Hi,
I finally found a compromisse and went for a AA Beethoven inspired dynamic open baffle speaker. Heard Orions and liked them alot. I also noticed couple problems with them and I'm familiar with some aspects that could be better in original Beethoven concept as well.
So the overall setup is WWMTM main panel with Beethoven styled W-baffle dipole colums following later. Mains have drivers selected and invested. 25W/8565-01 Scan Speak is serving for lower mid (and as bass before actual woofers appear), Seas W18E is selected for upper midrange duty and Dynaudio T330D Esotar should finish the job. Esotar is fitted into a waveguide to match dipole mids front radiation pattern. Considered xover frequences 250-300Hz 1st to 2nd order and 2-2,5Khz 2nd to 4th order. Lowest cross is going to be around 80-100Hz prefering 2nd order.
Already have two poweramps and Behringer DCX-2496 for prototyping. Finally I attend to upgrade Behringer into a solid analog active filter and perhaps use passive filtering if drivers efficiencys work along. Naturally the lowest cross is going to be active anyway.
I know. It's pretty much totally crazy project. Room is relatively small but I can get these placed pretty well considering the room. 1,5m space from speakers to front wall, 1,5m from listener to rear wall, 0,60m from panels midship to sidewalls. Nice 2,5m listening triangle and uses Linkwitz recommended positioning for dipole sources. Sidewalls, ceiling and floor are concrete so dipoles nulls should be able to bite most of the problems. Front wall is mostly window and a door, rear wall is 1/3rd doorway and 2/3rds thin wall that should pass some bass frequences. Separated dipolesubs can be placed along the sidewalls outside from main panels. Turning panels straight towards listener 1st reflections should arrive around 80 degree angle from the speakers view. With dipole radiation pattern they should be pretty tamed. Acoustic treatment is also possible.
As for the mechanical structure I noticed Orions shaking problem which in my friends case wakes up some floorboard resonances and kicks butt. Literally. This is one thing I'd like to skip. Other is Beethoven and Phoenix based stuffed spine structures.
From mms view this setup has it's advantages. 25W pair, W18 pair and Esotar have only single 10" XLS mms on them. And I'll make a lot more sturdy panels than Orions are. I'm using 16mm plasmacutted steelplate as a frame for the panels. Bottomplates are 12mm steel, welded together. Frame is covered with 16mm MDF sheets to create constraint layer damping. Already got the frames and bottomplates. Working on the wood pieces. Tweeter has 50mm offset from the vertical centerline. I'll clamp the drivers between steel frame and MDF sheet, propably glue them in. This leaves drivers mounted a bit deeped in the panel so I can use speakercloth just along the panel and don't have to worry about cones banging the cloth.
Structure is heavy. Here is a picture from one of the frames. It weights 25kg. Bottomplates are about 13kg, 10kg of drivers, perhaps 10kg of wood. Recently I added a lean spine to them. This increases depth dimension at frame-bottomplate connection and propably makes whole thing stiffer. It also allows drivers magnet support and is pretty lean character so it shouldn't mess up that much rear radiation.
But does this look/sound dull, too common or how these parts look together visually? Pretty straightforward panel and then some curved at the bottomplate.
All other comments are also very welcome.
Jussi
I finally found a compromisse and went for a AA Beethoven inspired dynamic open baffle speaker. Heard Orions and liked them alot. I also noticed couple problems with them and I'm familiar with some aspects that could be better in original Beethoven concept as well.
So the overall setup is WWMTM main panel with Beethoven styled W-baffle dipole colums following later. Mains have drivers selected and invested. 25W/8565-01 Scan Speak is serving for lower mid (and as bass before actual woofers appear), Seas W18E is selected for upper midrange duty and Dynaudio T330D Esotar should finish the job. Esotar is fitted into a waveguide to match dipole mids front radiation pattern. Considered xover frequences 250-300Hz 1st to 2nd order and 2-2,5Khz 2nd to 4th order. Lowest cross is going to be around 80-100Hz prefering 2nd order.
Already have two poweramps and Behringer DCX-2496 for prototyping. Finally I attend to upgrade Behringer into a solid analog active filter and perhaps use passive filtering if drivers efficiencys work along. Naturally the lowest cross is going to be active anyway.
I know. It's pretty much totally crazy project. Room is relatively small but I can get these placed pretty well considering the room. 1,5m space from speakers to front wall, 1,5m from listener to rear wall, 0,60m from panels midship to sidewalls. Nice 2,5m listening triangle and uses Linkwitz recommended positioning for dipole sources. Sidewalls, ceiling and floor are concrete so dipoles nulls should be able to bite most of the problems. Front wall is mostly window and a door, rear wall is 1/3rd doorway and 2/3rds thin wall that should pass some bass frequences. Separated dipolesubs can be placed along the sidewalls outside from main panels. Turning panels straight towards listener 1st reflections should arrive around 80 degree angle from the speakers view. With dipole radiation pattern they should be pretty tamed. Acoustic treatment is also possible.
As for the mechanical structure I noticed Orions shaking problem which in my friends case wakes up some floorboard resonances and kicks butt. Literally. This is one thing I'd like to skip. Other is Beethoven and Phoenix based stuffed spine structures.
From mms view this setup has it's advantages. 25W pair, W18 pair and Esotar have only single 10" XLS mms on them. And I'll make a lot more sturdy panels than Orions are. I'm using 16mm plasmacutted steelplate as a frame for the panels. Bottomplates are 12mm steel, welded together. Frame is covered with 16mm MDF sheets to create constraint layer damping. Already got the frames and bottomplates. Working on the wood pieces. Tweeter has 50mm offset from the vertical centerline. I'll clamp the drivers between steel frame and MDF sheet, propably glue them in. This leaves drivers mounted a bit deeped in the panel so I can use speakercloth just along the panel and don't have to worry about cones banging the cloth.
