The loading depends on the depth, and the lower limit for an 1" driver depends partly on the safety margin one desires, the drivers resonance frequency, and desired SPL levels. So far I am not even sure if I want a 1" or an 1.4" driver! Regardless of what I posted above I still feel more comfortable with the idea of an 1.4", because my intended WG is shallow, hence less loading below 1k.
Similarly, if 21" WG, then x-o must be lower than 900 Hz - that's what I meant in my previous post, that the WG forces the x-o point to some extent, due to size.
Pattern matching: depends how precise you want to get. Say a 15" driver has easily 2" cone depth and my prototype WG has 4" depth, remains delta depth 2", compared to x-o wavelength, not much. If so desired with an active x-o the remaining delay can be removed.
WRT ideal pattern matching: only coaxial designs can hope to match the patterns ideally, and those have their own problems (imprecise waveguide made of the cone of the midrange, which itself moves, so little hope to optimize for low HOM creation etc. Everything else will have vertical offset, so only the horizontal plane has a hope of good x-o pattern matching and some vertical lobing will occur. To alleviate all this, Paul intends to use brickwall filters, I use a DIY analog active x-o so I'll stick with 4th order LR maximum, and will have to live with some irregularities. OTOH unless you go to great digital lengths brickwall filters have their own issues, group delay for instance. So it's all in trade offs and experimentation will be necessary.
Beaming and transducer diameter: this should be covered in any good book on speaker design, for instance in my copy of the "Loudspeaker cookbook" (Dickason) there is a discussion on it, with graphs. The cutoff is not a brutal one, and depends on driver geometry / material to a small extent, but the general transition area from full space radiation to directional radiation (i.e., beaming) is usually given as wavelength = 1/3 circumference, which is roughly equal to wavelength = diameter (circumference = 2 pi R and 2R = diameter and pi = 3.14). So, for 15" diameter we're back to the 900 Hz you quoted, which coincides with my measurements of the 15" WG prototype above in this thread (pattern widens at that point), and with the x-o Earl Geddes chose in the Summa for transition between 15" cone -> 15" WG (with 1" compression driver).
Hello Mr. Paul W,
I remember seeing a picture with a a pair of 10'' 18sound drivers you posted on another forum. Did you had the chance to listen them or use them in your sistem?
I remember seeing a picture with a a pair of 10'' 18sound drivers you posted on another forum. Did you had the chance to listen them or use them in your sistem?
Member
Joined 2003
Yes, I listened to the 10NDAs for a few weeks. They were extremely well behaved from 200-1.5k. Didn't try them below about 200 and don't recommend them above 1.5k because it is so close to the (fairly benign) breakup above 2k. Factory graphs are legit, and the AIC really works.
I changed my goals along the way, so the 10's were not directional enough for the revised plan, but otherwise no problem.
I changed my goals along the way, so the 10's were not directional enough for the revised plan, but otherwise no problem.
Hello,
thank you for the reply. This thread and some others around here where particulary inspiring for me.
As I understand the goal was a certaim directivity pattern at about 900Hz to match the one from a waveguide. Could you tell me if you decided on a 15'' driver or you will use the 4 drivers baffle you posted pics with?
I'm asking because I really try to understand what are the efects of different directivity patterns at the xo point. I am somehow between this aproach, which works in the Summa speaker and Mr. Lynn Olson's opinions who believes constant directivity is not a big issue.
thank you for the reply. This thread and some others around here where particulary inspiring for me.
As I understand the goal was a certaim directivity pattern at about 900Hz to match the one from a waveguide. Could you tell me if you decided on a 15'' driver or you will use the 4 drivers baffle you posted pics with?
I'm asking because I really try to understand what are the efects of different directivity patterns at the xo point. I am somehow between this aproach, which works in the Summa speaker and Mr. Lynn Olson's opinions who believes constant directivity is not a big issue.
Member
Joined 2003
Sun Ra,
I agree with Earl regarding the benefits of constant directivity. An area where we differ is the use of a 15" up to 900/1k. Though certainly not a mainstream configuration, the 2x2 midrange works well and that's what I'll use for the front channels until something better pops up. Besides directivity control, dynamic headroom is dramatically increased.
Brett,
I tried the 10NDA520. The 610 was also considered but it seems better suited to a slightly higher range than the 520.
Paul
I agree with Earl regarding the benefits of constant directivity. An area where we differ is the use of a 15" up to 900/1k. Though certainly not a mainstream configuration, the 2x2 midrange works well and that's what I'll use for the front channels until something better pops up. Besides directivity control, dynamic headroom is dramatically increased.
Brett,
I tried the 10NDA520. The 610 was also considered but it seems better suited to a slightly higher range than the 520.
Paul
Paul,
have you made progress on the 21" horn (what are you using for the HF part now?).
