3D printed waveguides

[Pete Schumacher]

That's enough for me to figure it out, thanks Pete!

mouth opening is 1.625" wide.

[augerpro]

Wow those are totally different wg's. The 7" has a throat entry angle of only about 32 degrees, while 8" was quite wide, about 112 degrees. You liked the 8" better subjectively?

[TN Allen]

I find ~ the same angles. Were there any positive effects from the 7 X 2 guide, especially while listening?

[Pete Schumacher]

The shallower guide lets the tweeter sound more natural and transparent. I was never really happy with the sound of the deeper guide. It wasn't bad, just not as open as my reference Byzantium which uses the same tweeter.

The deeper guide definitely provided more "gain" but constrained that top octave and the shallow guide still has plenty of gain to offer great stress relief at the 1100Hz XO point. I'd think a happy spot might be 1.5" deep and 8" wide.

[augerpro]

Thanks, working my way through it and trying to find Toole's papers for the definition of the different types of tests (e.g. LW, PIR, etc.)

I thought those were defined, but if not, it is probably in here: https://www.dropbox.com/s/nif78zqk11...ntier.pdf?dl=0

[TN Allen]

Pete,

I think one of the earlier guides DanP made that you used was ~2" deep X 8" dia., how did the sound from it compare to those in your post of yesterday? That is, the 7" X 2", and 8" X 1". I realise there are many variables in any comparison, and listening is subjective, especially over time, nonetheless, in that you test and listen, your conclusions and/or impressions are valued and may help direct the dimensions for the guides to be tested in the future.

[Pete Schumacher]

Pete,

I think one of the earlier guides DanP made that you used was ~2" deep X 8" dia., how did the sound from it compare to those in your post of yesterday? That is, the 7" X 2", and 8" X 1". I realise there are many variables in any comparison, and listening is subjective, especially over time, nonetheless, in that you test and listen, your conclusions and/or impressions are valued and may help direct the dimensions for the guides to be tested in the future.

That was the guide we used in the MiniByzy build (Aurora). It was excellent (check the uniformity of the on-axis and off-axis plots). The only reason I did the 1" deep guide was simplicity of fabrication. To do the one in the picture you have to glue two pieces of 1" thick MDF together and then machine it opening up the possibility of having the seam "pop" like the one on the edge in the picture.





On-axis result




20 - 45 - 60 degrees off axis

[TN Allen]

I added two new guide surfaces for comparison to the previous image. One is for an 8" dia. X 2" deep guide using the 1.625" diameter throat opening, that is with the face plate in place. The second is the same 8" dia. but with the 1.475" throat opening and a 2.133" depth. The 1.475 and additional .133" depth represents the guide you used extended to compensate for the removed face plate. This is close to the guides I've milled for the RS 28 without the face plate.

You'll notice that most of the two surfaces for the 8" X ~2" guides are essentially the same. This is why the odd pattern where the aqua and red surfaces intersect. Were these different surfaces, the surface representations would be single colors, no mixed pattern. It also appears that, given the dimensions I used, the throat opening in the face plate was very close to tangent to the guide surface.

It would be interesting if Brandon were to test guides based upon the 7" X 2", 8" X 2" and 8" X 1" using the RS 28. The data might be quite useful, given these are very different guide surfaces.

[augerpro]

New waveguides in! Also a number of phase shield designs to play with, cuz, what the heck I'm already doing the work. Here is shot of the one with the throat that occludes the surround as Pete suggested:



I'll probably try one even closer to the dome, and maybe one with a slightly higher starting height. If this does indeed work better than phase shields then I might as well play with the variables a bit before I settle on one design. I'm also really happy with the studs I incorporated (pic a few pages back) to center the tweeter to the wg, works awesome!

BTW, if anyone finds this valuable, I would REALLY appreciate a donation to help out. I'm getting these waveguides at a pretty good price, but I'm already $250 in to this so far. And haven't even settled on basic design features so I can move on to other sizes and tweeters. So you can imagine how expensive this might get, and would probably stop progress. Anyway, you can paypal me at whippersnapper02 at yahoo. Thank you!

[augerpro]

Tim> 8" is next after I've nailed down best practices on these 5" wg's. That will be interesting! However I don't know when I'll design for the RS28A, possibly never. After all this is done and I feel I understand the requirements for a good wg, I could easily whip something up for anyone to fabricate an RS28A version on their own time and dime.

[TN Allen]

Brandon,

I have several 8" dia. guides ~2" deep, and would do a few more of varying depth to provide a good range for testing. I can also send you gages to verify cross section radii. Some can be used with different tweeters by turning or milling additional mounting plates. This changes the depth of the guide, but would make comparisions between tweeters somewhat legitimate.

[augerpro]

Sounds good, I'll let you know when I get to that point.

[TN Allen]

Reconsidering the initial idea for this thread, I decided to print a larger guide on a larger better 3D printer to evaluate the practicallity. I had done a small oval guide several years ago using a MakerBot. The photos are of an 8" dia. X 2" deep X 1.3" dia. throat WG printed at low density, meaning it's honeycombed inside. The surface is not as smooth as can be sanded after CNC milling in wood, MDF and various other materials.

In comparison, the time to mill this with setup, would be about 45-60 minutes, printing required over 13 hours. Had the part been high density or solid, the time would have increased considerably.
In addition to the long print time, there was considerable support material to break off, that's trash now.

Having milled and printed guides, my choice is milling. Perhaps someday, probably not soon, inexpensive 3D printed guides will be a practical choice, currently though, the finish and print time(=cost), are impractical.

[Pete Schumacher]

Reconsidering the initial idea for this thread, I decided to print a larger guide on a larger better 3D printer to evaluate the practicallity. I had done a small oval guide several years ago using a MakerBot. The photos are of an 8" dia. X 2" deep X 1.3" dia. throat WG printed at low density, meaning it's honeycombed inside. The surface is not as smooth as can be sanded after CNC milling in wood, MDF and various other materials.

In comparison, the time to mill this with setup, would be about 45-60 minutes, printing required over 13 hours. Had the part been high density or solid, the time would have increased considerably.
In addition to the long print time, there was considerable support material to break off, that's trash now.

Having milled and printed guides, my choice is milling. Perhaps someday, probably not soon, inexpensive 3D printed guides will be a practical choice, currently though, the finish and print time(=cost), are impractical.

Why the support material? Why not print it with the mouth down and the throat at the top?

[TN Allen]

There would still be support. The plastic filament can sag until it has cooled and hardened. This was a problem printing in the MakerBot a few tears ago.

It also helps maintain tighter tolerances. I haven't tried a multi-part print with parts inside of parts yet, or a print with very small intricate sections, but the support material would also be necessary in these, unless the latter were flat and thin, partially to provide clearance, partially for stability.

After printed, some parts require a long heated bath to disolve the support material. One of the sample files is of a detailed dinosaur skull complete with teerh, this would require considerable internal support while printing.

This printer will print quite high as well, I don't recall the Z dimension, but well over 12". It runs quite hot, so the plastic doesn't become stable until removed and cooled.

I considered turning the model over to print, but any dangling ends of hot filament, and the tendency to sag argued against this. As it is, I should have gone with a higher density print, the finish would have been better.

I suppose at some point 3D printing will build "wood" models by placing molecules rather than plastic filament, until then, I prefer the CNC mill, or turning guides.