Structure is heavy. Here is a picture from one of the frames. It weights 25kg. Bottomplates are about 13kg, 10kg of drivers, perhaps 10kg of wood. Recently I added a lean spine to them. This increases depth dimension at frame-bottomplate connection and propably makes whole thing stiffer. It also allows drivers magnet support and is pretty lean character so it shouldn't mess up that much rear radiation.
But does this look/sound dull, too common or how these parts look together visually? Pretty straightforward panel and then some curved at the bottomplate.
All other comments are also very welcome.
Jussi
Attachments
Those are going to be some crazily overbuilt baffles. Most impressive. You cutting the steel yourself?
Kensai
Kensai
Kensai said:
Well. Are they heavy enough. I don't know. Pretty much guessing since I don't have proper equipment and environment to experiment things through dozens of prototypes. Pretty much normal compromisse in DIY R&D... 🙁
At least they shound't dance around that much and overall structure is kept relatively slim and elegant and still obtain very high stiffness. This is one of the problems I see in many open baffle systems. It's impossible to get ultimate damping + stiffness and maximum open structure in once without using something exotic. Oh well, is steel exotic. Very rare, even in expensive commercial systems but it isn't that tough task if you really want such solution. There are plenty of metalshops to do the thing...
But plain steelplates need some serious damping. The frame alone ring like a bell. I hope the elastic fastening and MDF sheets do the job so I don't have to name these "Hells Bells". 😀
I guess the whole setup is blown off any reasonable proportion. But I like the idea of headroom if system needs considerable amount of planning, work, designing and of cource financial means.
Personally I can defend myself with this Linkwitz wisdom: "Extreme as these designs may seem, the goal here is to keep non-linear distortion low at high volume levels or, conversely, have very low distortion at normal levels.
I think it is important to be able to play back at live like levels. If the system has low distortion it will not sound loud, just natural. The level needs to be realistically high, so that the ear produces the same distortion and timbre as it would at a live event. Then you can experience maximum involvement and enjoyment. " 😉
This fits very well into my own perspectives and preferences. Not that I play at very high levels often but systems with almost ridiculous headroom play music with all its dynamics less stressed and more effortless. I played some of my normal test CD:s with one perhaps overkilled system and noticed how fearsome some of the dynamic parts really are. People commonly complain about compressed recordings and sure, it is a big problem. But in the same time many "normal" home systems have pretty significant compression problems with those "not that compressed" recordings. And is there another way to solve this than total overkill? 🙂
Nope. I have a friend in a metalworkshop and he made the frames and bottomplates. I have place where I can weld them together and do the finishing if whole thing works acoustically...
New ideas and suggestions are also welcome if something comes in mind... Just frames and plates are ready, woodplates aren't.
Jussi
Didn't mean anything derogative about your build. Just expressing awe and admiration for the magnitude.
Anyway, if they ring badly as is, then you may not want to weld them to the base plates. Use something that will isolate them accoustically, like wood or rubber, perhaps.
As for sheathing the steel, first off I don't like the idea of so permanently mounting the drivers between the MDF and steel. I would personally just sheath them and then mount normally. Also, I would probably use a nice solid plywood of some sort and then MDF on the back, simply for the aesthetics, though there have been some statements made as to the acoustics of ply over mdf in OB in a current thread. You'd probably also want to edge the frames in MDF (make the front ply plate bigger than the steel plate so you can put a tight edge of MDF behind the lip) so you don't get any ring off the baffle edge. On the back, you might even go with multiple layers of MDF, though not uniform. Thicker toward the edges would probably make a much better performing shape while progressively dampening the underlying steel toward the edge. Make sure to clear out the driver openings on the back side to maximize air flow.
Just the thoughts I had in mind. With your design, the crossover is going to be a huge factor in how good they sound, which is not my department (I usually hang out in the fullrange driver page over yonder). This will be fun to watch 🙂
Kensai
Anyway, if they ring badly as is, then you may not want to weld them to the base plates. Use something that will isolate them accoustically, like wood or rubber, perhaps.
As for sheathing the steel, first off I don't like the idea of so permanently mounting the drivers between the MDF and steel. I would personally just sheath them and then mount normally. Also, I would probably use a nice solid plywood of some sort and then MDF on the back, simply for the aesthetics, though there have been some statements made as to the acoustics of ply over mdf in OB in a current thread. You'd probably also want to edge the frames in MDF (make the front ply plate bigger than the steel plate so you can put a tight edge of MDF behind the lip) so you don't get any ring off the baffle edge. On the back, you might even go with multiple layers of MDF, though not uniform. Thicker toward the edges would probably make a much better performing shape while progressively dampening the underlying steel toward the edge. Make sure to clear out the driver openings on the back side to maximize air flow.
Just the thoughts I had in mind. With your design, the crossover is going to be a huge factor in how good they sound, which is not my department (I usually hang out in the fullrange driver page over yonder). This will be fun to watch 🙂
Kensai
to dampen the steel the best..
lead:microsorb:steel:microsorb:lead
2 things should be clear here:
1. the outer material should be heavier than the steel, NOT lighter (like mdf).. and the coupling material should be EXTREMELY lossy.
2. dampen both sides of the material.
