"Into The Abyss", Tweeters in low places

[scottq]

...Though, I've found with elliptical guides it's best to model it... ​​​​​​

What is used to model waveguides? #newbie

NM, answered it with a little searchy searchy. Sorry.
Good thread here, with Pete and TN contributing a lot: http://techtalk.parts-express.com/fo...guide-modeling

http://www.hornresp.net/ - Apparently, this has some accuracy limitations for small drivers. This doesn't seem like a good idea for tweeters.

http://www.randteam.de/ABEC3/Index.html

[TN Allen]

#newbie,

Other people will probably write about the software they use, possibly other methods as well. I use Kubotek's KeyCreator, which used to be Cadkey to design the wg contour, then use Key Machinist, which runs insude KC, to generate toolpaths/G-Code to run a CNC machine to mill the guide. Some of the guides are combined tweeter/midrange, which can be complex blending surfaces, and occasionally result in small surface anomolies. I have been looking for software soecifically for more complex surfaces, if anyone has suggestions.

I have also skipped designing wgs, and turned them on a lathe using a jig to turn a uniform arc cross section to a specific radius, tangent to the wg baffle/mouth surface.

[scottq]

Hi TN,

That's what I think of regarding "modeling" too. I would use CATIA for the modeling; it's plenty powerful and can do lots of things, including drawing to a defined (with formula) curve. I'm, at best, a novice user - it's quite expensive, insurmountably so (I have access at work). I've got combo/blended mid-tweeter straight 45 degree waveguide started for one of the speakers I'm conceptualizing. A friend has several programs he's played with for G-code to run his 3D printers, so I could get some advice from him. Too many ideas, lots of time.

I'll be sending you a PM about the CNC...

[Mike220]

Does it matter is its the fabric or aluminum rst-28 with this waveguide btw?

[Pete Schumacher]

Does it matter is its the fabric or aluminum rst-28 with this waveguide btw?

When I first did this kind of thing a while back, I think it was DanP who printed a bunch of different profiles of 2" deep waveguides that were 8.5" diameter circular and fit the RS28 chamfer ring around the dome mesh. The guide that worked the best was basically a circular arc cross section, and both the aluminum and fabric domes worked well. It even worked for the SEAS 27TBFC/G.

I then did a 1" deep version and again, both the fabric and aluminum performed well. But throat geometry is critical to top octave behavior. It needs to match the tweeter to avoid standing wave buildup between the dome and the guide walls.

[xmax]

I am wondering if the RST grill/phase shield combo (in WG) is creating some funkiness above 10K, has anybody successfully removed it?
It is on both the fabric and metal version of the New RST.

[xmax]

PE site appears to be down...

[augerpro]

I am wondering if the RST grill/phase shield combo (in WG) is creating some funkiness above 10K, has anybody successfully removed it?
It is on both the fabric and metal version of the New RST.

I tried prying it off but it seems pretty well attached. There is a huge gap between the phase shield and dome on these new ones. Maybe that is an issue? OTOH with how large the throat must be, I'm not at all surprised by the response wobble. I think the old tweeter working so well on waveguides was a fluke of lucky dimensions. I have a couple waveguides coming so I'll play more with it soon.

[Mike220]

Yes I'm using the rs28a in a seos guide and it really improved the sq on the rs28 in my oppinion. I want to try and recreate my build with the new tweeters. I crossed to a rs270paper 10".

[Erich H]

There is a huge gap between the phase shield and dome on these new ones. Maybe that is an issue? OTOH with how large the throat must be, I'm not at all surprised by the response wobble. I think the old tweeter working so well on waveguides was a fluke of lucky dimensions. I have a couple waveguides coming so I'll play more with it soon.

I agree with how wide the throat has to be on the new RS models causing some issues. On the past model the waveguide intake could be closer to the tweeters surround and I believe the dome section stood up higher off the faceplate. The older one also had an insert around the dome that had a slight upward flair that basically curved right into the waveguide. The new models are truncated and flat. So it's not as smooth of a transition.

[Pete Schumacher]

I tried prying it off but it seems pretty well attached. There is a huge gap between the phase shield and dome on these new ones. Maybe that is an issue? OTOH with how large the throat must be, I'm not at all surprised by the response wobble. I think the old tweeter working so well on waveguides was a fluke of lucky dimensions. I have a couple waveguides coming so I'll play more with it soon.

I don't think it was a fluke at all. If you look closely at the old RS28 flange and dome mesh assembly there is a small ring around the mesh that has a 45 degree chamfer. The opening of the guides I used continued from that chamfer, providing a very smooth transition to the guide walls. Other guides we did for the T-Labs tweeters took a different approach with the guide opening sitting over the dome itself. And the last one was accidentally cut so that the opening we outside the surround, and that one turned out absolutely fantastic. But they were round guides.

I have a feeling that because this elliptical guide has some pretty steep walls along the short axis that if the throat is too far away radially from the dome that cancellation could be happening. So the closer it can be to the dome, the better. The opening in this set of measurements isn't optimal for the RS28 anymore, but it does seem optimal for the Wavecor.

[augerpro]

I don't think it was a fluke at all.

I say fluke because the shape/dimensions that happened to work so well weren't intended to work so well with waveguides. It was luck.

[TN Allen]

I don't have the RST28 to look at, but am curious about the location of the surround. It seems to be glued to the plastic. Milling a guide throat to match the O.D. of the surround looks simple provided the faceplate and mesh can be removed. Is it possible the glue attaching the surround has produced a secondary bond to the faceplate? It appears there may be pins to align the diaphragm, which would be good as the alignment on the RS28 was sloppy once the faceplate was removed.

The photo is the replacement diaphragm for the RST 28, from the PE site.

[augerpro]

Tim something seems odd in that photo, when I took off the faceplate there was still an 1/8" ridge around the surround. You can see it there but it is flush with the rest of the plastic not raised. I assumed it was part of the plastic but maybe it was removable?

[TN Allen]

I am wondering if the RST grill/phase shield combo (in WG) is creating some funkiness above 10K, has anybody successfully removed it?
It is on both the fabric and metal version of the New RST.

How does the grill affect the response, especially the transition along the guide throat? I've wondered about that for some time. Has anyone tested the response with and without?
It seems to me that if the optimal guide throat is as close to the surround as possible to produce a smooth transition, the grill and any steps would detract.

"I say fluke because the shape/dimensions that happened to work so well weren't intended to work so well with waveguides. It was luck."
The RS28 had a somewhat better lower range than many tweeters, which contributes to its compatibility with wgs.

Playing the devil's advocate, most of what I've seen to substantiate any guide working well with a particular tweeter is graphs, which only tell part of the story.