Hi guys, I did some experiments this morning.
I don't like my current sealed subs, they don't play well with my OB setup, and they're so big I can't move them at the best place.
They sure bring the missing impact from OB, but most of the time it doesn't feel right, often I prefer turning them off.
Velocity VS pressure I guess? I couldn’t merge them perfectly (now sure I can maybe integrate them better, but I did everything I know).
The room is not treated other than with diffusors, and I really don't want to fall into this trap again.
So I wanted to try U & H frames, much smaller to optimise placement and/or closer to me to minimise the cancellation effect, and get an idea if it would worth redoing everything.
I have a bunch of ikea eket boxes around, and being 35x35x35 cm they're the perfect size for the drivers (ultimax 12").
So yesterday I cut a piece of mdf, glued it as a baffle and added 4 little corner braces just to better hold the baffle.
I also glued the box panels together, this woody version is not covered with plasticky stuff like other colors but with thin paper like material that can easily be sanded for better glueing (I don't think it's veneer).
Nothing very sturdy but just enough for the tests.
Quick measurements this morning, in between my mains, about 3m. in front, almost the middle of the room.
U and H frame compared, H with another empty box added in front gives +4db, I was surprised, it's such a small baffle.
And it's over the entire range, and even the U frame is not so bad.
(crossed at 80Hz/24db for all)
And once in hand I realised I could try pointed up, with the thick carpet and concrete below it's a sealed box:
Not bad! that is close to the level of one of the big box in almost the same place (it’s not here usually).
Less output below 30Hz but I can probably equalized that if needed:
They're so small I could build 4 and optimise their spots and/or mismatch open/close.
Could even make a new living room table there.
For now I will just build a second one and fully integrate them to listen.
If convinced I might even build big Hframes for my midbass (4x18"), and forget all about subs…
U frame is then a 35cm cube, about 30L. internally
H frame is 35x25x70,
I also tried the sub inverted mount in the « sealed » position, it didn't change anything, exactly the same response.
I don't like my current sealed subs, they don't play well with my OB setup, and they're so big I can't move them at the best place.
They sure bring the missing impact from OB, but most of the time it doesn't feel right, often I prefer turning them off.
Velocity VS pressure I guess? I couldn’t merge them perfectly (now sure I can maybe integrate them better, but I did everything I know).
The room is not treated other than with diffusors, and I really don't want to fall into this trap again.
So I wanted to try U & H frames, much smaller to optimise placement and/or closer to me to minimise the cancellation effect, and get an idea if it would worth redoing everything.
I have a bunch of ikea eket boxes around, and being 35x35x35 cm they're the perfect size for the drivers (ultimax 12").
So yesterday I cut a piece of mdf, glued it as a baffle and added 4 little corner braces just to better hold the baffle.
I also glued the box panels together, this woody version is not covered with plasticky stuff like other colors but with thin paper like material that can easily be sanded for better glueing (I don't think it's veneer).
Nothing very sturdy but just enough for the tests.
Quick measurements this morning, in between my mains, about 3m. in front, almost the middle of the room.
U and H frame compared, H with another empty box added in front gives +4db, I was surprised, it's such a small baffle.
And it's over the entire range, and even the U frame is not so bad.
(crossed at 80Hz/24db for all)
And once in hand I realised I could try pointed up, with the thick carpet and concrete below it's a sealed box:
Not bad! that is close to the level of one of the big box in almost the same place (it’s not here usually).
Less output below 30Hz but I can probably equalized that if needed:
They're so small I could build 4 and optimise their spots and/or mismatch open/close.
Could even make a new living room table there.
For now I will just build a second one and fully integrate them to listen.
If convinced I might even build big Hframes for my midbass (4x18"), and forget all about subs…
U frame is then a 35cm cube, about 30L. internally
H frame is 35x25x70,
I also tried the sub inverted mount in the « sealed » position, it didn't change anything, exactly the same response.
Ok so I've quickly glued a second one yesterday, tried today:
Even with the loss in efficiency the Spl is is similar to the big boxes from before, they were at 6/7m. now less than 3m.
But I will have to survey their excursions when loud, that will be easy they’re right in front.
They don't excite the room modes, maybe a tiny bit but much less than the big sealed.
As a result it feels more articulated, less one tone box.
I’ll keep them as is for few days and listen.
