![IMG_20210513_172226[1].jpg](/community/data/attachments/867/867150-f5e4badff150c198375eb38726d5e6d1.jpg)
after listening to V1 and V2 for long hours,
i started thinking sth was seriously wrong
i thought that the 4th order crossover might be to blame, and i was right.
with LR24db CO, songs i know well sounded like covers recorded by aliens.
i remember someone on the forum called it phasey and it fits,
when i switched to LR12db CO, the undercover alien invasion was immediately over.
however 2nd order CO brings its own problems (there is always a compromise).
i have tinnitus and am extremely sensitive to upper midrange/treble harshness and ringing. we tend to think that tweeters are the no 1 offender in this department. but i now firmly believe that cone driver break up is the most common reason for speaker generated harshness.
that is why i only use cone drivers with very light paper diaphragms and a mellow break up.
with symmetrical LR12db CO, both V1 and V2 have too much harshness and ringing for me (although the cone drivers here are known to be the most mellow in the industry and also the CO points are quite low in respect to the corresponding break up regions).
harshness becomes obvious when there is a busy mid/treble heavy crescendo in the music.
and i can clearly hear the ringing when ie a tenor holds a high note for long; ringing can be heard as a separate layer on top of the original note.
i also like ES and planar drivers because they don't have this problem; at least not in the critical 1-5khz range.
by chance, a few days ago i came across someone promoting asymmetrical 4th and 2nd order COs.
so i decided to try that. i did a LR24db lowpass on the cone drivers and a LR12db highpass on the driver they CO to.
it was a big improvement for harshness and there was also no sign of the aliens.
this set up would pass as an acceptable and enjoyable speaker but with difficult music/recordings 20-30% of the harshness and ringing is still there.
onwards to V3...
i started thinking sth was seriously wrong
i thought that the 4th order crossover might be to blame, and i was right.
with LR24db CO, songs i know well sounded like covers recorded by aliens.
i remember someone on the forum called it phasey and it fits,
when i switched to LR12db CO, the undercover alien invasion was immediately over.
however 2nd order CO brings its own problems (there is always a compromise).
i have tinnitus and am extremely sensitive to upper midrange/treble harshness and ringing. we tend to think that tweeters are the no 1 offender in this department. but i now firmly believe that cone driver break up is the most common reason for speaker generated harshness.
that is why i only use cone drivers with very light paper diaphragms and a mellow break up.
with symmetrical LR12db CO, both V1 and V2 have too much harshness and ringing for me (although the cone drivers here are known to be the most mellow in the industry and also the CO points are quite low in respect to the corresponding break up regions).
harshness becomes obvious when there is a busy mid/treble heavy crescendo in the music.
and i can clearly hear the ringing when ie a tenor holds a high note for long; ringing can be heard as a separate layer on top of the original note.
i also like ES and planar drivers because they don't have this problem; at least not in the critical 1-5khz range.
by chance, a few days ago i came across someone promoting asymmetrical 4th and 2nd order COs.
so i decided to try that. i did a LR24db lowpass on the cone drivers and a LR12db highpass on the driver they CO to.
it was a big improvement for harshness and there was also no sign of the aliens.
this set up would pass as an acceptable and enjoyable speaker but with difficult music/recordings 20-30% of the harshness and ringing is still there.
onwards to V3...
Last edited:
Hello. Nice build. Along the ways of something I've been experimenting with. Are your crossovers active or passive?
Mine is a 4-way stereo: dual 12" subs below 65Hz, 18" Faital 18HP1000 up to 275Hz, MTM with two B&C 8PE21 up to 1600Hz and Beyma TPL-150H open back with a foam waveguide behind. I use digital xo and initially I tried LR 2nd order and didn't like them, and moved on to Neville-Thiele 2nd order (they drop vertically after a transition region) and have been happy. Not saying this is my final xo, though! The design is far from finished and I've been exploring baffle shapes for the midranges. I am now on naked top mid and wings around bottom mid to allow 275Hz xo. Maybe you can try NT filters?
