Perry,
From your other thread where you show some interesting work on passive XOs for OBs.
From the first post:
Hard to reconcile that. I had to work really hard to make the 15a even listenable. I know they are often used in OBs because they are cheap and have a high Qt
Ever since i saw the lovey OBs in the first post i have been wondering: How well does the 15a do trying reaching up to over 1k?
From my personal experience, to put it mildly, it is lacking.
Choosomg a really good tweeter and using it with a not-so-goosd woofer i find strange.
dave
From your other thread where you show some interesting work on passive XOs for OBs.
From the first post:
Hard to reconcile that. I had to work really hard to make the 15a even listenable. I know they are often used in OBs because they are cheap and have a high Qt
Ever since i saw the lovey OBs in the first post i have been wondering: How well does the 15a do trying reaching up to over 1k?
From my personal experience, to put it mildly, it is lacking.
Choosomg a really good tweeter and using it with a not-so-goosd woofer i find strange.
dave
Funny you ask, because I admit it's not a stellar driver. I had a couple of extra on hand so I used it. It does have the virtue of having a Q of 1.26 which is desirable. Its 12dB peak at 1900Hz is a pain to deal with.
If you take a close look at the design, the crossover frequency is 1250. I use a pretty complex crossover network to tame it, and it blends very nicely with the tweeter.
Above: Nearfield measurements of both Alpha 15A and TW29R.
I'd say it's sort of like the KEF B110 in the LS3/5A: It takes a beast of a crossover to make it sound great, but once you do all that, it does sound great. It's the 2nd time I've used the Eminence 15 BTW.
Once tamed it performs very nicely. The SB Acoustics 15OB350 is a great choice as well. I discuss it here: https://www.diyaudio.com/community/...db-w-passive-xover-no-dsp.404171/post-7479701
If you take a close look at the design, the crossover frequency is 1250. I use a pretty complex crossover network to tame it, and it blends very nicely with the tweeter.
Above: Nearfield measurements of both Alpha 15A and TW29R.
I'd say it's sort of like the KEF B110 in the LS3/5A: It takes a beast of a crossover to make it sound great, but once you do all that, it does sound great. It's the 2nd time I've used the Eminence 15 BTW.
Once tamed it performs very nicely. The SB Acoustics 15OB350 is a great choice as well. I discuss it here: https://www.diyaudio.com/community/...db-w-passive-xover-no-dsp.404171/post-7479701
Last edited:
The nice FR tells us little of the sonics.
And i guess for someone to appreciate the LS3/5A comment, one has to actually think they sound good. They sound OK. I had a love/hat erelationship with them. The last pair i heard, i restored a buddies pair (he compared to a set of not buggered Rogers and they sounded teh same so i musty have done OK. I was very disappointed with the sound, particualrily how boxy sounding they were (I blame the BBC receipe there, the big Harbeths [30.2] have the same kind of box issues).
dave
And i guess for someone to appreciate the LS3/5A comment, one has to actually think they sound good. They sound OK. I had a love/hat erelationship with them. The last pair i heard, i restored a buddies pair (he compared to a set of not buggered Rogers and they sounded teh same so i musty have done OK. I was very disappointed with the sound, particualrily how boxy sounding they were (I blame the BBC receipe there, the big Harbeths [30.2] have the same kind of box issues).
dave
SB Acoustics 15OB350 15” Open Baffle woofer. Has very clean response to 2k.
https://www.diyaudio.com/community/...db-w-passive-xover-no-dsp.404171/post-7479701
https://www.diyaudio.com/community/...db-w-passive-xover-no-dsp.404171/post-7479701
Despite the limitations of the Eminence woofer, these speakers really do sound excellent. My reference for comparison is the Live Edge Dipoles which won 1st place on Parts Express Open Unlimited in the International Speaker Design Competition in August; and the Bitches Brews.
To me half the fun of speaker design is working within a constraint. When I designed OEM car speakers at Jensen in the 90s for Acura, Ford, Honda, Mazda and Chrysler, the most budget I ever had for any speaker was $6 per driver. Considering space and budget limitations, some of those systems were really good.
The idea behind the Walnut Dipoles was: "What dipole design can I execute that has the huge sound stage and imaging of my active designs, including Constant Directivity, but without the extravagant expense of Beryllium tweeters and the like, using ONLY a passive crossover to achieve strong bass down to 40Hz?"
I had a slab of walnut in the garage. I had the Eminence Alpha 15A woofers, one of which I won as a door prize at the 2022 Parts Express Audiofest. I did really like the 15As in a previous system. I also had the SB Acoustics TW29R tweeters from a previous project and I wanted to use an off the shelf waveguide, instead of a 3D printed custom waveguide.
