AVS Forum banner

781 - 800 of 12779 Posts

·
Premium Member
Joined
·
24,637 Posts

Quote:
Originally Posted by jzagaja /forum/post/19691920


Now with screws and 16deg throat we cut to exact diameter.

In the view of the WG's horizontal profile at the right, how does the radius of the transition from the 16 deg throat angle to the main section of the horn compare with other designs, especially Geddes'?


It looks kind of abrupt, and Geddes stresses the importance of minimizing curvature.
 

·
Registered
Joined
·
7,535 Posts

Quote:
Originally Posted by LBDiver /forum/post/19691749


Erich,


If you look at the shading in Jzagaja's 3D rendering the transition radius terminates right at the edge, where as my image (which you marked above) has 1/4" added around all sides. Adding to the height only would work, however I feel it needs to be 1/2" top and bottom so the screws can be countersunk without the risk of edge cracking. The 1/4" on the sides may be advantageous as well so there's wide enough corresponding rabbet on the baffle for gasket tape.

Screw holes won't be done by the manufacture because it will be too hard and mess up the finish used. It should have an option for screwless mounting, or 4 screws, or 8. That will satisfy the most people here. Anyone should be able to drill a hole where they want.


LBDiver, I agree about having the ability to put some screws in if people want them. That's fine. But I just don't see why we need a full 1/2" extra flange. With the placement I showed, there's no need for that much to be added unless the screws were going to be put right in the middle area. But the middle area only needs to have a 1/4" flat area because no hole will be put there, the spot where the screws will be placed would be closer to 1/2" anyway just because of the elliptical shape.


jzagaja's 2nd rendering shows a 3/16" flange. A person could shift those screws further towards the middle to close off the gap and still have plenty of room. This seems like a win/win situation to me.
 

·
Registered
Joined
·
429 Posts

Quote:
Originally Posted by Erich H /forum/post/19692468


Screw holes won't be done by the manufacture because it will be too hard and mess up the finish used. It should have an option for screwless mounting, or 4 screws, or 8. That will satisfy the most people here. Anyone should be able to drill a hole where they want.


LBDiver, I agree about having the ability to put some screws in if people want them. That's fine. But I just don't see why we need a full 1/2" extra flange. With the placement I showed, there's no need for that much to be added unless the screws were going to be put right in the middle area. But the middle area only needs to have a 1/4" flat area because no hole will be put there, the spot where the screws will be placed would be closer to 1/2" anyway just because of the elliptical shape.


jzagaja's 2nd rendering shows a 3/8" flange, and I still think we can go a bit smaller and still fit the screws in with no problem. He could shift those screws further towards the middle and close that gap off.

Yes I realize there will be no pre-drilled holes, and have just been discussing having a flange there for people to screw through if they choose.


Regarding it being a 1/2" flange, I think that it is necessary now knowing the Jzagaja's last rendering has and additional 3/8" lip. It makes clear that the SEOS profile tapers so slightly from the centerline that in order to keep the head off the transition radius they have to be moved much further to the sides. You must remember that this is not ABS or aluminum and that there is a real risk of splitting/cracking if the screws are placed too close to the edge. It seems 1/2" will be required to get the screws where you and I placed them, without risking it.
 

·
Registered
Joined
·
139 Posts
I haven't seen Geddes WGs so I can't comment. Construction is simple - at 16deg we cut OS profile and add conical duct, after lamination fine tune to desired diameter. There's no abrupt change in 3D model or a mold that is hand polished. Desired included angle is in both planes and required much more calculations than axisymmetrical type.




Quote:
Originally Posted by noah katz /forum/post/19692210


In the view of the WG's horizontal profile at the right, how does the radius of the transition from the 16 deg throat angle to the main section of the horn compare with other designs, especially Geddes'?


It looks kind of abrupt, and Geddes stresses the importance of minimizing curvature.
 

·
Registered
Joined
·
429 Posts
Erich,


I see you updated your post to reflect only a 3/16" (5mm) added flange. I mocked up the inset screw position you and I used, based on the screw diameter of 11mm in the picture, it looks like we need another ~9mm added to maintain proper spacing from the edge and transition radius, it's just the nature of the SEOS profile . Therefore a 14mm flange top and bottom should cover it.

 

·
Premium Member
Joined
·
24,637 Posts

Quote:
Originally Posted by LBDiver /forum/post/19692611


You must remember that this is not ABS or aluminum and that there is a real risk of splitting/cracking if the screws are placed too close to the edge.