Maybe try stuffing them a bit, and adding an empty box in front to extend the tunnel.
Then I’ll try pointing up in sealed versions but different spots, to see if I can bring back some impact without losing on definition.
Happy it worked!
Even with the loss in efficiency the Spl is is similar to the big boxes from before, they were at 6/7m. now less than 3m.
But I will have to survey their excursions when loud, that will be easy they’re right in front.
They don't excite the room modes, maybe a tiny bit but much less than the big sealed.
As a result it feels more articulated, less one tone box.
I’ll keep them as is for few days and listen.
Maybe try stuffing them a bit, and adding an empty box in front to extend the tunnel.
Then I’ll try pointing up in sealed versions but different spots, to see if I can bring back some impact without losing on definition.
Happy it worked!
I got my new coffee table! 
Doubling and adding the tunnel were needed, because I was mistaken it’s not the same SPL as before.
Right now I added 12db to match the mains, the old subs were boosted too but maybe by 6db (but I re-tuned everything so not really comparable). I’d be confortable with 4 of them for loud listening.
Non-optimal placement since close to the middle of the room and I get a dip at 70hz, but they’re also useful as a table.
Anyway it sounds good, it’s still early but I feel it's an improvement.
Next, sealed, closer to my listening position, maybe behind me.
Doubling and adding the tunnel were needed, because I was mistaken it’s not the same SPL as before.
Right now I added 12db to match the mains, the old subs were boosted too but maybe by 6db (but I re-tuned everything so not really comparable). I’d be confortable with 4 of them for loud listening.
Non-optimal placement since close to the middle of the room and I get a dip at 70hz, but they’re also useful as a table.
Anyway it sounds good, it’s still early but I feel it's an improvement.
Next, sealed, closer to my listening position, maybe behind me.
Well I played a bit more with them, sealed at different position in close front & sides.
Tuned to the same response than in dipole mode, but couldn’t A/B them of course.
it’s not great. Sure I get more tactile effect but it’s also blurry, the room reacts too much, the deep low end stuff becomes one note.
Not better than the mains left unfiltered for example.
Surprisingly they were easier to eq than the dipole but with at the end only few dbs more.
And it’s still less low end than the big ol’boxes so it doesn’t make sense.
Put everything back in dipole right in front, yes it’s clean.
Maybe I’ll double them.
Tuned to the same response than in dipole mode, but couldn’t A/B them of course.
it’s not great. Sure I get more tactile effect but it’s also blurry, the room reacts too much, the deep low end stuff becomes one note.
Not better than the mains left unfiltered for example.
Surprisingly they were easier to eq than the dipole but with at the end only few dbs more.
And it’s still less low end than the big ol’boxes so it doesn’t make sense.
Put everything back in dipole right in front, yes it’s clean.
Maybe I’ll double them.
I should get a 3rd ultimax next month, I can only fit 3 in front, 4 would be perfect spl wise as I need +6db but 3 should be ok.
And I want to experiment wit cardioid, with one sealed in the middle in its tunnel too just for the look.
Like dipole/sealed/dipole in line, I’m really curious. Maybe a good combo.
They already sound really good now, just need more oomph, it is "dry" for sure, but I like it:
And I want to experiment wit cardioid, with one sealed in the middle in its tunnel too just for the look.
Like dipole/sealed/dipole in line, I’m really curious. Maybe a good combo.
They already sound really good now, just need more oomph, it is "dry" for sure, but I like it:
Regarding your first post, don't confuse U frame and H frame with cardioid and dipole. An undamped U-frame cabinet has similar on axis response as a dipole with similar length below the dipole peak. That is, the U behaves as a dipole at low frequency when undamped. When you placed the tunnel in front of the U-frame you doubled the dipole length and thus higher SPL. So in effect you car comparing two dipole of different length. But if you correctly damped the U-frame so it behaved as a cardioid at low frequency you would see a similar increase in the on axis SPL and the damped U and double length H would be at about the same level..
See http://musicanddesign.speakerdesign.net/u_frame.html for discussion in depth.
See http://musicanddesign.speakerdesign.net/u_frame.html for discussion in depth.
Hi john K, thank you for your attention, your website is definitely a source that inspired me to play with these!
I did play with H and U frames with these two subs, and didn’t hear any difference, in this room and placement.
