Tweet
BTW, does anyone know how to setup wavelet analysis in SE?
-
Somehow I managed to not get printed the newest 5" and 8" for the SB26 :( But since I have three dome variants now and can measure out to 48khz, I thought I'd revisit the 5" vG and 8" vC. None these will have phase shields as I already cut them off months ago.
First is the fabric dome SB26STAC:
Aluminum done SB26ADC:
Ceramic dome SB26CDC:
I wanted to revisit using an ellipse as the profile for the flare. Like this:
Interesting results, I'll have to dig into this later.
Next is the little SB21SDC:
And last, the 8" RST28 designs:
Up next will be some 4" designs. I see some value in a small waveguide used to cross higher to a smallish mid. Benefits: still matches mid directivity, allows mid to cover much of vocal range without a wonky polar response, acoustic centers align.
After that I need to play with phase shields in shims and see what happens. I'd really like to get lucky and notch out those metal dome breakups.Comment
-
Noticed the fabric SB26 has a small felt ring on top of the pole piece, while the metal versions do not:
Let's see if this new phone can capture these waveguides better:
Comment
-
"I wanted to revisit using an ellipse as the profile for the flare. Like this:..."
I couldn't copy and paste the above(post 513) accompanying image I refer to, looking at it though, if the left and right cross sections are elliptical, what are the vertical cross sections?Comment
-
I stayed with the existing constant radius (a circle section). That model was a huge PITA so I stopped while I was ahead! I should mention also this was a 1/4 ellipse, so the throat angle was 0 and mouth was 90 (tangent to baffle).Comment
-
Here are the 4" waveguides. One keeps the mouth as a golden ellipse (1:1.618) but the others I went with 1:1.67 since it I didn't want to get too small vertically.
Comment
-
-
Brandon, I can't open the .f3d file, short of downloading Fusion 360. Can you export a .iges, .dwg, .stl, or .dxf file? I would like to see what the cross sections are. ThanksLast edited by TN Allen; 11-19-2018, 09:11 PM.Comment
-
.f3d is the best way to pick apart the model, come on Tim, all the cool kids are using Fusion360! I'll try to get an .iges up when I get a chance.Originally posted by TN Allen View PostComment
-
Just starting to pore over the results right now, and I think the versions G and H_e are interesting. G has a constant radius flare, basically a section of circle, with radius = 2.6", while H_e is using an ellipse for the flare (horizontally, I left the vertical flare the same as G).
And the two curvatures overlaid to see teh difference:
My assumption is that the difference was primarily due to the vertical throat of H_e causing the expansion in the 12khz area. So perhaps using a smaller radius on G would have a similar effect, since that will narrow the throat angle. However it will also push the wall out in front of the plane of the mouth just before it merges with the front baffle.
However just when I think I've found a causation I look at the 4" vA, which has exactly that smooth response around 12khz I want, but it has an even wider throat than G by about 4 degrees!
Incidentally, both G and this 4" A have a flare radius of 2.6".Comment
-
Don't bother, it's not that important.Originally posted by augerpro View Post
Perhaps the "cool kids" use it because it's free.Comment
-
-
John,
I don't know if the following link will be of any use:
https://www.autodesk.com/campaigns/f...for-hobbyists#
I'll ask a colleague who was using it if he has continued past the free offer. He's faculty though, and probably has it free for life if he wants it.
I suspect Autodesk has completed their initial marketing offer, and has moved on to signing up users and collecting license fees. It's a well thought out comprehensive program, and the training material is very well done, and it is particularly attractive to Mac users. But, Autodesk needs to make a profit on their investment of resources.
Perhaps Brandon has found a way to continue free use. He's a bit past 2 years using it.
Comment
Comment