First impressions of the LX521 pretty much always include some comment about the baffle . . . usually its aesthetics. SL says that the shape was arrived at empirically, but clearly there is some “thinking” which justifies the departure from the more “normal” straight sides with steps, and many “empirical” solutions can be incrementally improved if there is a theoretical understanding of them, and why they work. That it does work (in the context of what a dipole baffle is supposed to do) I can attest from listening to it . . . the dipole null is pronounced and broadband, and extends vertically to the extent that I could determine, at least 45 degrees above the horizontal. The c-to-c spacing of the combined mids should lead to some dipole-like vertical beaming around the crossover frequency, and the undercut below the lower mid may extend vertical dipole behavior as well. But is it “optimum”? Is it possible to do better?
I intend to start with the basic LX521 baffle shape for my “clone” experiments (without the extension for the rear tweeter, which to my thinking could go anywhere), but I’m altogether open to suggestions for “improvements”. Any ideas ? ? ?
I intend to start with the basic LX521 baffle shape for my “clone” experiments (without the extension for the rear tweeter, which to my thinking could go anywhere), but I’m altogether open to suggestions for “improvements”. Any ideas ? ? ?
Years ago I thought about making a 4-way variation on Orion by adding a small upper midrange, and I always thought that would suggest bumping the 22cm midwoofer to a 26cm. The new LX521 baffle leaves the midbass driver pretty close to nude and its output will be limited toward the bottom if its range. A hybrid of Orion / LX521 / NaO Note II would have an Excel W26 midbass and Discovery 10F midrange. This would require a steep low-pass on the W26 which you could do with MiniDSP.
Why the 10f? The vifa tc9 is 1/10th the price and a better performer.
In fact, I would just use that for the top and forget about the dome tweeters!
I would put an acoustic reflector lens on the rear of the vifa to diffuse rear HF.
Last edited:
Hi,
I have compared 2 variants of the LX521 lower baffle shape (eg. without tweeter part) to a rectangular baffle. The rectangular baffle is 256mm wide, the LX521 variants at the widest point 315 and 331mm. The measurements I made are a bit messy but if there is any interest it could upload them somewhere.
Regards
I have compared 2 variants of the LX521 lower baffle shape (eg. without tweeter part) to a rectangular baffle. The rectangular baffle is 256mm wide, the LX521 variants at the widest point 315 and 331mm. The measurements I made are a bit messy but if there is any interest it could upload them somewhere.
Regards
The lower part of the midrange baffle has to work up to 1 kHz. That's a wavelength of 34 cm. I don't see how the specific shape of the baffle in that region would make a real acoustical difference compared to vertical sides.
Of cause it does have some aesthetical and marketing advantages over the rectangular baffle.
The part around the tweeters isn't of too much acoustical concern either - regarding the crossover at 7 kHz. The tweeters already start to beam, which will control the radiation pattern increasingly more than the dipole null. If you would cut off the "excess" baffle around the tweeters (as indicated in the picture), I don't expect a tremendous difference - but would surely be interested whether I can be proofed wrong.
Now if you look at the "corrected" (red) baffle shape in my picture, it should strongly remind you of some other dipole 4-way speaker.
I can't help but find the LX521 baffle as it is now more of a "statement". Its shape isn't dull, it calls for discussion and identification, it has a strong resemblance of "I am a piece of art on a pedestal".I like it.
Rudolf
Wouldn't it be nice if reality weren't like that . . .
Did you get any sense of "difference", or was it "lost in the noise"?
That was my first reaction as well, and I commented to that effect. But . . . (and despite his occasional flair for "marketing"
) I've learned a lot by looking for the "reasons" SL does the things he does . . .I asked specifically about the "flair" at the top . . . the tweeters should be beaming enough for it not to matter . . . but he said that he observed interaction, and there was enough of it to make it worthwhile. I'm not going to co-locate the rear tweeter like that anyway, but still wonder about the overall effect of the "bigger" baffle.