BTW, I'm so happy with how open these sound. The 18" are hanging on a swing, naked. They provide such an openness to bass, which has improved with lowering the 18"-to-2x8" xo from 300 to 275Hz. I guess the 18" is rather heavy and the dual 8" move faster in that region and that's what I'm hearing.
Have you tried a higher xo for the Neos? My TPLs wouldn't do good that low. In fact my prior iteration had them at 1500 and I'm liking 1600Hz better.
Your soft panels around mids have me thinking I should try lossy wings around the bottom mids and see if that reduces the resonances (I'm guessing what I see on measurements are resonances from the wings).
Mine is a 4-way stereo: dual 12" subs below 65Hz, 18" Faital 18HP1000 up to 275Hz, MTM with two B&C 8PE21 up to 1600Hz and Beyma TPL-150H open back with a foam waveguide behind. I use digital xo and initially I tried LR 2nd order and didn't like them, and moved on to Neville-Thiele 2nd order (they drop vertically after a transition region) and have been happy. Not saying this is my final xo, though! The design is far from finished and I've been exploring baffle shapes for the midranges. I am now on naked top mid and wings around bottom mid to allow 275Hz xo. Maybe you can try NT filters?
BTW, I'm so happy with how open these sound. The 18" are hanging on a swing, naked. They provide such an openness to bass, which has improved with lowering the 18"-to-2x8" xo from 300 to 275Hz. I guess the 18" is rather heavy and the dual 8" move faster in that region and that's what I'm hearing.
Have you tried a higher xo for the Neos? My TPLs wouldn't do good that low. In fact my prior iteration had them at 1500 and I'm liking 1600Hz better.
Your soft panels around mids have me thinking I should try lossy wings around the bottom mids and see if that reduces the resonances (I'm guessing what I see on measurements are resonances from the wings).
hi Lewinski,
i have been following your project... nice.
i am indeed proud of my lossy wings idea. it does eliminate the resonance problem and also help better airload the driver cone. my idea is an extension of OllBoll's lossy baffle invention.
NT crossover seems really promising and the software you use is very cool. i am using nanoAVR from miniDSP. it is simple to use but does not have many options for CO (no NT filter option per se).
your TPL must be better sounding than the neo8s. but the neo8s can be crossed much lower; 800hz is no problem.
naked swinging drivers have many benefits. but when they swing with the move of the diaphragm, this results in loss of efficiency. you need to add mass to the driver to counter this effect (some say upto 250 times the mass of the diaphragm).
i have been following your project... nice.
i am indeed proud of my lossy wings idea. it does eliminate the resonance problem and also help better airload the driver cone. my idea is an extension of OllBoll's lossy baffle invention.
NT crossover seems really promising and the software you use is very cool. i am using nanoAVR from miniDSP. it is simple to use but does not have many options for CO (no NT filter option per se).
your TPL must be better sounding than the neo8s. but the neo8s can be crossed much lower; 800hz is no problem.
naked swinging drivers have many benefits. but when they swing with the move of the diaphragm, this results in loss of efficiency. you need to add mass to the driver to counter this effect (some say upto 250 times the mass of the diaphragm).
Last edited:
I like those vintage Seas drivers!
Still the same xo point? I currently use 12" drivers from electric organs. May have to try the Seas units again. Same problem with tinnitus and sensibility to harshness. I have resently figured out I have crossed them too high. Harshness comes from the mids here, not the compression tweeter.
