DIY cardioid monitors

Discussion in 'Speakers' started by Serious, Mar 14, 2023.

  1. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I consider my OBs a "subjectivist" speaker and I kind of wanted to build one that is more into the "objectivist" camp for a while now.
    There are actually three designs based on this concept. Omni bass transitioning to cardioid midrange/treble.

    This is the first one I built and might be the best sounding. If it didn't sound great I wouldn't bother spending hundreds of hours tweaking it to perfection :)

    Here's what they look like:

    [​IMG]

    First the current frequency response in-room:

    Hathor XO3 in room FR.jpg

    The peak at 8kHz is the only annoyance and it's not as bad as it looks IMO. That's not to say it isn't there. As you will soon see, these are brually honest and revealing speakers and the peak is the price we have to pay with the driver selection.


    The more interesting bit: Directivity (this is from the second crossover revision, however I normalized it so it should still look the same)

    Hathor XO2 full speaker angles Directivity (hor).png

    Again, not perfect. Some narrowing at 1kHz and widening around two or three resonances at 2.5kHz and 8kHz. However listening to how it sounds outdoors it has come a long way. It doesn't sound weird at 90 degrees off axis anymore, which is entirely due to acoustic mods I have applied. This is mostly based on things I learned when modding headphones.

    There are ways to get a smoother directivity plot, but I'm fairly certain it would hurt openness and would require a different (worse IMO) crossover.

    And finally, this wouldn't be an objectivist speaker designed by me without a good step response:

    Hathor XO3 in room step response.jpg

    I will likely get proper measurements of the current config outdoors as soon as the storm and rain has calmed down.
     
    • Like Like x 16
    • Epic Epic x 2
    • List
    Last edited: Apr 19, 2023
  2. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Well, my tease didn't really go anywhere. Still, this is what they look like:

    Of course I tweaked them some more. Better directivity.

    XO3 directivity tweaks.gif

    I don't think I want to go any wider in the sub 1kHz region so I'm pretty much done.
    Of course vertical directivity suffers a bit with this design, but preliminary measurements and simulations in VituixCAD don't show any nasty suckouts or peaks, just a midrange "tone control". I likely didn't quite match the microphone height for the left and right speaker, so this can be seen in the image below.

    Attached are the outdoor frequency response and step response graphs. The bass extension isn't nearfield, this is what I get just from a normal outdoor measurement. - 3dB is in the 25-28Hz region.
    The outdoor frequency response looks very similar to before and so does the in-room response in my room. In the living room (where there's more reverb) it's a bit more linear now.

    EDIT: Attached on axis at 2.83Vrms. Circa 85dB sensitivity with some extra bass. I think I can get some more midrange energy, which is what the next crossover revision will be focused on to get me closer to 86-87dB.
     

    Attached Files:

    Last edited: Apr 19, 2023
  3. Priidik

    Priidik MOT: Estelon

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    2,153
    Trophy Points:
    113
    Location:
    Estonia
    Cool tinkering. Is the mega tweeter a ceramic?
    People have different perspectives on this; I will wager it would sound more lifelike and normal with a tiny tweeter added from 6k upwards; 1. order xo is fine, even for objectivist purposes when you play around with it long enough.
     
  4. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    So a one bass driver approach, but cardioid via enclosure design? What is the size of the woofer?

    I wouldn't obsess about polar response at 1kHz, especially with the mega-tweeter. Tradeoffs with such a large tweeter. Don't be like Amir ASR and have tunnel vision. Can you lower the xover point for the tweeter?
     
    Last edited: Mar 24, 2023
  5. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    @Priidik As for an extra tweeter: I'd be worried about vertical off-axis lobing. But it's an idea. I would want a metal tweeter for coherency, though. Or maybe ceramic.
    The woofer had to be a metal cone aswell for it to be coherent. Basically a 9" woofer and 4-5" widebander. From cone size it's 8" + 4", from basket size it's 9" + 5".