On my side lack of play money precluded much progress. I tried crossing my 15" higher to check for differences with the 6.5". No real dramatic differences, but still - somehow smoother with the 6.5" crossing at 300 or 400 Hz. So I now tend more towards a WG-CD combo coming in at higher frequencies to replace my dome tweeter, say 1.5-2k, with a true midrange in the middle.
I did also experiment with Earl's suggestion in the Summa paper to toe in the speakers substantially, so that most first reflections come from the *opposite* wall with more delay than they would from the proximal wall. So far I like 20-40 degrees toe in best, more and it becomes too "phasey" for my taste and the image too diffuse. But maybe mine are not directional enough for this.
have you made progress on the 21" horn (what are you using for the HF part now?).
On my side lack of play money precluded much progress. I tried crossing my 15" higher to check for differences with the 6.5". No real dramatic differences, but still - somehow smoother with the 6.5" crossing at 300 or 400 Hz. So I now tend more towards a WG-CD combo coming in at higher frequencies to replace my dome tweeter, say 1.5-2k, with a true midrange in the middle.
I did also experiment with Earl's suggestion in the Summa paper to toe in the speakers substantially, so that most first reflections come from the *opposite* wall with more delay than they would from the proximal wall. So far I like 20-40 degrees toe in best, more and it becomes too "phasey" for my taste and the image too diffuse. But maybe mine are not directional enough for this.
Member
Joined 2003
MBK,
Speaker building has stopped for now. I am moving to southern US, and will get a dedicated AV room...so won't have to please everyone with speaker size and appearance 😉
The large WGs will use a 90 degree OS flare with a very large radius roundover at the mouth. Construction will probably be done by first building a male mold and then casting over the mold. (I need three WGs for left, center, and right main channels.)
Though more complicated, 3-way should give you improved performance over a 2-way. Will you stay with Scan Speak mid or go with something else?
Paul
Speaker building has stopped for now. I am moving to southern US, and will get a dedicated AV room...so won't have to please everyone with speaker size and appearance 😉
The large WGs will use a 90 degree OS flare with a very large radius roundover at the mouth. Construction will probably be done by first building a male mold and then casting over the mold. (I need three WGs for left, center, and right main channels.)
Though more complicated, 3-way should give you improved performance over a 2-way. Will you stay with Scan Speak mid or go with something else?
Paul
Nothing wrong with the SS, very smooth and natural to my ears. But you're right to insinuate... eventually it'll make more sense to upgrade it to a small diameter pro driver too. That way all the drivers will be high efficiency. Besides, those pro drivers seem to be somewhat addictive! So, say from 18Sound again, a 6ND410 would fit nicely... or their 8".
But first things first, my dome tweeter is clearly the limiting factor now. So I'll probably try an affordable-yet-good compression driver like the ND1020 on a 120 degree conical or OS WG designed to cut off around 1.6k to 2k. I want to try a conical as well, just to see how it behaves at angles >90 degrees (where there can be only one reflection within the WG, geometrically).
A dedicated room, well this makes some things a lot easier... Good lck with your move. I always need to keep in mind not just room space, appearance, and size, but also kid and cleaning friendliness and possible transportability in case of a move. So this introduces a lot of design constraints, but hey that's OK, it's part of the challenge...
But first things first, my dome tweeter is clearly the limiting factor now. So I'll probably try an affordable-yet-good compression driver like the ND1020 on a 120 degree conical or OS WG designed to cut off around 1.6k to 2k. I want to try a conical as well, just to see how it behaves at angles >90 degrees (where there can be only one reflection within the WG, geometrically).
A dedicated room, well this makes some things a lot easier... Good lck with your move. I always need to keep in mind not just room space, appearance, and size, but also kid and cleaning friendliness and possible transportability in case of a move. So this introduces a lot of design constraints, but hey that's OK, it's part of the challenge...
Member
Joined 2003
Actually, I was just curious about the mid...no insinuation intended as SS builds really great products.
Switching to a CD crossed higher gives you a huge range of possibilities. On the other hand, if you get a chance to try one, conical may be a good match for the existing dome. Conical with a CD may introduce a lot of diffraction at the driver/throat intersection?
Switching to a CD crossed higher gives you a huge range of possibilities. On the other hand, if you get a chance to try one, conical may be a good match for the existing dome. Conical with a CD may introduce a lot of diffraction at the driver/throat intersection?
Diffrection: yes, but only once:
The rationale for the conical is this: if you do a horn or Geddes WG, it starts out at almost 0 degrees angle to axis. Now if this angle is well matched to the driver, and the profile is an OS, then in theory you get no HOMs. In reality, you get less HOMs and the matching to the CD may turn out tricky.
