Hey all. I've fantasized for a while about building a bass tower to go with my HiVi 3.1's. The L6-6R is so cheap and I really quite like it's cousin, the L6-4R, which is already in the bookshelves. The goal is not to create a subwoofer, or necessarily modify my existing bookshelves, which are DSP corrected. It is to increase the headroom and bass output by essentially converting it to a 3.5-way with a null in the rear. I may have to plug the bookshelf's port to do this. I don't know yet. For simplicity's sake, I'm simulating a baffle and a sealed enclosure, but not using any sort of measurements. Again-- purely theoretical.
This all started when I saw a great video on bass and airflow on Youtube. I highly recommend it if vents, cardioid setups, or just bass in general is a bit mystifying.
I read a few articles on cardioid subwoofers and how to set them up. It's easier than I thought. Essentially you create a delay so the out-of-phase wave created by the rear speaker is arriving at the same time as the forward in-phase speaker. This negative pressure also creates a null behind the speaker, canceling noise near the rear speaker's source. This doesn't require as much energy as producing the actual original sound wave, so the out-of-phase rear speakers also play at reduced volume. I wonder if having all six drivers in the same enclosure would be better or worse than subdividing them individually. Maybe I'll find out if I build them!
I'd be happy to take any critiques or suggestions. I don't really know what's going on with the group delay or that resonance. In fact, I'm not sure if VituixCAD can appropriately demonstrate a cardioid setup. And yet, here it is. A heart in a polar response.
This all started when I saw a great video on bass and airflow on Youtube. I highly recommend it if vents, cardioid setups, or just bass in general is a bit mystifying.
I read a few articles on cardioid subwoofers and how to set them up. It's easier than I thought. Essentially you create a delay so the out-of-phase wave created by the rear speaker is arriving at the same time as the forward in-phase speaker. This negative pressure also creates a null behind the speaker, canceling noise near the rear speaker's source. This doesn't require as much energy as producing the actual original sound wave, so the out-of-phase rear speakers also play at reduced volume. I wonder if having all six drivers in the same enclosure would be better or worse than subdividing them individually. Maybe I'll find out if I build them!
I'd be happy to take any critiques or suggestions. I don't really know what's going on with the group delay or that resonance. In fact, I'm not sure if VituixCAD can appropriately demonstrate a cardioid setup. And yet, here it is. A heart in a polar response.
I modeled your system. It looks slightly different because I baked driver groups together in baffle diffraction simulator; however, it should behave the same.
450 mm depth may not be optimal for wide-range operation. I don't have enough expirience with aactive gradient sources to say for sure.
At present it behaves supercardioidish, with quite a thick manta-tail, and there is some lobing above 600 Hz. It may be possible to optimize it further; however, it definitely works as proof-of-concept.
Also, you probably have not rotated your back woofers (R=180 degres).
450 mm depth may not be optimal for wide-range operation. I don't have enough expirience with aactive gradient sources to say for sure.
At present it behaves supercardioidish, with quite a thick manta-tail, and there is some lobing above 600 Hz. It may be possible to optimize it further; however, it definitely works as proof-of-concept.
Also, you probably have not rotated your back woofers (R=180 degres).
Last edited:
Thanks for putting in the effort, dude!
I definitely did rotate them. I may have actually just modeled a four-woofer baffle for each driver. That would definitely cause a bad ripple. These parts of VCAD are confusing to me. Could you explain the "Manta Tail"? And how did you get away with not adjusting the XY axis?
Hm, strange. I've got similar comb filtering when I forgot to rotate the back drivers.
Manta Tail is a non-technical term for backward lobe of supercardioid. I made it up 🙂
Each "driver" in my model is a whole baffle with 4 or 2 drivers. As they share a cabinet and acoustic axis, I only had to enter Z offset and rotation.
Your back driver group is lower than front by 7 dB. This will nullify any directivity at low frequencies.