    It's the TB W4-1996. Flat Al honeycomb sandwich diaphragm. Neo magnet, large pole piece hole, etc. It's really an amazing driver, by far the best 4" widebander I've heard. Not quite as good as the Voxativs IMO, but I feel it puts the LS50s to shame. Yes, I think it's that good.
    The two main problems are of course the 8kHz peak and the protection grille is super ringy. I'd love to find a solution for it, but maybe it just has to go.

    @purr1n Yup, partially open enclosure for the tweeter, with added "flow resistance" (a combination of materials). Basically like the Dutch & Dutch 8C in that way.
    The woofer is ported to 28-30ish Hz. I did not expect to get this much bass from a 35L enclosure. Sensitivity is in the 85 to maybe 87dB region, though.
    (i just saw that the FR I posted was 5-6dB too low in level.)

    Crossover point is really low as is. 150 to 200Hz. Basically where the bump is located. The big coil is from a project I did 10 years ago and is in parallel to the coil from the PAP clone. Ideally I think I'd want a slightly larger coil than the parallel coils, so the bump around 100-200Hz should get a bit less.

    No high pass on the widebander, otherwise I couldn't get the crossover to work. The big tradeoff is excursion from the tweeter. Distortion doesn't seem to be an issue and personally I don't really hear any glaring problems even at a couple millimeters of excursion. This widebander can handle some abuse. But of course it's not ideal and these aren't disco speakers. I've tried bassy tracks at 100dB without damage, but I'd still be worried someone manages to rip the suspension :)

    The good thing is the more "closed" the enclosure is, the more it acts as a, well, closed enclosure and will limit excursion somewhat.

    I could make a notch filter for the peak, but it would need quite a high Q, so a large-ish cap (expensive and probably not that transparent).
    EDIT: Nevermind, absolutely tiny coil and small capacitor I meant. But tolerances are the main issue then. Pretty much impossible from simulations. Coils less than 0,1mH are hard to come by, etc.
     
    Last edited: Mar 25, 2023
  6. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I just ran some distortion measurements I wanted to share with you guys:

    Hathor XO 3 with tweaks - prototype distortion at 90dB.jpg

    Note that there was some rattling around 120Hz explaining the spike there.

    More distortion measurements can be found here: https://imgur.com/a/3PpN4qg

    Distortion in this case behaves linearly with volume level. Meaning that at 85dB distortion will be 5dB lower relatively. So -40dB in the bass instead of -35dB at 90dB. This is not the case with every speaker and is a good quality IMO.

    What's bad is the bass distortion. The woofer isn't the best. The Canton Reference 3K has two 9" woofers, but instead of 6dB better distortion like we might expect it's better by 10dB. Also there's a (mostly) D2 peak at 2kHz. Maybe I'm deaf, but this does not bother me. Ditto for the rise in distortion past 8kHz. I don't think I can discern distortion products at 16kHz anyway.

    Funnily enough it has lower bass distortion than the Dutch & Dutch 8C and also much lower D3 in the midrange. Attached.
     

    Attached Files:

    • Like Like x 3
    • Epic Epic x 2
    • List
    Last edited: Mar 24, 2023
  7. zerodeefex

    zerodeefex SBAF's Imelda Marcos

    Staff Member Pyrate BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    14,051
    Trophy Points:
    113
    This is a pretty badass project!
     
    • Agreed, ditto, +1 Agreed, ditto, +1 x 2
    • List
  8. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    The development is pretty much done for now. This is the crossover I will put on a board and in the speakers.
    After that it's time to finish the speakers. I want the backsides to be glossy white and the front to stay as is...


    This is one half of how the development works. Each crossover revision is generally a day of measuring outdoors to check against the simulaton. And they generally don't quite match perfectly in this case - first order crossovers seem to be harder in that sense because the drivers overlap so much more.
    Then lots of listening, tweaking the simulation and soldering a new crossover. Then listening and measuring again.
    The good thing with smaller speakers like these is that I can take outdoor measurements and listen in two different rooms all in one day. With my large speakers it took a whole day just to get measurements.