Now, if the inclusive angle of the straight WG section is >90 degrees, geometrically there can be at most 1 reflection (angle in = angle out w.r.t the orthogonal to the reflecting plane). So by *not* adding a less than 90 degree mouth section, i.e. by keeping it conical all the way, your worst case is one single diffraction/reflection event at the throat. By building a less than perfectly matched throat section, you have potential for a lot more bouncing reflections.
But, one has to try... and since conicals are easy to DIY I'll start out as you suggested, on the dome, to see what it is actually doing in practice. Who knows, maybe that's all I need and it sounds great ;-)
The rationale for the conical is this: if you do a horn or Geddes WG, it starts out at almost 0 degrees angle to axis. Now if this angle is well matched to the driver, and the profile is an OS, then in theory you get no HOMs. In reality, you get less HOMs and the matching to the CD may turn out tricky.
Now, if the inclusive angle of the straight WG section is >90 degrees, geometrically there can be at most 1 reflection (angle in = angle out w.r.t the orthogonal to the reflecting plane). So by *not* adding a less than 90 degree mouth section, i.e. by keeping it conical all the way, your worst case is one single diffraction/reflection event at the throat. By building a less than perfectly matched throat section, you have potential for a lot more bouncing reflections.
But, one has to try... and since conicals are easy to DIY I'll start out as you suggested, on the dome, to see what it is actually doing in practice. Who knows, maybe that's all I need and it sounds great ;-)
Member
Joined 2003
Yes, I remember the dome and >90 degree discussion from Earl's thread. Really curious how it turns out so please post results if you have a chance.
Hello,
Paul, thanks for the reply regarding constant directivity and I'm sorry for the confusion with the 10''-ers. I have the same doubts regarding the 15" drivers but that's because of my limited experience with drivers. From an intuitive point of view I would give a try to a 15'' pro audio driver with a leight cone, strong motor and most important ruler flat up to 2Khz at least. Alnico would be a bonus.
Until now the candidates would be the TL1601 (c version is out unfourtunately, remains a and b), BD-design 15'' neodynium woofer and maybe the supravox 400-2000 alnico or exc. Ofcourse various PA drivers could be suitable, not to mention the one Dr. Geddes is using (i'd like more than 96dB though). All these drivers have stunning frequency responce (ruler flat, with the TBX having a little rise at about), most of them beeing optimised for low distorsions. Supravox is probably more vintage orientated but that's just a "feeling".
I'd certainly try to compare these with more traditional midranges like SS, Seas, Audio Technology and others.
@MBK,
I know you own a 15ND930? I hope I'm not mistaking again how I did with Paul's midranges 😀. Did you tried to run it up to 900Hz and compare it to something more conventional? I remember seeing some comparative mesaurements against a seas W26 magnesium but I didn't spot any listening comparions. I just hope I'm not making some confusions again 😀.
Cheers
(edit) MBK, you mentioned something about a paper on the Summa speaker. As you know, Dr. Geddes is now working on another speaker and the page on his site about the Summa is not anymore. Could you direct me to a link for this paper?
Paul, thanks for the reply regarding constant directivity and I'm sorry for the confusion with the 10''-ers. I have the same doubts regarding the 15" drivers but that's because of my limited experience with drivers. From an intuitive point of view I would give a try to a 15'' pro audio driver with a leight cone, strong motor and most important ruler flat up to 2Khz at least. Alnico would be a bonus.
Until now the candidates would be the TL1601 (c version is out unfourtunately, remains a and b), BD-design 15'' neodynium woofer and maybe the supravox 400-2000 alnico or exc. Ofcourse various PA drivers could be suitable, not to mention the one Dr. Geddes is using (i'd like more than 96dB though). All these drivers have stunning frequency responce (ruler flat, with the TBX having a little rise at about), most of them beeing optimised for low distorsions. Supravox is probably more vintage orientated but that's just a "feeling".
I'd certainly try to compare these with more traditional midranges like SS, Seas, Audio Technology and others.
@MBK,
I know you own a 15ND930? I hope I'm not mistaking again how I did with Paul's midranges 😀. Did you tried to run it up to 900Hz and compare it to something more conventional? I remember seeing some comparative mesaurements against a seas W26 magnesium but I didn't spot any listening comparions. I just hope I'm not making some confusions again 😀.
Cheers
(edit) MBK, you mentioned something about a paper on the Summa speaker. As you know, Dr. Geddes is now working on another speaker and the page on his site about the Summa is not anymore. Could you direct me to a link for this paper?