The implication of the "modeling" suggests that on average the difference isn't large enough to matter, but the models don't account for specific axial behavior (so far as I can see). A "rectangular" baffle may behave like a circular baffle of some specific dimension overall but differ significantly depending on whether you are looking across the short dimension or the long dimension (especially significant where the effective baffle may be extended by the bass box below it). Is the gap between the LX521 baffle and the structure it sits on large enough to divorce it from influence or does the lower enclosure act as a baffle extension?
Are there issues here that are lost in the model's simplicity that nevertheless have significance in the "real world"?
Oracular pronouncements can be amusing
but are not particularly helpful . . . some explanation (and examples) of the testing that shows the "inferiority" of the Vifa driver might be more generally informative. Not asking for a full test suite like John Krutke does, or burst tests like SL often posts, but SOMETHING! more than CAPS and EXCLAMATION! . . . The difference is very small if any. You can (with a bit of optimism
) conclude that the lx521 style baffle has a slightly better polar response and that the smaller lx521 baffle has a higher dipole peak. But the difference is, as it should be at max 1KHz frequency, pretty small.I suspect most if not all of the mysteries lie either in the tweeter section, which I have omitted because I plan to use Neo3s, or in the vertical response (of all drivers) which I have not yet measured.
I'll upload the measurements later and perhaps I'll manage to do some vertical measurements this weekend.
Regards
Edit:
I have some pictures with me I can share now, all measurements are @1, without dipole compensation and gated @12ms, driver is a scan speak discovery 22w:
Rectangular baffle of 286mm width:
LX521 shape(ish) 331mm width:
LX521 shape(ish) 315mm width:
So basically the blooming at 600Hz is less pronounced and the increase in directivity near the dipole peak is less abrupt, but if you cross at 1Khz it shouldn't matter too much.
I'll post the specific dimensions of the baffles (angles and height) later.
Last edited:
Thanks for the plots, and yes, the difference is . . . subtle. But there is at least a hint that "the shape" is doing something better than (or at least different from) the bland rectangleThe difference is very small if any. You can (with a bit of optimism
.I'll be particularly interested to see if there are more notable differences on the vertical axis (where the baffle differences would appear to be greater).
If you increase the baffle size, you trade radiation pattern "quality" at the upper end of the pass band against load capacity at the lower end. This is true for the monopole tweeter case AND the dipole tweeter case. With the Orions SL always opted for improved load capacity. For the LK521 too?
Rudolf
Yes, but with ORION that was because he was trying to get the most out of the drivers in a 3-way. Hard to imagine there being any load related problem with the LX521 tweeters crossed that high . . . it must have something to do with pattern control (or "cosmetics"
).Oracular pronouncements can be amusing
It was as useful as the quote I was responding to.
By the way, baffle size and shape has little to do with the response in the 6dB roll off region other than to set where the peak occurs. What it does have to do with is what happens in the frequency region around the peak and what happens as the driver directionality takes over.
... there is at least a hint that "the shape" is doing something better than (or at least different from) the bland rectangle
I made the same conclusion, I can use the LX521 baffle shape and pretend that it's better than a simple rectangle
.Guys, I wouldn't waste a lot of time over analyzing this. Deward already mentioned that the LX521 baffle was arrived at empirically. I.e. SL cut a baffle, tested it and it seemed to work. Maybe he cut one baffle, maybe 20. My experience is that it isn't that hard.
Look at the lowewr mid polars for my system. That withthe side panels and all. I don't see any 600 Hz blooming. Looks pretty good to above the 1k crossover point.
Look at the lowewr mid polars for my system. That withthe side panels and all. I don't see any 600 Hz blooming. Looks pretty good to above the 1k crossover point.
An externally hosted image should be here but it was not working when we last tested it.
Oracular pronouncements can be amusing
He was just doing his Borat impression
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.