    Hathor outdoor angles setup.jpg

    I've completely changed the crossover topology and the measurements aren't really better now, but the sound is much cleaner. This is from a steeper woofer rolloff past 3kHz to prevent the peaks of the woofer from interfering with the MT sound. Before I basically didn't deal with the peaks from the alu woofer at all. The crossover is actually a bit simpler now, too.

    What's new is I've built a notch filter with a tiny coil I wound myself. I had all the parts except the coil and since it had to be exact I just wound it myself. It's easy to confirm the notch frequency and Q-factor with an impedance measurement, plus the coil's inductivity.
    The bad thing is I used an el-cheapo coil for this (didn't want to unwind a quality coil) with too thin a wire and the DCR is now too high - lowering the Q-factor of the notch. Ideally it should be flat at 8kHz, which it isn't.

    I attached a bunch of measurements, plus here are two of the current frequency response at the angle that I would call the listening angle. This is circa 17 degrees off axis.

    Hathor XO8 with DIY notch Listening angle at 50cm distance.jpg

    The midrange isn't the smoothest anymore. The sensitivity at 1m is barely 85dB again, it was higher when the woofer was contributing more to the sound, but the impedance is also a bit higher now.

    Step response doesn't have any initial overshoot at that angle anymore. Ringing is greatly reduced from the notch filter.
    Listening angle at 50cm distance step.jpg


    Quick comments on the attached measurements:

    Directivity plot:
    This one isn't normalized, so the 1kHz dip looks more severe now. I chose this visualisation because the widening at 2kHz (where there is a driver resonance) isn't too severe IMO. The line chart only goes to 90 degrees off-axis as it's hard to accurately measure the rear angles. Rear damping isn't the greatest with this design, though. But I do think it's enough to be able to call it a cardioid speaker. Maybe not quite - a sub cardioid/wide cardioid in the lower midrange and a cardioid in the upper midrange.

    On-axis response:
    Shows a broad peak in the top octave, extension up to 40kHz (the response drops off rapidly past 40kHz) and a supremely flat excesss phase and corresponding phase response. I'm especially happy with the excess phase response.

    RTA FR:
    This is one look at the in-room response. This is already an average over multiple positions RTA measurement, but change listening position or speaker position slightly and the midrange and bass will look different. I think I got lucky with my first in-room response.

    Power+DI:
    This is the graph ASR loves and takes as gospel. Whatever, it is a good tool for tuning a speaker, but it's not the only thing to look for. The simulation looks quite a bit more linear here, but I'm showing you guys the actual measurement which looked worse in this case. I may need to tweak the crossover slightly, but that means ordering parts and spending quite a bit of money that I currently can't allocate to this project due to life circumstances :(

    Impedance:
    Self explanatory. This is the speaker impedance. Minimum is at the port tuning frequency of 27.5Hz at 3.1Ohm. It's a bit bumpy, but not overly low. Raising towards high frequencies. I attached both a linear and a log graph.
     
    Last edited: Apr 19, 2023
  9. AdvanTech

    AdvanTech Friend

    Pyrate BWC
    Joined:
    May 13, 2016
    Likes Received:
    1,667
    Trophy Points:
    93
    Location:
    NYC
    What do you like about the presentation of these speakers vs your OBs?
     
  10. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Not all too much, really. They image worse, yet sound smaller. They're sharper, but also somehow slower sounding. The bass does dig deep, arguably deeper than my OBs, but it's not nearly as tactile an experience. It's just not as much fun. They resolve really well and they really scale, but they're no match for the Voxativs.

    What they do have is a totally different presentation. Paper drivers vs alu cones. Especially the sandwich driver has a very interesting presentation to it that I like. It's not sharp like regular alu, but still manages to have that laser precise hard cone sound. Maybe it comes down to different kinds of speed. Or decay, rather. As can be assumed from the phase response and the supremely flat response in the 2-10kHz area at the right angle, the CSDs look incredible.