Still the same xo point? I currently use 12" drivers from electric organs. May have to try the Seas units again. Same problem with tinnitus and sensibility to harshness. I have resently figured out I have crossed them too high. Harshness comes from the mids here, not the compression tweeter.
here is the final prototype: V3
mids are BMR units salvaged from Q acoustics soundbar
crossover points at 200 and 800hz
this version sounds really great; i can not hear any cone break up problems
can be due to small driver diameter or bmr technology (flat honeycomb diaphragm)
i used felt tweak to reduce vertical dispersion problem... but imo it was okay even without the felt tweak
in the final built, i will put the mids together ( :: ) and the neo8s on top
mids are BMR units salvaged from Q acoustics soundbar
crossover points at 200 and 800hz
this version sounds really great; i can not hear any cone break up problems
can be due to small driver diameter or bmr technology (flat honeycomb diaphragm)
i used felt tweak to reduce vertical dispersion problem... but imo it was okay even without the felt tweak
in the final built, i will put the mids together ( :: ) and the neo8s on top
Attachments
here is V3.1 (this time stereo)
while i wait marble people to build me a bass frame...
bass driver is a free hanging faital 18fh500
mid frame has lossy side walls.
neo8s is treated with felt front and back to increase vertical dispersion and also provide a lossy baffle to take the driver flat down to 800hz.
while i wait marble people to build me a bass frame...
bass driver is a free hanging faital 18fh500
mid frame has lossy side walls.
neo8s is treated with felt front and back to increase vertical dispersion and also provide a lossy baffle to take the driver flat down to 800hz.
Attachments
Looking good.
An idea to test, suggested to me by an experienced user here: try tilting you 18" back, so their centerline are approx pointing to your head. I liked it.
Did you test your lossy mid side walls vs solid? What did you observed, in what frequency range?
I forgot to say;
the surround is not flat at the back, ie it is an inverted rubber surround
looks cool
Neo 8
This is a cool project. I'd love to try those Great Plains woofers some day. I have used the Neo 8s in the past and wanted to pass along a comment if it's useful to you. I've run the Neo 8 from ~1KHz and up and really loved the open clean sound on mids and lower treble, but there was a 'tizz' or sibilance issue around 8-10 K that always bugged me. I know it add complexity, but if you're tweaking this further, you might consider using a dedicated (smaller) tweeter.
This is just a single comment so no worries if you're not hearing this issue. All the best with the project!
This is a cool project. I'd love to try those Great Plains woofers some day. I have used the Neo 8s in the past and wanted to pass along a comment if it's useful to you. I've run the Neo 8 from ~1KHz and up and really loved the open clean sound on mids and lower treble, but there was a 'tizz' or sibilance issue around 8-10 K that always bugged me. I know it add complexity, but if you're tweaking this further, you might consider using a dedicated (smaller) tweeter.
This is just a single comment so no worries if you're not hearing this issue. All the best with the project!
Not deep cone per se, but its curvilinear profile Vs JBL's or similar conical profile: Is it possible to cover the whole spectrum, high spl, low distortion with a 2-way?
Not surprised considering it's a steep decay combined with a 1 Hz timing offset at the XO point!
Yes, the neo range of drivers and their clones do have a pronounced HF peak. I do not hear it as sibilance though, This peak is supposed to be related to output summation between the holes on the drivers (pardon my poor language skills). So the good news is that it is not a resonance or other nasty. As such, it is easily dealt with DSP. In my current set up (with EQ) neo8s plays very musical, detailed and sweet.
Neo8s is actually the starting point of my speaker: ie everything is designed around it.
I did the initial prototyping in mono (pure laziness). LR4 sounded phasey in that setting. Once I decided on the driver configuration, I did build a final stereo prototype and went back to working on the XO.
Even with LR4 lowpass at 200hz, I was hearing the breakup of woofer (glare on some piano notes/recordings).
So I tried LR8/LR4 first and then LR8/LR8 later. To my surprise, I could detect no obvious phazeyness in stereo: it was just normal balanced music. I did revisit LR4/LR4, which sounded very problematic in mono, and no it also sounded quite acceptable in stereo. Just my observation; hope it can be verified by others.
So my current XOs are LR8/LR8 at 175 and 700hz which reduces cone breakup issues down to almost nil.
a happy side effect is that now, CtoC distance at both XO point is at or below 1/4 wavelength, a very rare thing in 3 way designs.