I'd prob. only use this at the outer edges of the steel baffle/support. The best course of action is a quick sink for the energy from frame to baffle that is dampened as far away from the frame/baffle coupling as possible.
lead:microsorb:steel:microsorb:lead
2 things should be clear here:
1. the outer material should be heavier than the steel, NOT lighter (like mdf).. and the coupling material should be EXTREMELY lossy.
2. dampen both sides of the material.
I'd prob. only use this at the outer edges of the steel baffle/support. The best course of action is a quick sink for the energy from frame to baffle that is dampened as far away from the frame/baffle coupling as possible.
I like to completely physically isolate the woofer array from the mid-tweeter with both a separate stand and a separate baffle. Having two separate structures also allows separate toe-in/out options. My room sounds best with a very tiny toe-in on the mid-tweet and a modest toe-out of the woofers toward the wall.
Kensai said:
I didn't take that as insult. Quite opposite. 🙂
Yep but that makes whole things overall stiffness much lower. One of the main reasons why I choose steelframe is that joint between baffle and bottoplate. There is no way to make that look so elegant and at the same time so stiff and sturdy by using MDF or such.
I didn't got your first point here but the woodparts of the baffle form a 50x25mm elliptic curve along both sides so they should be as diffraction free as possible. I know there is still going to be problems above 1Khz before tweeter come into play since there is at least one dipole null at 1-2Khz range.
Midrange drivers have 45 degree cuts on their opening and without cutting the overall thickness it's pretty much as open as possible. Allthought that's not enough. We measured single 7" Peerless HDS in free air, no baffle at all, and it just doesn't work as a dipole much above 1Khz. Spider and magnet structure just mess it up.
Thanks. Crossover and finally electronics. I have Behringer DCX for the prototyping so I can easily try different setups but finally it sure needs a well done passive network (with active lowest cross to actual woofers) or full active HQ analog filter.
LineSource said:
I agree. Positioning can be different, separated woofers can have sidewall boost without compromising main baffle response and most of the shaking also happens in physically separated box and doesn't disturb most of the frequency range.
Jussi
This is one serious project. The cost involved in any part of the construction (or theoretical construction like the frame damping ScottG is talking about) is more than the whole speaker of any of my currently built projects. I certainly can't wait to see how these monsters turn out.
Good luck.
Kensai
Good luck.
Kensai
Member
Joined 2003
Hi Jussi,
It is good to see your project moving forward...I'll be following it with great interest!
What are your plans for the region where the OB mids transition to the monopole tweeter? Most seem to disregard the different radiation patterns and I am curious about your current thoughts.
Paul
It is good to see your project moving forward...I'll be following it with great interest!
What are your plans for the region where the OB mids transition to the monopole tweeter? Most seem to disregard the different radiation patterns and I am curious about your current thoughts.
Paul
Paul W said:
On a baffle capable of supporting the W18 - the dipole radiation won't effect the dispersion patteren near the tweeters crossover of 2-2.5 kHz.
The problem here is the typical matching beaming mid to omni'esq tweeter.
Looking at the W18:
http://www.seas.no/PDF data excel 05-06/W18E001E0018.pdf
The - 6db point at 60 degrees is down about 6 db near 2.2 kHz. Now the sp loss is more linear from about 500 to 1.9 kHz. Considering a linear 4th order LR electrical crossover - I'd be looking at about 1.5 kHz for a crossover point with a waveguide that has a suitable matching dispersion pattern (..with the esotar).
Member
Joined 2003
Scott,
Guess I didn't state it clearly enough...I'm asking about the radiation to the rear from the back of the cone (frontside no problem).
Although many don't, a few designs address the fairly abrupt transition from dipole radiation to monopole and wondered what Jussi was planning.
Paul
Guess I didn't state it clearly enough...I'm asking about the radiation to the rear from the back of the cone (frontside no problem).
Although many don't, a few designs address the fairly abrupt transition from dipole radiation to monopole and wondered what Jussi was planning.
Paul
Paul W said:
It's a double blade sword. Most of the sound that reaches listeners ears are from the panels front radiation with first reflections from boundaries nearby. Then there are more reflections with longer delay that ear can define non related to direct sound. And in that aspect waveguide and mids radiation pattern matched tweeter is in order. Without the guide steep off axis angles don't have that much directivity and since tweeter is handling a pretty broad range I'd say it's a problem. I can't say for indefinitely did I heard it with Orion. Some slight problems but it could have been something else as well. That would require more intensive study on Orion than I could do on a normal visit.
In Orions case directivity is pretty uneven at different frequences. Measured power response is pretty good allthought it has a bit of a roll off towards couple top octaves. Does this make speaker more usable on different recordings and actually better playing varying music? Perhaps. When it comes to first reflection, the floor, Orions directivity is unefficient. Woofers are crossed away at 120Hz with steep loop so they don't assist that much, single midrange and nonwaveguided tweeter. Actually using two mids doesn't help much here. Lower midrange isn't still directive with bass solution like that, there are some vertical MTM beaming at midrange and then whole thing spreads all open again with the tweeter. Maybe even bigger problems than this used WWMT configuration. This might be one reason why SL dropped the other mid off from his Phoenix system while designing Orion?
As a floorbounce I consider about -45 degree vertical response. Depending on different listening distances it varies around -35-55 degrees. Driver mounted in open baffle has some directivity on a horizontal plane. Some of it also follows to vertical plane but dipoles natural -3dB directivity at 45 degrees angle isn't enough when floor is so close and directly between listener and the speaker. Sidewalls and ceiling are easier tasks. More distance, steeper reflection angles and so on. In my case pointing the panels towards the listener 1st reflection from sidewalls arrive to listening position around 80 degrees off axis from speakers viewpoint. With dipole radiation pattern this means over 10dB directivity which is pretty good. Even while distance to sidewalls is limited. More toe in and result is even better in this aspect. We measured Esotar in my present waveguide using a small MDF piece for installation and its about -13dB at 90 degrees off axis. Installing waveguide and tweeter into a much larger baffle it should be even better. Pattern is also in very close to dipolar front radiation in more common 45 degrees and 60 degrees off axis angles. -3dB and -6dB.
