Waveguide profile ideas for group experiment

[DanP]

As discussed in another thread, Lunchmoney has vounteered to do some CAD work on some different waveguide shapes - I'm going to make them out of glass-filled nylon via a laser sintering process, and Pete is going to throw some tweeters in them and measure them. So what profiles do we want to see?

There are no restrictions on shape - this process is capable of things not possible with conventional manufacturing techniques. The only restriction is size, and that is mostly an issue of time - I will be able to build smaller ones (6" dia. or less) very readily, but larger ones may take a while to be able to fit in the queue without costing my company too much money. Also, could we incorporate any geometry to minimize throat reflections or any other problems inherent with waveguides? I'm not knowledgable enough on this subject to get very creative with the ideas, but ANYTHING GOES with regard to geometry. If you can think it and Lunch can draw it, I can make it - slits, slots, bumps, fins, waves, etc., etc.

I'll kick things off:

I'd like to see a 104mm diameter waveguide with a 28mm throat and a 20mm depth and another with a 25mm depth to allow me to align the acoustic centers of my woofer and tweeter without comprimising the CTC spacing. Like I said, I don't know what the different profiles do, but I'd like to see the results of a few different ones. By the way, let's not all waste our time. If there is a reason this waveguide will be worthless, let me know and we'll skip it.

Lunch - can you whip up a a couple different profiles each for the 2 different descriptions above? Curves could be whatever you think would be interesting or usable.

Pete - what does your tweeter collection look like? Should we stick with a standard 28mm throat diameter or are there some smaller or larger sizes you'd like to experiment with as well? We'll probably start with just single waveguides, but if we stumble on a few profiles that the measurements look great on, I'll get you a duplicate so you can give us a little subjective description with a stereo pair.

Dan

[Pete Schumacher]

Re: Waveguide profile ideas for group experiment

As discussed in another thread, Lunchmoney has vounteered to do some CAD work on some different waveguide shapes - I'm going to make them out of glass-filled nylon via a laser sintering process, and Pete is going to throw some tweeters in them and measure them. So what profiles do we want to see?

There are no restrictions on shape - this process is capable of things not possible with conventional manufacturing techniques. The only restriction is size, and that is mostly an issue of time - I will be able to build smaller ones (6" dia. or less) very readily, but larger ones may take a while to be able to fit in the queue without costing my company too much money. Also, could we incorporate any geometry to minimize throat reflections or any other problems inherent with waveguides? I'm not knowledgable enough on this subject to get very creative with the ideas, but ANYTHING GOES with regard to geometry. If you can think it and Lunch can draw it, I can make it - slits, slots, bumps, fins, waves, etc., etc.

I'll kick things off:

I'd like to see a 104mm diameter waveguide with a 28mm throat and a 20mm depth and another with a 25mm depth to allow me to align the acoustic centers of my woofer and tweeter without comprimising the CTC spacing. Like I said, I don't know what the different profiles do, but I'd like to see the results of a few different ones. By the way, let's not all waste our time. If there is a reason this waveguide will be worthless, let me know and we'll skip it.

Lunch - can you whip up a a couple different profiles each for the 2 different descriptions above? Curves could be whatever you think would be interesting or usable.

Pete - what does your tweeter collection look like? Should we stick with a standard 28mm throat diameter or are there some smaller or larger sizes you'd like to experiment with as well? We'll probably start with just single waveguides, but if we stumble on a few profiles that the measurements look great on, I'll get you a duplicate so you can give us a little subjective description with a stereo pair.

Dan

Personally, I wouldn't go with less than a 6" diameter waveguide, like the one in the XT19 thread. Otherwise, the bump in response is too far up in frequency to facilitate a 2KHz XO point without a bunch of EQ. That's fine for something like a 20mm tweeter designed to cross at 3KHz, but for 28mm domes, you don't want something smaller than 7", which matches up well with a 7" woofer.

I wouldn't worry about throat diameter at this point, as long as it allows the 28mm domes and surround to fit inside. We could try smaller tweeters in there and see what the wider throat does.

One caveat to the CAD creator . . . no sharp transitions. smooth curves from the throat, to guide, to baffle. And keep in mind that too deep a throat profile will reduce dispersion.

[lunchmoney]

Re: Waveguide profile ideas for group experiment

Larger ones could easily be built in sections and glued/screwed/etc together.

I'll whip something up in the next couple of days. Won't take long, I'm just slammed for the moment.

Happy to help out. This should be fun.

[jim85iroc]

Re: Waveguide profile ideas for group experiment

I'd like to see some irregular shapes to see how they can control directivity. Perhaps some oval shapes, or rectangular.

[andy19191]

Re: Waveguide profile ideas for group experiment

Very interesting. Dan, is something like this doable (pic on page 3):

http://www.genelec.com/documents/publications/G0001.pdf

I am developing some 3D aeroacoustic software and was considering looking at mid/tweeter waveguide integration and cone/waveguide integration as possible example simulations to help with the development.

Related to this, is anyone aware of any available CAD drawings of drivers?

[lunchmoney]

Re: Waveguide profile ideas for group experiment

Very interesting. Dan, is something like this doable (pic on page 3):

Easy.

[andy19191]

Re: Waveguide profile ideas for group experiment

Easy.

Good.

What file formats are required? STL is easy for me.