    Hathor XO8 with DIY notch listening angle CSD.jpg

    The Voxativs are incredible and do things I've not heard another speaker do. These are more like your traditional multi-way speakers. They do remind me of Accutons somewhat.

    They also have some of that open OB sound, no doubt due to their half-open design and the resulitng cardioid dispersion pattern. They have a similar sound "projection". But unlike OBs these are easy to position. They sound good wherever I put them, the OBs only sounded good in my room (and outdoors lol).

    It's also about compromises. The OBs are sort of a no-compromise speaker for me. These aren't and I wanted to see what I could do with size and budget constraints. I already think they turned out quite well overall. I'm biased but let's say I think they easily outresolve the LS50 Meta. Tbh I think they might outresolve my dad's Canton Reference 3K.
    I mean the Boenicke B10s used these drivers 10 years ago and they sold for circa 20k a pair.

    Finally, it's variety and a breath of fresh air. Almost cold air from the raised top octave, hah.


    So generally it's not really all that much about their presentation, but rather about size, positioning, transportability and generally having a more normal speaker.
     
    Last edited: Apr 20, 2023
  11. AdvanTech

    AdvanTech Friend

    Pyrate BWC
    Joined:
    May 13, 2016
    Likes Received:
    1,667
    Trophy Points:
    93
    Location:
    NYC
    Yeah, I laughed a bit when you said you wanted to build an objectivist speaker because, by most performance metrics, your OBs were still probably going to be objectively better. :)
     
  12. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I was going for something like my variant of a Dutch&Dutch 8C or Kii Three cardioid speaker. Distortion is quite high with both of these, but dispersion and overall frequency response is very good. And I wanted similar full-range sound (good bass extension) from a small enclosure.

    You could say that these are more neutral than my OBs, especially since they work better in the living room where there is a lot less direct sound and the RT60 is in the 550ms range instead of 250ms like in my listening room. The OBs sounded too thin there.

    The dispersion graph doesn't look too pretty for the OBs:
    2022 Directivity (hor).png

    Much narrower, yes. But also kind o firregular. Especially the rear angles aren't pretty.
     
  13. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Looking at the above dispersion graph I realized that I probably fucked up somewhere. I'll have to eat my words that these are done. They weren't very cardioid :oops:

    To confirm I measured the LS50 Meta. Yup, not too much narrower in the lower midrange that that: f**k!


    So back to the drawing board. This is what I got by removing some damping material from the MT section. Frankly I had simply overdamped them by concentrating too much on the 1kHz dip, making the enclosure closer to closed than ideal.

    Before:
    Hathor XO8 with DIY notch full speaker angles Power+DI.png Hathor XO8 with DIY notch full speaker angles Directivity (hor).png

    After:
    Hathor XO8 with DIY notch and tweaks full speaker angles Power+DI.png Hathor XO8 with DIY notch and tweaks full speaker angles Directivity (hor).png

    Much better, but also quite a bit leaner. You have to be careful not to get a strong dip at 150Hz, but I think this is probably closer to "cardioid" than it was at any time before.


    In terms of sound: More open, more direct, better imaging. Leaner, bordering on too lean.
     

    Attached Files:

    Last edited: Apr 21, 2023
  14. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Quick demonstration of the speakers' dispersion pattern:
    https://youtube.com/shorts/eUP9ZSAyyaQ?feature=share

    Bought parts for a better notch filter (higher Q):
    Hathor XO10 notch filter.jpg
    This is part of XO10 and kills the 8kHz peak pretty much completely. Now there's still too much air, so if you prefer a KEF sound you could add a 0.05mH series coil to the tweeter. I might give it a shot.
    The crossover is now finally on a board inside the speaker. I think that makes it a real speaker, or at the least a prototype of the finished speaker now. I'm happy with XO10.


    FR is better now. Dispersion may look worse, but I measured at a larger distance, which makes it look worse. It's sadly more realistic this way. However we can also judge LF extension better now. Dispersion likely hasn't changed much as I didn't really touch the damping materials.