So those were my constraints.
Biggest challenge was the nasty peak of the Alpha 15a. These are my measurements in the actual baffle with no xover components. (The 175Hz peak is a measurement artifact BTW). As you can see the bass without the Passive Bass Boost is a little anemic as well.
The question is, how do you tame the peak at 1900Hz? Most designers will just use a steep 3rd or 4th order filter. I avoid that wherever possible. One of my beliefs is, "don't use brute force when finesse will do" and "don't use nuclear weapons when you can use a bow and arrow."
I combined a 1st order high pass (which is bypassed by a resistor!) with a notch filter at 1900Hz to essentially make a high Q shelf filter.
This is the drive signal received by the woofer:
I've carved out a 16dB notch, but as you see as you go towards 20KHz the level returns to 0dB. So Acoustically it's still a 12dB/octave filter with no phase rotations. Subjectively I think this approach sounds much better than a brute force approach.
Above shows the measured response of the woofer with crossover and Marshall Bass Boost in place. The 1900Hz peak has been completely tamed with no more coercion than absolutely necessary. My ears tell me that this minimalist approach sounds most natural.
Even though the crossover network is complex with lots of components, the overall strategy of the crossover is pretty simple as you see from the drive signal curves.
These do not have the liquid transparency and razor sharp detail of the Live Edge Dipole Beryllium tweeters and they don't have the bass output of an 18" woofer with 8mm xmax; after all this is a 15" woofer with 4mm Xmax. But the high end is very detailed and not fatiguing and the imaging is deep and very 3D. They reach very solidly down to 40Hz; what is missing below that is not easily noticed; and the subsonic filter very effectively limits undue excursion so they can play quite loud. They're almost party speakers in that respect.
So I think the Alpha 15As with this crossover have fantastic midrange. It it smooth and uncolored, integrates very nicely with the tweeter (only some minor radiation pattern ripples between 1500Hz and 2Khz) and they will easily hold their own with commercial designs in the $10K to $20K range.
Last edited:
Hey Perry,
I think that last reply was most useful to my situation. I have a pair of beautiful Butternut slabs that I bought to build your "Cottonwood" design with, also the DSP, but not the drivers just yet. I am now sufficiently reassured that the (far) more expensive drivers (and DSP) will be well worth the added cost. I might want to try the passive route you described later on though, so I can hear for myself the differences between the two approaches. While there is always "more than one way to skin a cat" some ways are less messy, and it becomes a question of tradeoffs - the essence of engineering IMO. Thank you for continuing to do the hard R&D coupled with empirical results & explanations for laymen builders like me!
I think that last reply was most useful to my situation. I have a pair of beautiful Butternut slabs that I bought to build your "Cottonwood" design with, also the DSP, but not the drivers just yet. I am now sufficiently reassured that the (far) more expensive drivers (and DSP) will be well worth the added cost. I might want to try the passive route you described later on though, so I can hear for myself the differences between the two approaches. While there is always "more than one way to skin a cat" some ways are less messy, and it becomes a question of tradeoffs - the essence of engineering IMO. Thank you for continuing to do the hard R&D coupled with empirical results & explanations for laymen builders like me!
I posted more examples of substituting notch filters for LP filters here:
https://www.diyaudio.com/community/...-pass-xovers-with-notch-shelf-filters.404415/
https://www.diyaudio.com/community/...-pass-xovers-with-notch-shelf-filters.404415/
@perrymarshall Thank you for sharing this project it looks super and a very practical build.
Would you share the FR and impedance files too? And the 15OB350 files...? Thanks!
@Rod Coleman here you go - zip file of several items: FRD files of Eminence Alpha 15A in the Walnut Dipole Enclosure; FRD files at 0 15 30 45 60 75 90 for it as well as the SB Acoustics TW29R tweeter in the Visaton WG 148R waveguide; and impedance files of the PRV WG230TI tweeter. I don't have a raw frequency response curve of the PRV here. Also included are nearfield and farfield curves for the 15OB350 taken from the middle woofer of the Bitches Brew speakers.
Great looking and performing system you put together there Perry. I hope to share my experience with a dipole AMT soon…..i figure no added cost of a separate tweeter or time based offset……I don’t think at this point I can assume it would perform any better than your approach though……the vertical window is obviously much smaller so we’ll soon see how that sounds.
The Alpha 15a is loved by many……but the cost has gone up significantly in a short period of time and I’m not sure it’s the value leader it once was. We’ll see if B&C can right that ship and get Eminence back on track.
The Alpha 15a is loved by many……but the cost has gone up significantly in a short period of time and I’m not sure it’s the value leader it once was. We’ll see if B&C can right that ship and get Eminence back on track.