ OBB issues, from a quick run-through, all I can usefully offer is that (a) The 6dB baffle peak is (typically) substantially mitigated by having a rectangular baffle plus the floor reinforcement . Maybe if you have a limited size for the upper-bass baffle, this may not hold .
(b) the comb-filtering of the OBB sides can definitely be 'blended' and improved by slanted sides , as per the Quasar .
"Once you find out ... the Circumstances ; then you can go out"
#530
Clearly I need to get my post count up, so I will take this opportunity to introduce myself.
I serve as assistant conductor, personnel manager and principal timpanist for the Bremerton Symphony Orchestra. I am the choir director at Summit Ave Presbyterian Church. My "day job" is teaching general music at Green Mountain Elementary.
I was raised in rural Montana, earned my BA in Music Ed at Pacific Lutheran University in Tacoma, WA, and my M Mus in Conducting at Western Washington University in Bellingham, WA (1990). I had further studies in conducting with Gunther Schuller at Sandpoint, and in timpani with Michael Crusoe of the Seattle Symphony. In elementary music, I have extensive training in Kodaly, Dalcrose and Orff Schulwerk. I was conductor of the Okanogan Valley Orchestra for 12 years, and have also conducted for Kitsap Opera, Puget Sound Opera, Bremerton Community Theatre and Kitsap Youth Symphony.
In my spare time (ha ha!) I enjoy building and listening to musical electronics.
Over the years I have had what could best be called a love-hate relationship with high end audio, because most of it (at least in the US) doesn't really sound to me like real music. The so-called "close-up" sound isn't really like what instruments sound like up close. I am accustomed to being on stage, either in front of or alongside other musicians, which has led to my interest in creating a system that sounds realistic to me.
Gary Dahl
I serve as assistant conductor, personnel manager and principal timpanist for the Bremerton Symphony Orchestra. I am the choir director at Summit Ave Presbyterian Church. My "day job" is teaching general music at Green Mountain Elementary.
I was raised in rural Montana, earned my BA in Music Ed at Pacific Lutheran University in Tacoma, WA, and my M Mus in Conducting at Western Washington University in Bellingham, WA (1990). I had further studies in conducting with Gunther Schuller at Sandpoint, and in timpani with Michael Crusoe of the Seattle Symphony. In elementary music, I have extensive training in Kodaly, Dalcrose and Orff Schulwerk. I was conductor of the Okanogan Valley Orchestra for 12 years, and have also conducted for Kitsap Opera, Puget Sound Opera, Bremerton Community Theatre and Kitsap Youth Symphony.
In my spare time (ha ha!) I enjoy building and listening to musical electronics.
Over the years I have had what could best be called a love-hate relationship with high end audio, because most of it (at least in the US) doesn't really sound to me like real music. The so-called "close-up" sound isn't really like what instruments sound like up close. I am accustomed to being on stage, either in front of or alongside other musicians, which has led to my interest in creating a system that sounds realistic to me.
Gary Dahl
So many tubes, so little time...
-
- User
- Posts: 141
- Joined: Tue Dec 17, 2013 3:48 am
- Location: Northern Colorado, USA
#531
I should mention that Gary Dahl and I have similar but not identical tastes, and that Gary D has owned Ariels in the past, continues to own Amity amplifiers, and has heard my system when I lived in Washington State and in Colorado. He has a lot more experience with Altec than I do, though, and has owned Altec A7 828 models, the eXemplar, and Altec Duplex systems.
My bias is towards electrostats, actually, and the Ariels are voiced similar to stacked Quad ESL57's ... in fact, they were designed for a friend in Portland that owned stacked Quads.
The new system (I really must give it a name one of these days), in turn, is voiced similarly to the Ariels, with a bit of Altec flavor thrown in (mostly the dynamics and tone colors). I am not a fan of most horn systems, which is probably why I get a lot of stick from horn enthusiasts in other forums ... my goals are different than theirs, and the lack of commitment to the cause of horns is not appreciated.
Side note: It's true the full 6 dB of baffle peak is only seen with a circular baffle without side wings. Any kind of asymmetry in the baffle spreads out the arrival time of the back-wave, which then reduces the magnitude of the baffle peak. Angling the side wings "steers" diffraction around the edges and alters the time spread of the back wave.