I also tried dampening when in U frame, and I could measure slight variations but couldn’t hear the difference.
So I finally kept them both in simple H-frame version for the look, and practicality (as coffee table).
Now with 3 subs I will also try mixing sealed and H-frame, more for the experience and the fun, I don't remember where I read this but it could be a way to get a cardioid pattern.
Although I’m not sure that for only the first octave it would be noticeable.
So I will try 3 subs, in the same position and layout than the 2 up in this thread, but central one in sealed version, and both sides in Hframe. With dedicated dsp and amp channels I expect to be able to tune the response a bit.
I will add a front tunnel to the sealed one but just to keep the "one block" look and I don't expect any effect.
If not convincing I will then go back to H-frame for all 3.
It’s surprising how well the H frame version works, they may not be in the best placement in the room (I didn't search for the nulls), but for sure they are not in the worst one.
The room interaction is very different, it is still excited but it looks like it's different points than with the sealed, corners are still a bit too hot for example, but at my listening position it’s greatly reduced, in result the bass is much more articulated.
It’s not a setup that is easy to A/B, but I swear I hear more variations and precision in the low bass notes now.
It’s just much more satisfying than before.
I did play with H and U frames with these two subs, and didn’t hear any difference, in this room and placement.
I also tried dampening when in U frame, and I could measure slight variations but couldn’t hear the difference.
So I finally kept them both in simple H-frame version for the look, and practicality (as coffee table).
Now with 3 subs I will also try mixing sealed and H-frame, more for the experience and the fun, I don't remember where I read this but it could be a way to get a cardioid pattern.
Although I’m not sure that for only the first octave it would be noticeable.
So I will try 3 subs, in the same position and layout than the 2 up in this thread, but central one in sealed version, and both sides in Hframe. With dedicated dsp and amp channels I expect to be able to tune the response a bit.
I will add a front tunnel to the sealed one but just to keep the "one block" look and I don't expect any effect.
If not convincing I will then go back to H-frame for all 3.
It’s surprising how well the H frame version works, they may not be in the best placement in the room (I didn't search for the nulls), but for sure they are not in the worst one.
The room interaction is very different, it is still excited but it looks like it's different points than with the sealed, corners are still a bit too hot for example, but at my listening position it’s greatly reduced, in result the bass is much more articulated.
It’s not a setup that is easy to A/B, but I swear I hear more variations and precision in the low bass notes now.
It’s just much more satisfying than before.
The damping of the U-frame is critical to get a cardioid and as I have showed, it degrades as the frequency rises. The room interaction is differ net than a dipole as there is some room pressurization with a cardioid. So they can have a little more impact.
Yes, a dipole and a monopole, correctly aligned (acoustic centers) will yield a cardiod response. But they have to have the same axial response.
This idea can be used, for example, when crossing a dipole midrange to a monopole woofer if you use and LR type crossover. The response above the crossover is dipole (red), at the crossover point it is cardioid (green) and below the crossover point it morphs to a mono pole (yellow, then blue).
Yes, a dipole and a monopole, correctly aligned (acoustic centers) will yield a cardiod response. But they have to have the same axial response.
This idea can be used, for example, when crossing a dipole midrange to a monopole woofer if you use and LR type crossover. The response above the crossover is dipole (red), at the crossover point it is cardioid (green) and below the crossover point it morphs to a mono pole (yellow, then blue).
They will be aligned but horizontally only, facing me but I could play with:
- the placement of the sealed baffle, since the added tunnel would vary the distance from the other two H baffles by +/- 35cm.
That seems small compare to the wavelength but maybe there will be impact.
I don't expect any resonance issue with the tunnel on the sealed, it's crossed so low acoustically at 70Hz.
- phase and time delay
something like,
All H frames:
With sealed in the middle, all baffles on same plan, facing me:
or facing rear wall, phase inverted, I assume here the cardioid effect would also be reversed
Or playing with baffle placement since they’re all cubic and easy to move around.
I might also try inverted mount of the sub if it asks for a little more volume.
I should receive the 3rd one soon so I will come back with test results.
- the placement of the sealed baffle, since the added tunnel would vary the distance from the other two H baffles by +/- 35cm.
That seems small compare to the wavelength but maybe there will be impact.