    Distortion is better, but not all of the rattle has been eliminated for some reason. I'll have to check again what's causing the rattling. Might be a wire inside the speaker.
     

    Attached Files:

    Last edited: May 3, 2023
  15. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    FINISHED PROTOTYPE:

    I spent quite a bit of time trying to get them to look how I want, while also going up one revision in the crossover department, The back is now glossy white and they're at XO11. This time around I think they are truly "finished" (get it, because they have a proper finish now :p)

    I'm planning to make an extreme version of these speakers using the 8" alu cone Purifi woofer and better crossover components in a balanced configuration at some point.
    I want to make a loaner of either one of the speakers and start selling them after some feedback and subsequent tweaks. At least that is the plan. If you guys think they suck, I'll need to make another pair of speakers that I can sell. I initially wanted to sell a three way first-order monitor with a coaxial midrange + tweeter, but I didn't think that one sounded good enough. This one does.

    Hathor L side.jpg

    I put a grille on the side so nobody can press in the damping material. It's essential for the sound for it to stay that way and it's a PITA to disassemble the speaker just to move it back into position.
    I also removed the grille on the widebander and in doing so I found out that the diaphragm is indeed very thin. When trying to put the grille back on it got attracted to the magnet and bumped into the diaphragm :/. Sounds painful, but there wasn't too much damage and neither the measurements, nor my ear can detect a difference. Phew! Live and learn. It's a bit ironic that trying to put the grille back on is ultimately what scratched/dented the driver. Oh well.


    The only real change going from XO10 to XO11 is that I added a second notch filter to tame the top octave. I used an MKP and the special coils along with a very good quality wire-wound resistor. To me it's close to transparent when comparing to digital EQ. Plus I find it well worth doing so because the treble quality is now something else entirely.

    Hathor XO10 full speaker angles with 2nd notch Power+DI.png

    The in-room response is maybe even somewhat smoother than the predicted in-room response, with less of an 8kHz peak and a better midrange balance.

    Hathor XO11 RTA at LP (different position).jpg

    I cannot stress this enough: The treble from this small aluminium widebander has to be heard! Sibilants now sound supremely natural and everything comes through in such a life-like fashion it's uncanny. As a whole it sounds more realistic than most mid-tweeter combinations I have heard. Very few widebanders manage to reproduce treble in a way that it sounds this natural. The Voxativs are similar, but a different timbre with their paper diaphragm. Still, this one doesn't sound harsh or "metallic" at all. This is true hifi.

    Monitor Audio's Hyphn flagship should be interesting in this sense as they appear to use 6 of the smaller 3" drivers around an AMT in the center.

    Website link: https://www.seriousaudiodesign.de/audio-out/hathor

    EDIT: Replaced the FR with one at a different position and angle that I feel is somewhat more representative. Other one included below.
     

    Attached Files:

    Last edited: May 17, 2023
  16. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Have I been listening to them the wrong way all the time?

    20230525_205826.jpg

    I gave listening to them upside down a try just now. At least on paper it has the advantage of there being less floor bounce. Not just because the woofer is physically farther from the floor, but also because of the lobing with this design. There's more energy towards the bottom than there is towards the top.

    (The way you would normally deal with this would be to move the woofer forward, closer to the listener. But then the top lobe sounds even thinner. I chose this dZ as a compromise and I think it works well.)

    So what do I think?
    • Imaging could be said to be marginally more precise in the y-axis. Will depend on the floor and if your ear height really matches the widebander.
    • It looks dorky and it becomes tall and top heavy. Not good.
    • It sounds worse when standing up.
    Hardcore audiophiles or hardcore objectivists might prefer them upside down.
    But for a speaker inside a living room that isn't purely an audio dungeon I think having them the right way around works better.

    So in conclusion? No, I think woofer down is better in this case. Even if I generally prefer speakers with a midrange above their tweeter, as long as the tweeter is kept at ear height.
     