Another technique might also be interesting; the side wings can be made of pegboard, which "leaks" the back wave gradually through the perforations. If the pegboard is covered with heavy felt on the inside, or even a layer of felt sandwiched between two pegboards, the time release of the back-wave would be even more progressive and gradual.
My bias is towards electrostats, actually, and the Ariels are voiced similar to stacked Quad ESL57's ... in fact, they were designed for a friend in Portland that owned stacked Quads.
The new system (I really must give it a name one of these days), in turn, is voiced similarly to the Ariels, with a bit of Altec flavor thrown in (mostly the dynamics and tone colors). I am not a fan of most horn systems, which is probably why I get a lot of stick from horn enthusiasts in other forums ... my goals are different than theirs, and the lack of commitment to the cause of horns is not appreciated.
Side note: It's true the full 6 dB of baffle peak is only seen with a circular baffle without side wings. Any kind of asymmetry in the baffle spreads out the arrival time of the back-wave, which then reduces the magnitude of the baffle peak. Angling the side wings "steers" diffraction around the edges and alters the time spread of the back wave.
Another technique might also be interesting; the side wings can be made of pegboard, which "leaks" the back wave gradually through the perforations. If the pegboard is covered with heavy felt on the inside, or even a layer of felt sandwiched between two pegboards, the time release of the back-wave would be even more progressive and gradual.
#532
I can't contribute to the theory of sound waves but here is a pic of my OB's with (piano hinged) adjustable wings..we live in small rooms where it is impossible to have OB's in free space so alot of theory goes out of our small windows..
The room corner in the pic. is a problem and I have to place felt behind the drivers with the OB further out in the room...the other corner is o.k.
Nice to see you here..
The room corner in the pic. is a problem and I have to place felt behind the drivers with the OB further out in the room...the other corner is o.k.
Nice to see you here..
#533
Hi LynnLynn Olson wrote:
My bias is towards electrostats, actually, and the Ariels are voiced similar to stacked Quad ESL57's ... in fact, they were designed for a friend in Portland that owned stacked Quads.
It's the first time I've noticed you mention this and it struck a real chord with me. When I first built Ariels back in 2003/4 they were the nearest I'd ever heard boxes come to my electrostats. I found that hard to explain to others though. Very intersting observation nonetheless.
I still use my original Ariels in the project studio, nothing I've played with since does the job so well. Incidentally, since our last correspondance I havn't had a single enquiry to build a pair, although prior to me asking I was getting regular requests. Go figure....
I'm a bit frustrated because I would love to try OBs but like Will I have a small room and they would seem to be impractical...
There's nowhere you can be that isn't where you're meant to be
- IslandPink
- Amstrad Tower of Power
- Posts: 10041
- Joined: Tue May 29, 2007 7:01 pm
- Location: Denbigh, N.Wales
#534
I'll be (mostly) away from the computer for a few days now.
Have a nice Christmas everyone !
One question for Lynn - what baffle area/dimensions can you make available for the 515 , given you have the low-bass solution somewhere underneath ?
Have a nice Christmas everyone !
One question for Lynn - what baffle area/dimensions can you make available for the 515 , given you have the low-bass solution somewhere underneath ?
"Once you find out ... the Circumstances ; then you can go out"
- slowmotion
- Old Hand
- Posts: 282
- Joined: Fri Aug 21, 2009 4:07 pm
- Location: Norway
#536
A nice Christmas for everyone!
I'm wondering what happens if one makes the baffle wider
on the outside so it touches the side wall?
What does the theory say?
And it's nice to see Lynn and Gary here.
I've enjoyed the thread in that other place a lot,
but it did seem to go astray.
I'm wondering what happens if one makes the baffle wider
on the outside so it touches the side wall?
What does the theory say?
And it's nice to see Lynn and Gary here.
I've enjoyed the thread in that other place a lot,
but it did seem to go astray.
- Jan -
- Paul Barker
- Social Sevices have been notified
- Posts: 8863
- Joined: Mon May 21, 2007 9:42 pm
#538
You would have seen me and RayP from here at VSAC as half the English (Edit British, I think Scot who came may have been Scottish, he certainly liked pure Malt.) contingent.
We had a great time.
We had a great time.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
#539
Indeed, we did have a great time, though it was a decade ago now - how time flies nowadays!Paul Barker wrote:You would have seen me and RayP from here at VSAC as half the English (Edit British, I think Scot who came may have been Scottish, he certainly liked pure Malt.) contingent.
We had a great time.
Ray