3D printed waveguides

[Pete Schumacher]

In other words the depth of the guide is the same as the depth of the baffle?

That would be correct.

[DanP]

That would be correct.

So Pete, would you recommend making the WG as shallow as possible once your boost and directivity requirements are met?

Dan

[augerpro]

Dan that's how I'm leaning. We should consider AC depth too though. I did some measurements yesterday of the depth from flange face to spider, which I've found is a good estimate of AC. 3/4" -1" for 12-15cm drivers, 1.25" for 18cm driver, and 1.75" for 22cm driver.

[DanP]

Yes, AC is a good thing to consider as well and it looks as though it will fall into a good range with the rest of the parameters.

Dan

[TN Allen]

Perhaps testing a series of graduated guides varying only the depth and consequently the cross section would be worthwhile. The geometry and tool paths, or the .stl file are easily done. Doing this might reveal some guidelines for optimal depths given specific throat and mouth diameters, or throat diameter and ellipse dimensions. It is easy to scale the solid or surface model to different depths. This changes the cross section from arcs to a more complex curve, but that might not deviate sufficiently from an arc to significantly affect performance. This might also need testing.

[augerpro]

Definitely Tim

[TN Allen]

Once the printing routine is worked out, so we know for certain what file format and scale is required, I'll do a quick series of designs with a uniform mounting configuration.

[Pete Schumacher]

Dan that's how I'm leaning. We should consider AC depth too though. I did some measurements yesterday of the depth from flange face to spider, which I've found is a good estimate of AC. 3/4" -1" for 12-15cm drivers, 1.25" for 18cm driver, and 1.75" for 22cm driver.

👍

That's how I am leaning these days.

[TN Allen]

Assuming 3D printing waveguides proves practical and not too expensive, and that the surface printed is acceptable, what would be a useful range of dimensions for a comprehensive evaluation of waveguide effects? If the guides are round, it is easy to use a given guide with various tweeters mounted on adapter plates that maintain a uniform guide contour to the surround. This is not possible with oval guides of course because the contour transition from ellipse to circular changes with depth.

I haven't seen a comprehensive, methodical study of waveguides that would provide the quantitative data that would make decisions regarding acoustic response a little more objective.

Depending upon the dimensions chosen, some of the test guides might too large to be printed practically, and might consequently be milled, but that aside, it would perhaps be interesting to explore ideas about a useful range of test guides. It might also be worthwhile long term, to think of ways to relate data from round test guides to future data from a similar test project for oval/elliptical guides, and guides for ribbon tweeters.

[augerpro]

When you say range of guides or range of dimensions, what do you mean exactly? Do you mean a series of guides where only one variable is changed (say diameter) to compare the effect? Or a series of guides for different applications? I intend to investigate both. Any promising guide should be compared to a variant that is slightly wider, or deeper, etc. And as far as a range of sizes, I'm thinking more a range of defined applications. For example we defined this first as being mated to a 12-15cm driver with certain characteristics. Next we should do the same for a 18cm and then 22cm driver too.

For quantitative data what did you have in mind? If there is an Excel expert out there I could supply the SPL in a .txt file and they could calculate a Directivity Index and plot like you find with JBL's datasheets. Is that the sort of thing you mean?

I'm going to check out printers today.

[augerpro]

Which printing process/machine should I specify: http://redrockspd.com/3d-printing-services/

[augerpro]

So uploading your files Tim to their site shows absurdly large XYZ dimensions, about 146" x 108" x 26", so too large to get a quote.

[TN Allen]

Brandon, The file is metric, but they read it in inches. 146mm/25.4" & etc. yields the actual size. I spoke with Matt Filkin at Redrock. They can work in either metric or English units. They evidently read the file, that was the initial objective. If you want that file printed, either they can read it in in metric units, or I can convert it to inches for you. However, I suggest developing a new file that addresses some of the questions expressed on the forum. I can send you the guide I have, why duplicate it. Now that we know they can read the file, and a file in inch units is no problem, what will the first guide be? Sent from my LG-V495 using Tapatalk

[TN Allen]

When you say range of guides or range of dimensions, what do you mean exactly? Do you mean a series of guides where only one variable is changed (say diameter) to compare the effect? Or a series of guides for different applications? I intend to investigate both. Any promising guide should be compared to a variant that is slightly wider, or deeper, etc. And as far as a range of sizes, I'm thinking more a range of defined applications. For example we defined this first as being mated to a 12-15cm driver with certain characteristics. Next we should do the same for a 18cm and then 22cm driver too.

For quantitative data what did you have in mind? If there is an Excel expert out there I could supply the SPL in a .txt file and they could calculate a Directivity Index and plot like you find with JBL's datasheets. Is that the sort of thing you mean?

I'm going to check out printers today.

I worded my suggestion vaguely on purpose, dimensions, methods and objectives probably should be decided by the various Forum members discussing the possibility of printing WGs. Presumably they would be the users of the data, and have a good sense of what would be useful. I may be wrong, but it seems to me much of waveguide design is based upon vaguely substantiated ideas, rather than testing and some uniform method of recorded subjective evaluation using an incremental series of guides, for example:

3-16" diameters in 1" increments X 1-6" depths in 1/4" increments using diverse drivers. Perhaps the data could be compiled in a matrix that could be referenced for specific application objectives. I realize that is quite a few guides, and likely quite a few drivers. But if the guides are round, it is easy to produce suitable adapter plates for the drivers that would ensure a smooth uniform throat transition.

Excel could work nicely, documenting guide dimensions, driver specifications and acoustic effect. Lots of work! Could be fun though.

[augerpro]

I'm thinking start with what we've discussed so far of the first design: 5" x 1" deep, with whatever constant radius is required, matching to whichever tweeter you fancy. If it is not the RS28A we'll need to include a phase shield. My test baffle will use a 9" x 9" square insert, 1/2" thick (I could maybe do 1/8" thick and then use a 3/8" spacer if that would be cheaper and less time consuming). A screw hole can be placed in the corner 3/4" from each edge, or 1.06" in from the corner, depending on how you prefer to look at it.

I'm going to run to Redrock right now and talk to them in person. Fedex and UPS stores near me didn't have 3D printing.