While working on my car stereo, I think I accidentally invented a new type of waveguide.
Here is a rendering of the waveguide. The reason that it has an odd shape is that it's designed to be crammed into a corner. (The corner of my car, to be exact.) Basically the compression driver fires DOWN, and then the waveguide radiates into the space you're listening in.
Here's the 3D printed version of it
Here is the polar response. On axis, 22.5 degrees off axis, and 45 degrees off axis. The entire crossover is a single cap, 1mfd. (The 5dB peak at 1800hz is due to the cap being too small to offset the impedance peak of the compression driver; if you used an active xover or a 2nd order passive you wouldn't get this peak.)
Dare I say it, this is very nice performance.
To put this in perspective, I'm getting polars from this waveguide that are about as good as the JBL econowave guides. Which are very good.
Inspiration came from the lens that Manny LaCarrubba invented for Sausalito Audio Works, as licensed by B&O in all their speakers. What you are looking at is what happens when you use a SAW lens with a waveguide.
I built half a dozen waveguides before I settled on this one, check out my build thread at diyma (28 Weeks Later - Car Audio | DiyMobileAudio.com | Car Stereo Forum) if you're curious about the waveguides that preceded this. There was a LOT of trial and error to get here.
Here is a rendering of the waveguide. The reason that it has an odd shape is that it's designed to be crammed into a corner. (The corner of my car, to be exact.) Basically the compression driver fires DOWN, and then the waveguide radiates into the space you're listening in.
Here's the 3D printed version of it
Here is the polar response. On axis, 22.5 degrees off axis, and 45 degrees off axis. The entire crossover is a single cap, 1mfd. (The 5dB peak at 1800hz is due to the cap being too small to offset the impedance peak of the compression driver; if you used an active xover or a 2nd order passive you wouldn't get this peak.)
Dare I say it, this is very nice performance.
An externally hosted image should be here but it was not working when we last tested it.
To put this in perspective, I'm getting polars from this waveguide that are about as good as the JBL econowave guides. Which are very good.
An externally hosted image should be here but it was not working when we last tested it.
Inspiration came from the lens that Manny LaCarrubba invented for Sausalito Audio Works, as licensed by B&O in all their speakers. What you are looking at is what happens when you use a SAW lens with a waveguide.
I built half a dozen waveguides before I settled on this one, check out my build thread at diyma (28 Weeks Later - Car Audio | DiyMobileAudio.com | Car Stereo Forum) if you're curious about the waveguides that preceded this. There was a LOT of trial and error to get here.
Nice work, & good results
Thank you!
Up until six months ago, I thought the SAW lens was mostly a way to mount a tweeter in a novel way. (90 degrees off axis.)
But you can see from the measurements it does more than that; it really smooths out then entire polar response.
My theory on this is that the lens is concentrating all of the output into a single point, and by virtually 'shrinking' the size of the tweeter, it's flattening out the polars.
I don't think I've ever seen a waveguide where the polars were so eerily consistent, over a wide bandwidth. Basically it sounds the same no matter where you are.
The ideal loudspeaker would cover ten octaves and it would be infinitely small. If there was a speaker that measured 3/4" in diameter and could do 20hz to 20khz, that would be ideal.
Obviously, that type of speaker doesn't exist. Even a 1" tweeter will beam above 13,500 because 13,500hz is smaller than the tweeter diaphragm.
The Sausalito Audio Works lens takes the output of the driver and it focuses it at a single point. So you basically get a polar response that's pretty darn close to that idea. Although our tweeter may be 1" in diameter, it's being "concentrated" into a smaller size. That's why the polars looks so crazy; they look like the polar response of a speaker that's much smaller than 1" in diameter.
I don't think B&O has published polars. But Sausalito published some polars of the prototypes that preceded the Beolabs. In Sausalito's polars you can see similar behavior as mine, basically the response curve is really really tight inside of the waveguide's beamwidth. In the measurement above, you can see that the SPL level is basically the same, whether you're directly in front of the speaker or 30 degrees off axis.
