audio-talk

I've been thinking about this a bit, but i think my questions would be broader.
What is unsatisfying about your current bass solution, can you put it into words and in any detail ?
Are you sure this new solution will cure the problems ?
Do you definitely like the drivers, have you listened to them already in some way which makes you confident ?

The current issue I have is positioning in the room. I am using a clone of the LF unit from the linkwitz system.
[attachment =0]B4A14559-B655-47A4-A968-E65DE00C4693.jpeg[/attachment]
I like the sound of these subs, they integrate well with AMT’s that I am using. The challenge is positioning. The LX 521’s need to be given space, need at least a 1m from boundaries. I have tried them in the corners and as they are dipole they seem to go muddy.
I only have a relatively small listening room. 5x4m, so don’t really want the speakers well into the room, as I need at least 2m preferably 3m for the midrange and HF to integrate well at the listening position.
So my mission is to repurpose the seas subwoofer drivers from the LX521 bass units into something that can go closer to a boundary. The drivers don’t work well in sealed cabinets, hence why I am looking at a cardioid aperiodic design. Been taking inspiration from this page.https://web.archive.org/web/20210122211 ... nssikotelo

Thanks. That's a very interesting link. Where ( freq ) is the crossover from the bass unit to the OB lower-mid ?

The crossover is somewhere a round a 100 hz, in the LX521. I cross over around 130hz to take advantage of a natural roll off of the mid bass driver

After some thoughts about the 2nd chamber with the rear vented “leaky” cabinet, converting it to masses and springs. How I see it, and please someone correct me. A traditional leaky cabinet will have its own natural frequency Fb with said driver, slightly lower than a sealed cabinet of the same volume, dependant on leakage QI. The second chamber, the Helmholtz resonator, is tuned to reduce the impedance response at Fb. So it doesn’t effect the cardioid response. Whether it offers much benefit over just a larger cabinet, and an electrical notch filter we will have to see.

So sensible approach would be to build test cabinet, blank off second chamber, measure response, and then tune the resonator with the mouth dimensions, to the test data.

It’s coming to the time of year where the speaker “skunk works” opens for the winter.

Amongst many ideas to work, I want to use these Beyma 12p80nd drivers I have had for sometime( doubled in price since I got them). Challenge has been beaming, it’s physics - hard to design around. Saw a thread on Diyaudio, which may helps solve the challenge. Not sure what you call it, a wave guide?? similar to the manger surrounds maybe?? Anyone got any design rules for these things, or is it test and modify approach. I know KEF used something similar, and others Ave used the reverse to support piezo tweeters above the dust cap.

Attachment for above....

Variation of slot loading (which is itself a simple form of waveguide), with the shape adjusted to shape the polars. Strictly speaking, it turns it into a variation of the 6th order bandpass, just as the old Karlson designs were, although unless you go OTT it shouldn't affect tuning to a huge extent.

It’s the directionality is the only thing I want to change. In simple turns, halving the diameter in the orifice in front of the driver should double the beaming frequency. At the expense of what IM distortion?? A search of the web came up with lots of weird and wonderful shapes, but not a lot of science.

You can do a finite or boundary element model easily enough if you have access to the software. However, the lack of 'science' is largely because there isn't necessarily perfect consistency from one design to another. Your initial polars depend on the behaviour of the driver itself & its cone geometry, and most models of this sort of cavity / slot loading will assume a flat piston as the source with falling accuracy as you move toward the point-source limit ultimately defined by coil diameter, as the actual cone behaviour affects the real-world response to some extent or another. The flat plane / 1d model is reasonable as a starting point though so that's pretty much the way to go unless you prefer ye traditional cut & shut. As for distortion -totally dependent upon the driver, its cone, suspension, motor, then the slot profile, compression ratio etc. so there's little way of actually predicting this in advance.

Very similar thinking to my speakers there. My advice it to try it, but ive found anything in front of the driver needs at least a couple of layers of car carpet on the speaker side , and one on the outside will help with the coloration that's invariably produced.

Yeah, you need to watch out for cavity resonance, especially if there's a bit of a gap between baffle & driver.

... Although if you plan things just right, you might find the cavity resonance can help with the crossover.

Chris

I am hoping not to encounter any issues with the 12p80nd as a midrange driver, from a resonance perspective. I had in my mind something like the 9 starred arrangement of manger audio. https://6moons.com/audioreviews/manger2/2.html Which they only claim reduction of edge deflections from. Looks like I will have to generate a few different arrangements and the rope in Chris to get some meaningful measurements for the options. Not an issue to cnc some shapes, lots of perspex appearing as Covid screens are removed.

Generating something irregular isn’t easy..... but here’s 1st attempt, not equally loading the cone is my main concern, if too asymmetrical which this shape is.