The same model with back group 7 dB down. At LF it is essentially omni.
If this is supposed to be a subwoofer, shouldn't we focus on what happens below 100Hz? Only the setup in post #2 is cardioid there.
I mostly meant it's supposed to operate between 50/60-250hz. Just like the original woofer. I am very invested in what's happening below 100hz, just not sure I completely understand it. The goal of understanding cardioid subwoofers instead of just cutting slots on the side of the bookshelf/tower is to control that "bass" range from 80-300hz. My sources indicated this happens by adding 4-6ms of delay, reversing and inverting 1/3 of the woofers, and reducing the volume of the rear speakers by a relative 3db. I think I'm getting flustered by the baffle diffraction simulator and need to rethink my use of VCAD in the first place. I'll take another stab at it tonight.
100-300 range is important, and also very problematic with all those reflections yes!
Here are my measurements of four speakers in same placement. Responses are not normalized, that explains huge differences below 100hz. They do sound different, but there are many important factors... AW-7 spl was 55dB, others around 80-85dB.
AINOgradient is cardioid 100-250Hz. MR183w is 3-way with downfire woofer to 300Hz. A-7 and ER18DXT are typical 2-way boxes, AW-7 is sealed.
Here are my measurements of four speakers in same placement. Responses are not normalized, that explains huge differences below 100hz. They do sound different, but there are many important factors... AW-7 spl was 55dB, others around 80-85dB.
AINOgradient is cardioid 100-250Hz. MR183w is 3-way with downfire woofer to 300Hz. A-7 and ER18DXT are typical 2-way boxes, AW-7 is sealed.
I did replicate post 2 effectively. I'll be sticking to polar heatmaps for now instead of the 2D balloons. Caught a fundamental flaw in my logic; I can't change the frequency the polar balloon is set to. At least I can make an active midbass cardioid tower successfully!
What causes so much loss in bass when using the @VoxCelestial post #2 design? With a sensitivity down in the low 80s below 100hz, I have to wonder what the practical value will be after I seal up the port in my bookshelves. Each tower is really only capable of handling 60 watts before bottoming out. With the sub-bass PEQ and preamp, they just don't play as much midrange comparatively.
Chasing my tail
What causes so much loss in bass when using the @VoxCelestial post #2 design? With a sensitivity down in the low 80s below 100hz, I have to wonder what the practical value will be after I seal up the port in my bookshelves. Each tower is really only capable of handling 60 watts before bottoming out. With the sub-bass PEQ and preamp, they just don't play as much midrange comparatively.
Chasing my tail
Cardioid pattern behaves like a dipole below peak spl, it loses spl by acoustic cancellation 6dB/octave. Then add the woofer's own spl loss down low...
https://www.linkwitzlab.com/models.htm#A1
https://www.diyaudio.com/community/threads/cardioid-bass.121590/
https://www.linkwitzlab.com/models.htm#A1
https://www.diyaudio.com/community/threads/cardioid-bass.121590/
No? Do a sanity check: even with all filters disabled (shorted, Ctrl+S when selecting a filter) your model should be dipolar for the 20-500 Hz range.
Cardioid systems shape their directivity through destructive interference, a lot of energy get lost; they exchange efficiency for directivity, unlike horns.
There is a scrollbar below the charts that does exactly this. For some reason it is invisible until you hover your cursor over it or click it, and even then sometimes it takes couple of clicks or sweeps in the right spot.
@kimmosto could you please improve the visibility of the frequency slider? The best way I could think of is a slide rule, a semi-transparent slider over the numbered scale. This way user immediately see it is a frequency control element, not a regular scrollbar for dragging display left or right. Of course, if you have time and possibility for it.
The picture was from an earlier prototype from when I was scratching my head. I just didn't see a reason to repost what you'd already done. Good to know about the baloons, though! Wish I could wrap the polar heatmap circularly instead of on the line. It's so much easier for me to visualize that way.