Well. My point here isn't to mud Orions. Or any other similar design. I just use it as example since most people that have played with open baffles know this design. Personally I considered Orion a very good speaker and most usable with various music material without being picky. This is important for me and Orion was a big source of inpiration that lead me to this project. Maybe I'm too stubborn not to build Orions straight away but I consider there are some problems that could be fixed.
Problems are a relative term. I don't think SL even considered Orion to be that directive speaker. People have wondered this aspect when he come up with Pluto which is pretty much totally different approach. In FAQ SL states that the most important thing in these designs is even and smooth power response. And thats achieved in Orions case. Using broadly radiating single front firing tweeter it doesn't leave a notch in power response that a front pattern matched waveguided tweeter causes. In theory that's 3dB in power response when rear radiation drops off and tweeter continues with just the front radiation.
Most of directivitys advantage is bass and lowmidrange. And this is done in Orion. But for example against floor reflection configurations directivity isn't enough. Not even if the midrange would be without baffle and have its schoolbook -3dB at 45 degree directivity. It just isn't enough if floor reflection killing is in mind. And then the treble. Local magazine measured Orions and did the normal floor reflection measurement. Response with just direct radiation from speaker and 1st reflection from floor with numeric dB value how much different floor reflection is from direct sound. At their normal range 800-10000Hz (in this measurement) Orions value is 5dB which isn't that different from any 2-way or such.
So the solution. More drivers in vertical plane and waveguide for the tweeter. It's still uneven but more directive. Best configuration would be Beethoven Grand like WWMTMWW but I can't have such colossal panels in this room so I have to settle for Beethoven style WWMTM. Using gentle slope between lowmids and uppermids it creates some vertical directivity. MTM enhances and extends it up in frequency while waveguided tweeter finishes. WWMTMWW would make this symmetrical but I can't have that. In theory I'd need a nonsymmetric waveguide as well. It would need mensioned -3dB at 45 deg, -6dB at 60 deg horizontal directivity but also perhaps double or triple that in +-45 deg vertical window. I think this kind of configuration suffers even more from unwaveguided tweeter than Orion style that isn't that directive at other frequencies either.
I consider even directivity to be the key here. It doesn't work that well if some range beams fiercely and some other is spread all over the place. Some large midwoofer 2-ways suffer from this problem. Even if the radiation patterns are matched. Cabinet works as monopole bottom of it's range and top of the treble is very directive. This causes pretty extensive roll off towards treble in power response and this is normally also very audible in actual room. All frequences aren't radiated the same amount. Problem can be fixed a bit by using bright freefield response but I consider that pretty woodleg solution. It doesn't fix the problem, just covers it and not even that if acoustic environment changes. Problem is even bigger if tweeters radiation pattern isn't matched and it causes camelback effect in power response. Monopole bass, midwoofer beams, tweeter lowrange spreads it back open and tweeter beams.
Here is a visual example. Orions case. 2x10" woofers, 8" midrange, 1" nonwaveguided tweeter. 120Hz and 1400Hz 4th order crosses. Used -500mm and -750mm for woofers positions, -200mm for W22 (which is stated in Orions specs). As you see +-45 degrees vertical response is very close to on axis. Naturally this is just the driver configuration and crossovers created directivity. Dipole radiation patterns directivity isn't simulated here. But like you know it's -3dB at 45 degrees at best.
ScottG said:
Actually midrange should have a bit wider dispersion in open baffle at the top if its range. At least Linkwitz Phoenix showed this sort of behavior. 8" mids didn't beam 60 deg off axis at 1500Hz but around 2000Hz.
Jussi
Attachments
And here is the other example. 2x10", 2x7" and 1". Crossed 250Hz 2nd and 2200Hz 4th order. Woofers -400mm and -670mm from tweeter, mids +-150mm from tweeter. Matches my project dimensions. With 2nd order upper cross behavior is even smoother but I don't expect to reach 2nd order loop with waveguided Esotar. Could be a nice bonus thought...
There are some configuration and crossover frequency / order based vertical directivity here. Between 300Hz and 2200Hz it's over 6dBs down from on axis and both 45 degree responses start dropping from on axis just above 100Hz.
This simulation doesn't count waveguide in so the tweeters response is pretty identical to Orions. In my case imagine there -3dB for +-45 deg response for tweeters range. Like I said, it isn't enough, should be -6dB or even more but I don't have tools to make nonsymmetrical waveguides. Or not at the moment at least. Perhaps it could be done from 22mm steel with CNC lathe. 😉
With symmetrical (and bigger) WWMTMWW configuration vertical directivity would start a bit lower and behave symmetrically to both directions. Basically this means room height linesource type system with crossovers to cut it shorter while frequency goes up.
Most effective floor reflection appears around 250Hz with normal height speakers. In this case systems vertical directivity can punish some of it and create more punch towards the listener.
After this very messy text you propably consider me as a total measuring freak but thats not correct. These are just some of my preliminary thoughts, simulations and such that I made before getting into this task. System is still far from perfect. For example at least one dipole null is going to appear around 1-2Khz and I can't help it. So it's still going to be far from measure perfect even when it isn't here to measure yet. I think that MTM null at that range cause enough notch for power response that dipole null created momentary pattern spreading isn't audible.