    • Like Like x 2
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
  17. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Updated measurements (brutally honest):


    Frequency response:

    Hathor FR in room.jpg


    Distortion:
    Hathor distortion.gif


    Still using the same crappy woofer and it's really not too terrible all things considered. There's quite a bit of port chuffing at the higher volume levels and the baffle is still not glued, which I think explains the D4 and D5 in the bass.

    The distortion spike at 2kHz is from the surround, as far as I know. Still, I do not like seeing D3 and D4/D5 this high. Otherwise an excellent result with very low distortion in the midrange, especially D3. This is kind of surprising for a metal driver. Most metal and ceramic drivers seem to be D3 dominant in the midrange.

    I think that's the main reason I like the presentation of this driver so much. It's kind of hard to explain, because while it is a metal driver sound it also avoids some of the glaring faults of many metal drivers.
    It sounds pretty darn transparent to me.

    Compare to Amon here:
    https://www.superbestaudiofriends.o...al-ob-speaker-project.4083/page-4#post-416190
     
  18. Aklegal

    Aklegal Friend

    Pyrate Contributor
    Joined:
    Mar 1, 2016
    Likes Received:
    623
    Trophy Points:
    93
    Location:
    Atlanta, GA
    Fantastic.
    I was (still sort of am) considering on embarking designing a speaker with a cardioid mid and tweeter section with a U-Frame or H-Frame bass section similar to these by Penaudio https://penaudio.fi/floorstanders/43-karelia.html

    I was going to add a rear-firing high-frequency driver as well.

    It's awesome that you had success with this.

    I have been an open baffle guy for about 12 years - my current speakers are the Serenity Acoustics / GR-Research Super 7s.
     
  19. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    It's been a while, but I may want to give this a shot later on... I tried simulating it with the 26mm SB ceramic tweeter and got a decent result after some time.

    It's cheap and interesting enough to just buy one and a couple components and if the speaker measures fine then I can always buy a second tweeter and actually listen to it.

    Obviously a simulation without the proper baffle and relying on two different measurements that I found on the internet + extrapolating it to 180 degrees is not going to be accurate. I just wanted to see if it can work in theory.

    Some comments based on the simulation only:
    • Had to add an impedance correction circuit to the midrange to compensate for the peaks from the notch filters. Thankfully one works just fine. And then a single series coil works well enough.
    • The tweeter was similar. Impedance correction circuit, a series resistor (normal L pad didn't really work) and only a capacitor.
    • Overall I tried to keep the changes to the crossover minor. The series resistor doesn't need to have a super high power rating, etc. It seems to work in theory.
    I tried to keep the resulting FR and the horizontal dispersion the same. Power response is also similar. Step response was similar except now there's a resonance around 26kHz that results in quite a bit of ringing (as was the case with my LS50 Meta measurements, for example).
    With one caveat: It takes a 1dB lower sensitivity to make it work, it seems. That also means it'll be bassier. Impedance is fine.

    So in general it could work and it seems to be worth a shot.
     
    Last edited: Dec 13, 2023
  20. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    So what's the problem here, why haven't I done this before?

    Vertical dispersion with the tweeter:
    Sim 1 Theory Directivity (ver).png

    Instead of without the tweeter:
    Hathor XO11 full speaker angles imp fixed Directivity (ver).png
    (actual measurement)

    Ugly comb filtering starting at 3kHz. Though I still think it could sound fine. Considering that the tweter is only 50 bucks and I have a feeling it could mate well it could work. I wonder if the tweeter can be modified for a lower Q at Fs, which would improve the crossover a lot. Hopefully it's the volume and not the magnet. Fs is quite low, but a slightly lower Fs would be even better, of course. And a smaller faceplate, too :D
    I think it's also interesting since we can try it with the phase shield and without.

    Another thought: I think it would take a very high quality series capacitor to match the treble quality from the widebander. From previous projects with tweeters the treble quality and resolution always suffered from the capacitors I used. Probably quite a bit more expensive than the tweeter itself.
     

    Attached Files:

Share This Page