I don't expect any resonance issue with the tunnel on the sealed, it's crossed so low acoustically at 70Hz.
- phase and time delay
something like,
All H frames:
With sealed in the middle, all baffles on same plan, facing me:
or facing rear wall, phase inverted, I assume here the cardioid effect would also be reversed
Or playing with baffle placement since they’re all cubic and easy to move around.
I might also try inverted mount of the sub if it asks for a little more volume.
I should receive the 3rd one soon so I will come back with test results.
Some news, before more mods on the system I tried some of the previously shown combos.
Honestly using one sealed didn't change anything whatever its place, I could barely see the change in measurement and even less by ear.
So I went with 3 H frames as is:
I have not measured the voltage out of the amp before on the previous 2 H-frames, so I still didn't know what to expect on the power needs.
I assumed much less than in boxes but without a clear idea I wired them in parallel/serie (they are UM12-22 so dvc).
Here are some plots for one driver, in // or serie:
So between 1 and 4 ohms in //, and from around 4 to 14 ohms in serie in the bandpass used: 20-65hz.
I selected a serie of 3 in parallel, for final 3-12 ohms load.
Kept them like that few weeks but I had a lot of noise from the motors.
Yesterday I moved and rewired the mains a bit, in the same time rewired the 3 subs VC all in serie to test.
So in theory something like 12 to 40+ohms load for the amp.
Retuned a bit but it didn't need much change, I was surprised, just +2db for the same response.
And the motors are more quiet, at least from listening position, surprising too for me but there must be an explanation.
Anyway today I finally measured the voltage out, to see if I could replace this amp for something quieter.
(also to check if the little FOSI lc30 could survive in its path to get needles fun, this device is limited to 250w).
What a surprise, for 50Hz, I got 1.7v at -30db, 5.4v at -20db, 17v at -10db, and I didnt measure above.
This gives 0.2w / 2w / 20w / 200w at -0db full potential, where I have never been.
Loud means around -15db for me, and very rarely go near -10db with music.
It turns out I could even use a tube amp here if I stay safe with the volume, or at least something more classic/smaller and silent.
I knew they would need less power, if only by their excursion limit (that I didn't simulate).
But still it's a shock, and opens a lot of possibilities for a future amp!
Edit:
Oh wait, I’m an idiot...
Yes 20-200w only, but at 12ohms!
Silly me, I was so excited to look out for other amps, forgot any regular amp will not be able to do the same.
Honestly using one sealed didn't change anything whatever its place, I could barely see the change in measurement and even less by ear.
So I went with 3 H frames as is:
I have not measured the voltage out of the amp before on the previous 2 H-frames, so I still didn't know what to expect on the power needs.
I assumed much less than in boxes but without a clear idea I wired them in parallel/serie (they are UM12-22 so dvc).
Here are some plots for one driver, in // or serie:
So between 1 and 4 ohms in //, and from around 4 to 14 ohms in serie in the bandpass used: 20-65hz.
I selected a serie of 3 in parallel, for final 3-12 ohms load.
Kept them like that few weeks but I had a lot of noise from the motors.
Yesterday I moved and rewired the mains a bit, in the same time rewired the 3 subs VC all in serie to test.
So in theory something like 12 to 40+ohms load for the amp.
Retuned a bit but it didn't need much change, I was surprised, just +2db for the same response.
And the motors are more quiet, at least from listening position, surprising too for me but there must be an explanation.
Anyway today I finally measured the voltage out, to see if I could replace this amp for something quieter.
(also to check if the little FOSI lc30 could survive in its path to get needles fun, this device is limited to 250w).
What a surprise, for 50Hz, I got 1.7v at -30db, 5.4v at -20db, 17v at -10db, and I didnt measure above.
This gives 0.2w / 2w / 20w / 200w at -0db full potential, where I have never been.
Loud means around -15db for me, and very rarely go near -10db with music.
It turns out I could even use a tube amp here if I stay safe with the volume, or at least something more classic/smaller and silent.
I knew they would need less power, if only by their excursion limit (that I didn't simulate).
But still it's a shock, and opens a lot of possibilities for a future amp!
Edit:
Oh wait, I’m an idiot...
Yes 20-200w only, but at 12ohms!
Silly me, I was so excited to look out for other amps, forgot any regular amp will not be able to do the same.
Last edited: