Calling all waveguide supergeeks!

[br85]

Alright, I've decided to build a waveguide-based design, from the ground up, and I am in need of some guidance (no pun intended).

The goal is to achieve a low acoustic crossover point (1khz-1.5khz) on a tweeter that would not normally withstand it, the SS R3004 illuminator ring radiator. I deliberately wanted something with an ultra high top end but poor dispersion in the top octave. I also wanted to avoid waveguide loading a conventional dome, and I didn't want to use a compression driver (both for reasons I don't want to go into just now).

I would like to know if there's ANY way to "model" the shaping of a driver's frequency response (on axis is all I need but polar would be handy too) in a waveguide of known dimensions (depth, radius to roundover, roundover radius, throat termination diameter and waveguide geometry as known parameters), or some pointers as to how I can obtain the sort of data I'm after without building hundreds of waveguides?

Also, does anyone know of any published or third party THD graphs of the R3004's? I can't find them ANYWHERE, and I'm really interested in seeing if the sub-2khz distortion problem is congruent to the XT25's on a flat baffle.

Since more information is generally asked for, I can't say exactly where I plan to cross the tweeters yet, since I'm doing an MTM (seas W18s, if you must know), and the waveguide's diameter (and therefore the CTC distance) will determine the final crossover points and slopes.

I'll be using some pretty hardcore dsp crossovers with shaping networks, so any passive crossover complications are not an issue.

ANY help whatsoever in the field of modelling waveguide responses or comprehensive data on the R3004 would be incredibly helpful. Thanks,

Jeff.

[doug20]

You should post the measurement question on HTGuide.com. They might be able to help you.

What DSP will you be using?

[br85]

With any luck the guy I'm building them for will be getting a DEQX HDP-3, so I'll use that once they're in his studio.

Until then, I'll probably do all of the preliminary testing with one of the poverty pack Behringer dsp's. At first, I'm not particularly interested in a nice flat response, the priority is to push the R3004's as low and loud as possible. If distortion figures and my ears tell me that, in a waveguide, they can hit above 100db @ 1.5khz I'll be happy. This would potentially solve the main problem with seas W18's as well.

[Paul W]

You might try AxiDriver for WG modeling.

[ZilchLab]

Practical DIY Waveguides - Part 1

[br85]

Thanks guys, I have read that article before, and I will be downloading Axidriver next time I'm at a PC. The only issue I had with the article presented is that it deals with conical and composite conical waveguides only. I am definitely interest in using spherical or OS waveguides and I'm going to have to figure out a way to plot and model them.

I might even do something like modify a simply MCM waveguide - just to get a bit more depth happening.

[br85]

has anyone with MTM + waveguide building experience got anything to note before I dive into this?

[Paul W]

If you want to model only Spherical and OS, HornResp will do those...though I do not know how your chosen driver will react in those those flares. As far as I know, only AD lets you model custom flares and may let you get fairly close to the ring radiator geometry. Of course, with the additional flexibility, AD is a bit tricky to use.

I've built an MTM with horns (W26FX/ND1460A/XT1464)...no unusual tricks or hints.
Here's a link including some of the process HTGuide Forum - The Raptor ... a 10" MTM

[br85]

Another query (probably not too related to the thread topic) for people who may know, is:

Assuming LR24 slopes, how low would I need to cross the W18's to prevent the need for a notch filter in the 4-5k breakup? From what I understand, it's still somewhat audible at 1.6k/LR4. Would 1.2k make a big enough difference?

[Loren42]

What's wrong with using Geddes' SPEAK software?