Hi SunRa,
yes I own the 15ND930. Usually I cross it at 320 Hz to SS8543 (polyethylene 6.5"). I spent a week or so crossing it at 630 Hz for comparison, that's the highest I can do right now because 1) in my current physical layout the driver separation becomes too large above that; 2) the mid crosses to the dome tweeter at 2k. If the mid HP is too high it starts interfering with the 2k LP; that's fixable by tweaking Q and gain but I didn't want to go to the trouble (no DEQX here, analog active X-O). If I take out the mid I still can't reach below say 1.5k with the dome and 1.5k is too high for the 15" (breakup at 1.7k, mild 3 dB peak only, but still, a clear breakup).
But anyway: the immediate impression with the 15" LP'd at 630 is not too different as to having it LP'd at 320 Hz. In the long run though it does feel a bit smoother with the midrange going down to 320 Hz. It is not a rigorous comparison, just an impression. A major factor that makes the comparison questionable is that the dispersion changes as well between the 6.5" and the 15", precisely in the 300-800 Hz region they measure quite differently in dispersion. So unless I'd build two sets of speakers optimized for their driver characteristics, I can't truly make a fair comparison. It amounts to gut feeling, and my gut feeling is that I get a smoother presentation with a true mid doing the midrange.
Incidentally the nearfield FR of the 15ND930 becomes more raggedy above say 400 Hz, I highly suspect all 15" behave similarly, but MF's data are nicely smoothed out.
Summa paper: I read it when it was still online, a while ago. Maybe you can pm Earl Geddes to get it.
yes I own the 15ND930. Usually I cross it at 320 Hz to SS8543 (polyethylene 6.5"). I spent a week or so crossing it at 630 Hz for comparison, that's the highest I can do right now because 1) in my current physical layout the driver separation becomes too large above that; 2) the mid crosses to the dome tweeter at 2k. If the mid HP is too high it starts interfering with the 2k LP; that's fixable by tweaking Q and gain but I didn't want to go to the trouble (no DEQX here, analog active X-O). If I take out the mid I still can't reach below say 1.5k with the dome and 1.5k is too high for the 15" (breakup at 1.7k, mild 3 dB peak only, but still, a clear breakup).
But anyway: the immediate impression with the 15" LP'd at 630 is not too different as to having it LP'd at 320 Hz. In the long run though it does feel a bit smoother with the midrange going down to 320 Hz. It is not a rigorous comparison, just an impression. A major factor that makes the comparison questionable is that the dispersion changes as well between the 6.5" and the 15", precisely in the 300-800 Hz region they measure quite differently in dispersion. So unless I'd build two sets of speakers optimized for their driver characteristics, I can't truly make a fair comparison. It amounts to gut feeling, and my gut feeling is that I get a smoother presentation with a true mid doing the midrange.
Incidentally the nearfield FR of the 15ND930 becomes more raggedy above say 400 Hz, I highly suspect all 15" behave similarly, but MF's data are nicely smoothed out.
Summa paper: I read it when it was still online, a while ago. Maybe you can pm Earl Geddes to get it.
Paul W said:
Hello Paul, I know this thread has died, but I would like to know what was your absorption method for the rear radiation?
Member
Joined 2003
Hi tnargs,
The backside of the baffle is covered by a 1/2" layer of F7 wool felt wrapped in the form of a half cylinder. Two sheets of heavy wire screen immobilize the felt, inside and out. The screen and felt are screwed to the upper, middle, and lower lips in the baffle through three 1" aluminum strips.
I haven't measured 360 degree polars to validate, but this should result in a quasi-cardioid pattern in the midrange, evolving to dipole in the lower bass. It definitely prevents backwave reflections from spoiling imaging across the front wall.
BR,
Paul
The backside of the baffle is covered by a 1/2" layer of F7 wool felt wrapped in the form of a half cylinder. Two sheets of heavy wire screen immobilize the felt, inside and out. The screen and felt are screwed to the upper, middle, and lower lips in the baffle through three 1" aluminum strips.
I haven't measured 360 degree polars to validate, but this should result in a quasi-cardioid pattern in the midrange, evolving to dipole in the lower bass. It definitely prevents backwave reflections from spoiling imaging across the front wall.
BR,
Paul
Hi Paul,
How "permeable" would 1/2" F7 felt be to sound? I would think that it would be a pretty good "deadening agent".
...love them technical terms.
How "permeable" would 1/2" F7 felt be to sound? I would think that it would be a pretty good "deadening agent".
...love them technical terms.
Member
Joined 2003
Although the F7 wool looks and feels "very heavy and thick" it is surprisingly permeable, particularly at lower frequencies. At higher frequencies, it begins to act more like an absorber.
I've found the variable resistance over a fairly wide frequency range can be used to advantage when matching dipole low bass, through cardioid mid, with a monopole tweeter.
I've found the variable resistance over a fairly wide frequency range can be used to advantage when matching dipole low bass, through cardioid mid, with a monopole tweeter.
- Status
- Not open for further replies.