[Mayhem13]

Re: Waveguide profile ideas for group experiment

And keep in mind that too deep a throat profile will reduce dispersion.

I thought the 'constant' directivity was the goal over the frequency range used by the device. It's gotta hold pattern control as low as possible. The shallow MCM doesn't do that very well to 2khz. But then there's the dreaded 'HOMs' with deepening WG profiles and the need for non-descript foam.....and around and around it goes.

I've used the MCM in several designs with the 27TDFC, SB29, DX25 and while i like them all, i haven't fallen in love with any of them. The surreal imaging is fun at first, but then gets kinda annoying after a few listening sessions. The deeper and much better directively speaking XT120's from 18Sound i'm working with do sound much better in that regard, but the ragged top end is also quite fatiguing...to my ears anyways.

Honesty, my favorite iteration whether it's accepted or not is a wide baffle flush mounted dome AKA Troel's Poor Man's Straads. Silky smooth performance with incredible imaging and directivity. I've got no measurements to support this of course, but it's the direction i'm heading myself just the same. maybe a shallow but large WG profile milled into the baffle.....who knows.

[DanP]

Re: Waveguide profile ideas for group experiment

Very interesting. Dan, is something like this doable (pic on page 3):

http://www.genelec.com/documents/publications/G0001.pdf

I am developing some 3D aeroacoustic software and was considering looking at mid/tweeter waveguide integration and cone/waveguide integration as possible example simulations to help with the development.

Related to this, is anyone aware of any available CAD drawings of drivers?

Like Lunch said, that would be a piece of cake. Is this something that you'd be testing yourself and posting results for the DIY community? I'd prefer not to build these otherwise extemely expensive parts for individual (especially commercial) ventures. Like I said, the owner of the company is very generous and doesn't mind me making personal items, but I always use discretion when doing so.

For what it's worth, STL is the format I use to build the parts. I didn't mention it because most people don't use that format.

Dan

[Pete Schumacher]

Re: Waveguide profile ideas for group experiment

I thought the 'constant' directivity was the goal over the frequency range used by the device. It's gotta hold pattern control as low as possible. The shallow MCM doesn't do that very well to 2khz. But then there's the dreaded 'HOMs' with deepening WG profiles and the need for non-descript foam.....and around and around it goes.

. . . maybe a shallow but large WG profile milled into the baffle.....who knows.

It's a matter of how far off axis can you still "see" the dome at the bottom. You can pick whatever angle you want, with 45 deg horizontal being the minimum I'd consider. But if you wait too long to begin the flare of the throat, you'll reduce that dispersion even more.

You're really trying to get the off axis performance of the WG/tweeter to approximate the off axis performance of the woofer in the crossover region, so that as you move off axis, response falls uniformly across the band, maintaining as flat a response as possible as far off axis as possible.

I would think the eventual result will be to have the flare machined right into the baffle, or at least have the outer edge tangential with the baffle.

[andy19191]

Re: Waveguide profile ideas for group experiment

Like Lunch said, that would be a piece of cake.

That is good to hear. How well does it hold it's shape?

Is this something that you'd be testing yourself and posting results for the DIY community?

My primary interest is the software and demonstrating how it can address engineering problems. The point of the exercise is to make the results available to interested people like DIYers. Ideally others within the project would build and test some of the ideas/designs. Perhaps driving the design process around their available hardware.

For what it's worth, STL is the format I use to build the parts. I didn't mention it because most people don't use that format.

Good.

[timw]

Re: Waveguide profile ideas for group experiment

If you are designing a tweeter waveguide from scratch, I would prefer a waveguide that allows a 1.0 to 1.3kHz crossover. This is in an effort to reduce the sound level being supplied by the woofer/midrange in the 1.0 to 2.0 kHz range, as numerous quality woofer/midranges have peaking 3rd and 5th harmonic distortion in the 1.0 kHz to 2.0kHz band. Check Zaph's charts.

[Mayhem13]

Re: Waveguide profile ideas for group experiment

If you are designing a tweeter waveguide from scratch, I would prefer a waveguide that allows a 1.0 to 1.3kHz crossover.

That would and does require quite a large WG, and if round would set the C to C spacing outside of acceptable XO limits....and there's not too many domes that can play that low with any real authority,and none that can do it at reference or HT levels. You'll need a compression driver for that. The subjects been beaten to the inevitable before.

[lunchmoney]

Re: Waveguide profile ideas for group experiment

That is good to hear. How well does it hold it's shape?

SLS is incredibly strong and stable. It's quite literally solid nylon, and can be made very thick if required. Way more solid than mdf, that's for sure.

You can drill it, sand it, machine it, whatever.

Definitely more than useable for this.

Polyjet or SLA would be better, because it has higher resolution. But I'm overstating this. SLS will be just fine.

[rpb]

Re: Waveguide profile ideas for group experiment

I'd like to see a 104mm diameter waveguide with a 28mm throat and a 20mm depth and another with a 25mm depth to allow me to align the acoustic centers of my woofer and tweeter without comprimising the CTC spacing. Like I said, I don't know what the different profiles do, but I'd like to see the results of a few different ones. By the way, let's not all waste our time. If there is a reason this waveguide will be worthless, let me know and we'll skip it.

Have you considered this tweeter?

SEAS Prestige 27TBCD/GB-DXT (H1499) Tweeter