Personally I'm after a long term musical enjoyment. I like clean, effortless and non tiring sound. System should have enough dynamics to give justice to symphonys I started collecting some time ago while smaller performers should have their slice as well...
Jussi
There are some configuration and crossover frequency / order based vertical directivity here. Between 300Hz and 2200Hz it's over 6dBs down from on axis and both 45 degree responses start dropping from on axis just above 100Hz.
This simulation doesn't count waveguide in so the tweeters response is pretty identical to Orions. In my case imagine there -3dB for +-45 deg response for tweeters range. Like I said, it isn't enough, should be -6dB or even more but I don't have tools to make nonsymmetrical waveguides. Or not at the moment at least. Perhaps it could be done from 22mm steel with CNC lathe. 😉
With symmetrical (and bigger) WWMTMWW configuration vertical directivity would start a bit lower and behave symmetrically to both directions. Basically this means room height linesource type system with crossovers to cut it shorter while frequency goes up.
Most effective floor reflection appears around 250Hz with normal height speakers. In this case systems vertical directivity can punish some of it and create more punch towards the listener.
After this very messy text you propably consider me as a total measuring freak but thats not correct. These are just some of my preliminary thoughts, simulations and such that I made before getting into this task. System is still far from perfect. For example at least one dipole null is going to appear around 1-2Khz and I can't help it. So it's still going to be far from measure perfect even when it isn't here to measure yet. I think that MTM null at that range cause enough notch for power response that dipole null created momentary pattern spreading isn't audible.
Personally I'm after a long term musical enjoyment. I like clean, effortless and non tiring sound. System should have enough dynamics to give justice to symphonys I started collecting some time ago while smaller performers should have their slice as well...
Jussi
Attachments
Well, its obvious you're in deep intellectually ont his project. That's a staggering amount of analysis.
All I can say on the subject is experiential, first and second hand.
Second hand, there was a gent around here that has built a pair of OB line arrays using large B&G planar elements as the basis. One of these giant line planars with a row of like 5"-7" woofers next to them. He decided he wanted to try a supertweeter with these (they spec up to 18kHz, but he said he was hearing a lack of air). He used some sort of dome tweeter. He had them set up in a large room to where he could walk around, between them during normal use of the house and was enamored with the ability to walk "through" his recordings like this. Putting in the dome tweeter killed this sensation, so he put another one in each line, rear firing, which fixed the issue.
First hand, I would recommend going with a B&G Neo3PDR tweeter (remove the back cup). I love those things in dipole.
Kensai
All I can say on the subject is experiential, first and second hand.
Second hand, there was a gent around here that has built a pair of OB line arrays using large B&G planar elements as the basis. One of these giant line planars with a row of like 5"-7" woofers next to them. He decided he wanted to try a supertweeter with these (they spec up to 18kHz, but he said he was hearing a lack of air). He used some sort of dome tweeter. He had them set up in a large room to where he could walk around, between them during normal use of the house and was enamored with the ability to walk "through" his recordings like this. Putting in the dome tweeter killed this sensation, so he put another one in each line, rear firing, which fixed the issue.
First hand, I would recommend going with a B&G Neo3PDR tweeter (remove the back cup). I love those things in dipole.
Kensai
We tested B&G Neo3 a bit over six months ago. Decent for its price but still cant compete with standard series dome tweeters. Response has it's problems and distorsion isnt clean either.
In this aspect I attend to stay with Esotar. At least if I don't run into serious problems installing it into a waveguide. Esotar can produce very high volumes with very low stress and I don't see a point why to choose lower quality tweeter and cause a weakpoint there while lower and uppermids can deliver plenty of HQ performance.
If I don't remember incorrectly T330D has <0,3% THD at 95dB SPL in its range. Waveguide assists up to 6dB and THD should remain constant. But I don't have measured information of this waveguided Esotar business. Not yet.
I think I'll mension few of my perspectives in this design. First it's a long term project and meant to serve as my personal music enjoyment for a long time. Perhaps permanently. This is why I don't want to make that much compromisses althought some are unavoidable. This thinking lead me into a rather massive 4-way concept even while overall physical appearance is kept pretty elegant.
I quoted Linkwitz in post #6. Check it out. This is one thing I consider important in "no compromisse" systems. I understand it can't be achieved if the goal is to design a 2-way in compact size but it's one serious issue to take into count if "whole packadge" is the target.
I have studied, experimented and tested things for about five years to come up with "ultimate solution" for my own purposes. Naturally I've build few smaller systems, prototypes and such during this time so don't think that this guy has just wondered this one thing for five years. 🙂
Unfortunately I noticed that more I learn, more I know what kind of compromisses had to be made. So this is just one vision, one compromisse. Excellent all around response, seamless integration, excellent details and imaging with lifelike dynamics are my primary objectives. I don't know how well those aspects are met but that can be reported later...
Orion experience was good. Not unique. I've heard similar quality performances and in some aspects even better. But the concepts all around music serving operation got me thinking. I know that I can have much more resolution even up to brutal levels but that just doesn't serve music. It squeezes guts out of recordings and puts them up for display. Makes poor recordings sound extremely poor. Naturally every HQ system does some of it. Shows limited dynamic range when compared to quality recordings and so on. But for example Orion did it putting music first, gut wrenching resolution second. And this is one thing I consider important when it's meant to enjoy the music, not to suffer from it. 🙂
I've also done some analysis on different music genres. Primary on their frequency range and dynamics. What is adequate frequencyrange and what kind of SPL levels system has to reproduce to be authentic with different recordings. Naturally every album can also be played at modest levels, this is just one aspect I wanted to pay attension in this design.