This wouldn't be an issue if the screws were counterbored instead of countersunk.


That causes only compressive stress under the head, whereas the wedge effect of the conical head of a countersunk screw is an ideal way to cause splitting.


The screws could be low head height socket screws like McMaster 92220A166 or; 10-24 x 1" long, .100" head height, .361" head diameter.


If you don't have a counterbore you could use a 3/8" brad point or Forstner bit.
 

·
Registered
Joined
·
1,957 Posts

Quote:
Originally Posted by jzagaja /forum/post/19692744


I haven't seen Geddes WGs so I can't comment. Construction is simple - at 16deg we cut OS profile and add conical duct, after lamination fine tune to desired diameter. There's no abrupt change in 3D model or a mold that is hand polished. Desired included angle is in both planes and required much more calculations than axisymmetrical type.

Noah, is correct, this is not how Geddes implements the OS profile to match the exit angle of the CD.


Instead he traverses the X-axis until the tangent line of the OS profile forms the targeted exit angle and begins the OS progression at that point.


You can see a graphical explanation by downloading this spreadsheet:
http://www.musicanddesign.com/codes/OS_wave_guide.zip


I believe it is rather material that this principle is adhered or the OS is basically wasted.
 

·
Premium Member
Joined
·
24,637 Posts
Quote:
Originally Posted by coctostan
Instead he traverses the X-axis until the tangent line of the OS profile forms the targeted exit angle and begins the OS progression at that point.
It looks like jzagaja's does that, but it happens over a small distance because of the high curvature/small radius.
 

·
Registered
Joined
·
26,476 Posts
Quote:
Originally Posted by noah katz
+1


Penngray, I'm puzzled by your comments about looks, I thought you said you didn't care?


I think bare horns above a box are very unsightly.


Besides that, many of us want to minimize box volume and don't want to lose that available around the WG.
I have always cared a great deal about looks. I hated even the QSC for looks but its a great waveguide. I always use boxes but I have learned that the waveguide needs to be protected from the woofers back wave too....the seal exists still no matter what has been posted. My way or other ways..the seal exists so you either care about looks or you dont care.....if you do not care about looks then do not screw everyone else out of having better looks. Difraction occurs at the mouth and baffle as much as it does at the throat....we are trying to minimize that are we not??


I think the whole discussion has been way overblown. To think that people need 8 screws to seal a waveguide just lacks common sense and its full of stubborn attitudes. There are substances called adhesives that actually work in the world of speakers. Many use them ALONE without screws to seal the design. We can seal this design without 8 to 10 screws that is the crux of the discussion.


Anyways, I posted 4+ times already that Im on board with the final product whatever it is and still debate goes on



Although if we are having a flange then a 1/2" flange is again another extreme overkill IMO. People need to learn the stress points and the amount of thickness needed for common screws. This product won't split like they think it will if done properly. Npw we have 1" more width in the waveguide we also have a curve to flat to curved transition in the waveguide. Less flat surface the better, it should transition from curve to curve as best as possible. Few here seem to have little experience with JMMC horns and how the curve all the way around.


Maybe its just me but Im just not interested in another PA looking product.
 

·
Registered
Joined
·
1,957 Posts
Quote:
Originally Posted by noah katz
It looks like jzagaja's does that, but it happens over a small distance because of the high curvature/small radius.
It isn't really a choice but dictated by the OS angle, the throat diameter and the throat entry angle. You essentially start part of the way down the OS throat, at the point where the tangent equals your exit angle. The problem is that point is wider than the expected throat diameter if you start your OS math at that point.


Instead you need work backwards from the throat angle to find the point where the tangent line intersects with the intended diameter.


I've marked J's model. I assume he created a conical section that starts at the red arrow. That is incorrect. Instead he should have started at the green arrow. (sorry for the rough drawing)




Edit: To further clarify my point. It appears Jzagaja has added a conical section that is 16 degrees to bridge the gap where the tangent occurs and the 1" mouth. I can see now why he said it was so difficult. Once he takes a look at that spreadsheet I think it will actually make his model much easier to create.


IMO, this concept is central to the OS profile. It incorporates the the CD's throat angle into the OS profile, transitioning as smoothly and soon as possible to the OS profile. When Geddes has spoken about his perfect CD I think he would like to start to the OS profile deeper into the CD itself thus eliminating the conical portion altogether.
 

·
Premium Member
Joined
·
24,637 Posts
Quote:
Originally Posted by penngray
1. I have always cared a great deal about looks.