While screwing around with these lenses, I found something interesting. Basically, the lens concentrates the energy of the tweeter. So if you have a 1" tweeter that's flat to 20khz, when you put it on the lens, you wind up with a falling response. (This is because the 1" tweeter is beaming above 13,500khz.) You can see this problem in the measurements from Sausalito. That's why I wound up going with a compression driver. (I am using the same compression driver that you are, the Celestion CDX1-1445
Only if you didn't want directivity (I do). An omnidirectional point source speaker (or even a 180degree speaker) is pretty much the last thing kind of polar response I'd want, it would illuminate all the room boundaries like a light bulb and have vague (though ethereal) imaging. I made a system that approximated that once (not as neatly as your setup, though) but didn't really like the effect much after a week or so. It sounded kind of neat, but in the same way that drive-in movie theaters used to sound if you walked around, with all the speaker sound coming from everywhere and nowhere. Just a matter of tastes, of course. (And you'd have to be somewhat elder to remember drive-ins to follow my babbling).
But I was asking about the waveguide shape/effect. It almost appears that the side walls don't do anything, like the bottom surface is doing it all. Or maybe they will beyond +/-45 degrees?
This lense is more a diffusor than a contractor type of waveguide. Usually we have our thinking locked to horn-like waveguides that do the opposite. The smoothing comes by attenuating "baffle step" and width-related directivity that appear in conventional setups. That's why it is wise to use small compression driver instead of a dome.
Nice 3d printed effort. One of the smoother frequency response curves Inhave seen from you. However, it looks to me, very similar to SAW's lens but with angled side walls. If you looked at the SAW's patent claims, it probably covers sidewalls that are narrower in subtended angle, in which case you have a variant of the SAW lens in "scope of metes and bounds" as the patent language usually says to cover obvious adaptations of its main function.
Its "it's designed to be crammed into a corner" and in a corner is restricted horizontally to 90 degrees. Uniform across those 90 degrees is ideal.
The tighter it fits into the corner the higher in frequency the walls of the room act as extension of the waveguide sides. With the CD firing down instead of out, this goes deeper into the corner than what can be done with a conventional horn or waveguide. Patrick is no doubt targeting a car where height doesn't matter so much. In a home theater, firing up from the floor might not be the best but one can raise it up and then have to deal with floor bounce....
Definitely. I've messed around with Danley Paralines a lot, Sausalito Audio Works lenses quite a bit, and V-DOSC waveguides a little. I've found that the Paralines were the easiest to build, but it's REALLY difficult to get smooth response out of them. Conversely, the SAW lenses are difficult to build, but easy to get good results.
It wasn't until I had a 3D printer that I could start to do some crazy things with the lens. (Side note - I nearly took off my right thumb years ago trying to make the lens out of wood. It's really hard to work with due to it's tiny size.)
Anyways, long story short, I always figured you could put these at the throat of a waveguide or horn, but didn't have a way to get good results until I bought a 3D printer.
Intriguingly, a member of the forum pointed me in the direction of a new patent from the inventor that license the lens to B&O. The new patent is obviously for horns. But there's another interesting thing in there, which is that he's changed the lens shape from elliptical to conical. This might seem like a small change, but if it works, then it means that it could easily be built in wood. Because all you would need is a conical drill bit, which can easily be found at any Home Depot. A conical lens also reduces the height of the waveguide.
Keep up the good work, & post with updates etc
+1 and subbed.
Very interesting stuff which kinda makes the rest of the ICE world look decidedly old hat.
Hmm, kinda, sorta as Altec made an extremely truncated 90 deg Mantaray WG for its A8 VOTT. Unfortunately, this is the only picture I could find in a quick search: http://images.audioasylum.com/usr/y2009/07/26014/untitled.JPG
Regardless, great work! Gotta love all this new and relatively affordable technology. Can it easily make vintage multi-cell horn parts and their complex throat adapters to bond together?
GM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.