Then there are few milestones in overall design that I considered important when it comes to natural sound reproduction. Room problems is very important one. Couple earlier posts had some thoughts about this aspect.
Adequate SPL and dynamics capacity is another. Low distorsion and high effieciency I'd consider key features here. One of them isn't enough, we need both. High efficiency creates easily high SPL but without low distorsion it's just noise. Unbearable. And low distorsion with low efficiency isn't much better. Dynamic drivers don't take that much RMS power without significant thermical problems. For example normal home midwoofers can handle maybe 30-40W and then adding more RMS power doesn't do any good. Of cource stated power handling is higher but that's just the point where voicecoil burns. Actual problems are met earlier. And if driver needs relatively high power to sustain needed RMS level it suffers from thermical problems already and isn't cool enough to handle transients. This is what I personally call the "lightbulb-effect".
This is why I don't even want to use big amps. It's easier to build system 3dB more efficient than use two times more watts to drive it. Don't have the measurements yet but I expect to have pretty decent figures from my own construction. Esotar is 92-93dB sensitive 8 ohm unit without the waveguide. Guide adds 0-6dB depending on frequency. Most to the bottom of the range rolling off towards mid treble. This is ok since most of music materials intensity is below 3500Hz. Above from there that much power isn't needed. Uppermidrange units with 1st to 2nd order 250-300Hz cross should be pretty much dipole eq free and so they can play as loud as they can. W18W is stated about 87-88dB sensitive and pair of them creates 93-94dB sensitive nominally 4 ohm load. 25W pair can't reach those figures but they gain some boost from the floor on their range and lowest cross to actual woofers (80-100Hz) assists more. Overally not that much power isn't needed, RMS power stays low, voicecoils cool and transient response fast.
I've heard that some big high capacity speakers have suffered tweeter failures under high SPL listening. I think this is caused by the lack of efficiency on the tweeter. They can't handle that much power and are "just" 90-92dB sensitive. So demanding 110dB+ levels in large room isn't exactly riskfree. Other frequencyareas can have multiple drivers with larger high power voicecoils so no problems for them. Mechanically tweeters can handle quite much. For example Dynaudio tested their Esotar up to 1kw 10ms burst and it didn't cause compression. Esotar doesn't have much linear excursion but there is plenty of mechanical excursion potential so they don't bottom.
Using Linkwitz spl_max spreadsheet I predicted that Esotars runs out of linear range with 2Khz cross in that waveguide somewhere around 114dB level. Uppermids can handle almost 130dB, lower mids almost 120dB. Naturally these depend on used crossover orders and numbers are pretty theoretical. I won't be having amp big enough to run W18s over their linear range. This arrangement should keep the drivers secure and on their usable excursion range in every situation that listeners ears can handle.
Constructions mechanical structure is also important. For example Wilson has pressed pretty excellent performance from relatively inexpensive units by using very inert and well build cabinets. Acoustic designing could be better but I don't think Wilsons are played in a small room anyway. Linkwitz has/had very good experimental and theoretical information on his site. At least the "recoil" research wasn't there when I searched some time ago. But this can also mean that I didn't searched properly enough.
Mechanical problems can cause a lot of unwanted things. In Orions case whole panel is shaking. Partially because woofers have high mms, panel (as a kit) isn't that heavily build and woofers also move front-back motion. There is no force cancellation. And the woofers are also in same cabinet with the mid-treble units. Compact but compromissed. (Still excellent speaker)
Linkwitz had this recoil research some time ago on his site. He had experimented and measured the effect of panels opposite motion from cones motion. So if cone moves forward it causes the panel move back with the same force. And if the panel is light and not supported enough it will move and shorten cones effective excursion and therefore cause compression to played transient. Intresting point.
Other related problem is panels modulation problems. Panel shouldn't shake at all to keep drivers operation linear. There isn't that much use to have low distorsion driver motors and pistonic cones if whole system shakes and modulate signal. I once calculated one very pessimistic model of this problem. Lets consider panel is 100% stiff and 100% of drivers motion is transfered to move the whole panel. My panel is about 130 x 34cm. That's 4420cm2. Install couple 10" woofers on it. That's 660cm2 leaving 3760cm2 bare panel surface. So 660cm2 should be creating sound and 3760cm2 should be totally quiet. Calculating relation to there it means that moving 660cm2 cones 1mm forward panel has to move just 0,17mm back to cancel whole sound.
Like said. This is very pessimistic model. Panel isn't 100% stiff, it doesnt create 100% modulations and so on. But the problem is there. Just think how small 1" domes movements are. And how messed up it's sound is if the panel rocks even small amounts. Modulation distorsion.
Therefore my panel project propably looks like a helluva overkill but I'm hopeful that I can dodge some of the mensioned problems. 130g mms, very stiff propably over 60kg panel with large area bottomplate and spikes. Should be sturdy platform to operate.
This is also why I wanted to use separated woofers in my own system. Bass needs most displacement, bass needs most conearea. So bass propably has most moving mass and therefore more shaking problems. So using physically separated woofers doesn't allow mechanical shaking to transfer into main panel. Even if the woofers dance around the room frequences above 100Hz aren't messed.
One idea is to make the W-baffle dipole colums from 12mm steel. Quickly calculated mass would be way above 100kg for each unit. Four 12" Peerless SLS woofers on each side with force cancellation installation. If they dance around they also cause significant damage to decoration... 😀
Am I delusional or what.. 😎
Jussi
In this aspect I attend to stay with Esotar. At least if I don't run into serious problems installing it into a waveguide. Esotar can produce very high volumes with very low stress and I don't see a point why to choose lower quality tweeter and cause a weakpoint there while lower and uppermids can deliver plenty of HQ performance.