2. Now we have 1" more width in the waveguide we also have a curve to flat to curved transition in the waveguide. Less flat surface the better, it should transition from curve to curve as best as possible.
1. I must have mixed you up with someone else then


2. What's the matter with flat? The radius is infinity = no diffraction
 

·
Registered
Joined
·
26,476 Posts
Thanks coctostan! that is a very important post.


The connection of the CD should be starting at the green arrow, correct? but that isnt 16deg.
 

·
Registered
Joined
·
26,476 Posts
Quote:
Originally Posted by noah katz
1. I must have mixed you up with someone else then


2. What's the matter with flat? The radius is infinity = no diffraction
1. I sometimes spend more on veneer then I do on drivers and wood
and I do spend more on drivers then most others. I do not like


2. maybe! I have just seen the continous curve of the JMMC to create the least amount of issues in wavelet plots.



Either way, I think the discussion about the throat is far more important.
 

·
Registered
Joined
·
1,957 Posts
Here is a screenshot from the spreadsheet further explaining my point about the throat:




You can see that where the two lines intersect it is at some point past x in the y(x) OS equation. This is the tangent point but it is also .5", the throat radius. The OS must start there.


I'm very confident in this, but as always I'm just an amateur so if someone disagrees please feel free to post. I've been wrong before.
 

·
Registered
Joined
·
1,957 Posts
Quote:
Originally Posted by LTD02
what is the purpose of the region between the red arrow and the green arrow in coctostan's post?

http://www.avsforum.com/avs-vb/showt...8#post19694158


in other words, why not just cut the horn at the green line and mount the cd there?
I believe he did that because the point at the green line is a diameter greater than the 1" specified. That will always be the case unless the driver exit angle is 0 degrees.


That is why, for each waveguide angle on the ellipse, the x value for the tangent point corresponding with the driver exit angle must be found. From that point, you then use the remaining OS equation to chart the points.


You have two extremes:
  1. Exit angle = 0 degrees. If we were targeting a CD with this exit angle, the entire OS profile would be used and the calculation would be very simple.
  2. Exit angle = waveguide angle. For instance if the exit angle was 45 degrees and the waveguide angle was 45 degrees, the waveguide would actually just be conical. It would only match the OS profile at OS profile's termination.


You shouldn't simply calculate the OS profile for 0 degrees, find the 16 degree point, chop off the portion before that and replace it with a conical section to complete the transition back to the intended entry diameter. It is sub-optimal and unnecessary.
 

·
Registered
Joined
·
21,343 Posts
Discussion Starter #797
so you are saying that chopping off the horn at the green arrow would be a fail because it would not follow the olbate spheroid shape, but that once considering the geometry inside the driver itself, it may be better to go ahead an chop off the horn at the green arrow because the driver itself includes something of the area between the red arrow and the green arrow?
 

·
Registered
Joined
·
7,535 Posts

Quote:
Originally Posted by LBDiver /forum/post/19692800


Erich,


I see you updated your post to reflect only a 3/16" (5mm) added flange. I mocked up the inset screw position you and I used, based on the screw diameter of 11mm in the picture, it looks like we need another ~9mm added to maintain proper spacing from the edge and transition radius, it's just the nature of the SEOS profile . Therefore a 14mm flange top and bottom should cover it.


LBDiver, I was looking at this again because something just didn't seem right. I then realized that the screws were awefully big in that 3d drawing. 11mm heads are massive for this project aren't they? Maybe I'm having a slight brain lapse, but I don't think anyone is going to use a screw with a head that big for these WG's. Any thoughts?
 

·
Registered
Joined
·
429 Posts

Quote:
Originally Posted by Erich H /forum/post/19694941


LBDiver, I was looking at this again because something just didn't seem right. I then realized that the screws were awefully big in that 3d drawing. 11mm heads are massive for this project aren't they? Maybe I'm having a slight brain lapse, but I don't think anyone is going to use a screw with a head that big for these WG's. Any thoughts?

Erich,


I did note the rather large size, but figured Jzagaja spec'd them for a reason. A #8 panhead is still ~8mm and a #10 ~9.5mm, thereby only gaining .75-1.5mm on the inside radius over spec, as I would not move the holes any closer to the edge of the flange. This would equate to a reduction of suggested flange width from 14mm to 12.5mm if using #8's vs Jzagaja's spec'd screws.
 
781 - 800 of 12779 Posts
Top