If I don't remember incorrectly T330D has <0,3% THD at 95dB SPL in its range. Waveguide assists up to 6dB and THD should remain constant. But I don't have measured information of this waveguided Esotar business. Not yet.
I think I'll mension few of my perspectives in this design. First it's a long term project and meant to serve as my personal music enjoyment for a long time. Perhaps permanently. This is why I don't want to make that much compromisses althought some are unavoidable. This thinking lead me into a rather massive 4-way concept even while overall physical appearance is kept pretty elegant.
I quoted Linkwitz in post #6. Check it out. This is one thing I consider important in "no compromisse" systems. I understand it can't be achieved if the goal is to design a 2-way in compact size but it's one serious issue to take into count if "whole packadge" is the target.
I have studied, experimented and tested things for about five years to come up with "ultimate solution" for my own purposes. Naturally I've build few smaller systems, prototypes and such during this time so don't think that this guy has just wondered this one thing for five years. 🙂
Unfortunately I noticed that more I learn, more I know what kind of compromisses had to be made. So this is just one vision, one compromisse. Excellent all around response, seamless integration, excellent details and imaging with lifelike dynamics are my primary objectives. I don't know how well those aspects are met but that can be reported later...
Orion experience was good. Not unique. I've heard similar quality performances and in some aspects even better. But the concepts all around music serving operation got me thinking. I know that I can have much more resolution even up to brutal levels but that just doesn't serve music. It squeezes guts out of recordings and puts them up for display. Makes poor recordings sound extremely poor. Naturally every HQ system does some of it. Shows limited dynamic range when compared to quality recordings and so on. But for example Orion did it putting music first, gut wrenching resolution second. And this is one thing I consider important when it's meant to enjoy the music, not to suffer from it. 🙂
I've also done some analysis on different music genres. Primary on their frequency range and dynamics. What is adequate frequencyrange and what kind of SPL levels system has to reproduce to be authentic with different recordings. Naturally every album can also be played at modest levels, this is just one aspect I wanted to pay attension in this design.
Then there are few milestones in overall design that I considered important when it comes to natural sound reproduction. Room problems is very important one. Couple earlier posts had some thoughts about this aspect.
Adequate SPL and dynamics capacity is another. Low distorsion and high effieciency I'd consider key features here. One of them isn't enough, we need both. High efficiency creates easily high SPL but without low distorsion it's just noise. Unbearable. And low distorsion with low efficiency isn't much better. Dynamic drivers don't take that much RMS power without significant thermical problems. For example normal home midwoofers can handle maybe 30-40W and then adding more RMS power doesn't do any good. Of cource stated power handling is higher but that's just the point where voicecoil burns. Actual problems are met earlier. And if driver needs relatively high power to sustain needed RMS level it suffers from thermical problems already and isn't cool enough to handle transients. This is what I personally call the "lightbulb-effect".
This is why I don't even want to use big amps. It's easier to build system 3dB more efficient than use two times more watts to drive it. Don't have the measurements yet but I expect to have pretty decent figures from my own construction. Esotar is 92-93dB sensitive 8 ohm unit without the waveguide. Guide adds 0-6dB depending on frequency. Most to the bottom of the range rolling off towards mid treble. This is ok since most of music materials intensity is below 3500Hz. Above from there that much power isn't needed. Uppermidrange units with 1st to 2nd order 250-300Hz cross should be pretty much dipole eq free and so they can play as loud as they can. W18W is stated about 87-88dB sensitive and pair of them creates 93-94dB sensitive nominally 4 ohm load. 25W pair can't reach those figures but they gain some boost from the floor on their range and lowest cross to actual woofers (80-100Hz) assists more. Overally not that much power isn't needed, RMS power stays low, voicecoils cool and transient response fast.
I've heard that some big high capacity speakers have suffered tweeter failures under high SPL listening. I think this is caused by the lack of efficiency on the tweeter. They can't handle that much power and are "just" 90-92dB sensitive. So demanding 110dB+ levels in large room isn't exactly riskfree. Other frequencyareas can have multiple drivers with larger high power voicecoils so no problems for them. Mechanically tweeters can handle quite much. For example Dynaudio tested their Esotar up to 1kw 10ms burst and it didn't cause compression. Esotar doesn't have much linear excursion but there is plenty of mechanical excursion potential so they don't bottom.
Using Linkwitz spl_max spreadsheet I predicted that Esotars runs out of linear range with 2Khz cross in that waveguide somewhere around 114dB level. Uppermids can handle almost 130dB, lower mids almost 120dB. Naturally these depend on used crossover orders and numbers are pretty theoretical. I won't be having amp big enough to run W18s over their linear range. This arrangement should keep the drivers secure and on their usable excursion range in every situation that listeners ears can handle.
Constructions mechanical structure is also important. For example Wilson has pressed pretty excellent performance from relatively inexpensive units by using very inert and well build cabinets. Acoustic designing could be better but I don't think Wilsons are played in a small room anyway. Linkwitz has/had very good experimental and theoretical information on his site. At least the "recoil" research wasn't there when I searched some time ago. But this can also mean that I didn't searched properly enough.
Mechanical problems can cause a lot of unwanted things. In Orions case whole panel is shaking. Partially because woofers have high mms, panel (as a kit) isn't that heavily build and woofers also move front-back motion. There is no force cancellation. And the woofers are also in same cabinet with the mid-treble units. Compact but compromissed. (Still excellent speaker)
Linkwitz had this recoil research some time ago on his site. He had experimented and measured the effect of panels opposite motion from cones motion. So if cone moves forward it causes the panel move back with the same force. And if the panel is light and not supported enough it will move and shorten cones effective excursion and therefore cause compression to played transient. Intresting point.
Other related problem is panels modulation problems. Panel shouldn't shake at all to keep drivers operation linear. There isn't that much use to have low distorsion driver motors and pistonic cones if whole system shakes and modulate signal. I once calculated one very pessimistic model of this problem. Lets consider panel is 100% stiff and 100% of drivers motion is transfered to move the whole panel. My panel is about 130 x 34cm. That's 4420cm2. Install couple 10" woofers on it. That's 660cm2 leaving 3760cm2 bare panel surface. So 660cm2 should be creating sound and 3760cm2 should be totally quiet. Calculating relation to there it means that moving 660cm2 cones 1mm forward panel has to move just 0,17mm back to cancel whole sound.
Like said. This is very pessimistic model. Panel isn't 100% stiff, it doesnt create 100% modulations and so on. But the problem is there. Just think how small 1" domes movements are. And how messed up it's sound is if the panel rocks even small amounts. Modulation distorsion.
Therefore my panel project propably looks like a helluva overkill but I'm hopeful that I can dodge some of the mensioned problems. 130g mms, very stiff propably over 60kg panel with large area bottomplate and spikes. Should be sturdy platform to operate.
This is also why I wanted to use separated woofers in my own system. Bass needs most displacement, bass needs most conearea. So bass propably has most moving mass and therefore more shaking problems. So using physically separated woofers doesn't allow mechanical shaking to transfer into main panel. Even if the woofers dance around the room frequences above 100Hz aren't messed.
One idea is to make the W-baffle dipole colums from 12mm steel. Quickly calculated mass would be way above 100kg for each unit. Four 12" Peerless SLS woofers on each side with force cancellation installation. If they dance around they also cause significant damage to decoration... 😀
Am I delusional or what.. 😎
Jussi
Delusional? No, but you seem pretty committed to the closest to 100% solution you can come up with. I can envy that, as well as your apparent budget. I mean those B&G Neo3PDRs I recommended have been the priciest drivers in my house for a long time now ;-p
Anyway, I can see you're a couple orders of magnitude more dedicated to researching these things than I am. I don't have any measuring gear other than my ears, so I can determine some things while I'm trying to build a detailed EQ set through my PC (E-Mu 0404 sound card; gives me nearly limitless control in the digital domain), but mainly, when sit back to listen to some real content, I hear alot of mistakes made while setting up the EQ (mainly needing to relax cuts made in the mids and highs; sounded flat on the frequency sweep but sounds like crap for playing real music.
The proof is in the pudding, as they say. If what I hear from a given bit of kit is good (or I can tweak it to make it sound good), then its good for me. Just to let you know, my bias is toward Magneplanar speakers, if we're talking commercial (and if I actually had a budget, I would buy them and hopefully be able to tone the the DIY fervor). I've got a high end B&W shop in town, and have listened to their $16,000/pair 800 series speakers, and while they did sound good, but but my $50 OBs in the office sound at least 80% as good and my modified Yamahas in the living room (figure about $200 since this is where the Neo3s are right now) are more like 85%. My rough guess from checking your work here is that you're going to get something that would IMO sound 110%-120% as good as these B&W for what, about a tenth of the price (note I will be biased toward your approach as I much prefer OB/dipole sound to even the best boxes in the world).
My question is, will your sound be able to beat Magnepan? My favorite that I've listened to were the 1.6QR (sure the 3.6R had better more/lower bass, but I didn't care for the 3way nature or the ribbon tweeter). I figure if you build something that's more enjoyable to listen to than a Maggie 1.6 for reasonably similar price, then you've done something remarkable.
Kensai
Anyway, I can see you're a couple orders of magnitude more dedicated to researching these things than I am. I don't have any measuring gear other than my ears, so I can determine some things while I'm trying to build a detailed EQ set through my PC (E-Mu 0404 sound card; gives me nearly limitless control in the digital domain), but mainly, when sit back to listen to some real content, I hear alot of mistakes made while setting up the EQ (mainly needing to relax cuts made in the mids and highs; sounded flat on the frequency sweep but sounds like crap for playing real music.
The proof is in the pudding, as they say. If what I hear from a given bit of kit is good (or I can tweak it to make it sound good), then its good for me. Just to let you know, my bias is toward Magneplanar speakers, if we're talking commercial (and if I actually had a budget, I would buy them and hopefully be able to tone the the DIY fervor). I've got a high end B&W shop in town, and have listened to their $16,000/pair 800 series speakers, and while they did sound good, but but my $50 OBs in the office sound at least 80% as good and my modified Yamahas in the living room (figure about $200 since this is where the Neo3s are right now) are more like 85%. My rough guess from checking your work here is that you're going to get something that would IMO sound 110%-120% as good as these B&W for what, about a tenth of the price (note I will be biased toward your approach as I much prefer OB/dipole sound to even the best boxes in the world).
My question is, will your sound be able to beat Magnepan? My favorite that I've listened to were the 1.6QR (sure the 3.6R had better more/lower bass, but I didn't care for the 3way nature or the ribbon tweeter). I figure if you build something that's more enjoyable to listen to than a Maggie 1.6 for reasonably similar price, then you've done something remarkable.
Kensai
This is what i waiting for.I hope that i see differences between fabric and metal domes.
You know discussion about fabricdome in waveguide and its bad behaviour in it?Please say to me,myth busted!
- Status
- Not open for further replies.