I've modeled a different slot loaded enclosure... The 21" on the bottom and 18" on top. 2.83v
I still wish there were more scientific data on this idea of machinal vibration cancellation... If its so great or necessary, why don't we see it else where in studio products?
I am starting to look at the removal of excursion as another source of reduction to mechanical vibration. I think, maybe its overkill when I can keep excursion below a certain margin.
One of the reasons why I sought to delete the 15" from the equation is because in the situation it was in, little off center of the living room and me sitting 54" away..... The ppsl had the better FR, as a result of being closer to the floor, I assume. The room caused more FR distortion at the position of the 15".
In combination, the 15" would only worsen the FR, correct? From there I committed to the idea of FR quality. I think I am right to believe that a woofer closest to the floor will have the better FR? Its one step away from being infinite baffled into the floor? Its like a pseudo-corner load even. So yeah that happened. Will the 15" in the same height but corner loaded, have the same problem in my basement? I have not tried it yet. But according the room sim in VituixCad, the woofer by the floor wins.
Still waiting on Acoustic Elegance to see if he will fix the surround, or if I should attempt it.
This is not automatic, these actions are taken by the team.. so forego the logic in this case
By 'accents', I was referring to bold/underline originally used in post #12,679 and others.
If you have any further questions, please take it to PM, or alternately report a post for review.
I guess one way to put it is... In multi subwoofer approach, We say that the subs combined FR is better, due to one woofer improving an area where another is suffering. The other aspect is that the woofer that is suffering will degrade the same area of FR the other woofer is performing correctly in.
In the instance of a woofer stacked vertically over another, you are only changing position on height Axis, and in my room, the other 2 axis's remain the same.
This makes me wonder, maybe there is merit to spacing drivers on the left and right within certain parameters. Facing diagonally within front and side walls would give variation 2 out of 3 axis's.
2 - 18"s next to each other should just barely fit in 1/4WL spacing.
This makes me think, it sounds feasible that if you had a sub with woofers on all panels... if they are all within 1/4WL spacing of the passband, you can treat the source as one, only needing one delay in time alignment.
In the instance of a woofer stacked vertically over another, you are only changing position on height Axis, and in my room, the other 2 axis's remain the same.
This makes me wonder, maybe there is merit to spacing drivers on the left and right within certain parameters. Facing diagonally within front and side walls would give variation 2 out of 3 axis's.
2 - 18"s next to each other should just barely fit in 1/4WL spacing.
This makes me think, it sounds feasible that if you had a sub with woofers on all panels... if they are all within 1/4WL spacing of the passband, you can treat the source as one, only needing one delay in time alignment.
Yes, this is true and exactly what one doesn't want to do. One wants more independent sources at LFs not less.
The huge peak/dip that results from slot loading has hardly changed. If used as a subwoofer, (<100Hz) those peaks are not much of a problem.
Mechanical vibration effects can be reduced to an acceptable (virtually inaudible) level with bracing and damping.
The cancellation dip that results from slot loading can't be reduced by any means.
There are studio products that use high mass subwoofers that do use opposed drivers for reduction of mechanical vibrations, but none that I know of are using slot loading.
Your front loaded 15" frequency response is far better (more linear)in the 200-1000Hz range than the dual driver slot load, whether "in combination" or separately.
Boundary reflections will cause peaks and nulls dependent on the distance relationship between the acoustic point of origin, the boundary, and the listening position.
https://mehlau.net/audio/floorbounce/
At close distances the inverse distance loss and a higher directivity index increase the direct to reflected ratio, reducing the depth of cancellation.
A woofer has a frequency response.
The location of the woofer in relation to boundaries (floor, corner, any wall) will affect the measured response.
Whether the peaks and nulls at a particular listening location are "better" depends on what range you look at.
Take a look at Genelec's latest, 8381a.
My take on vibration reduction is you need it in two situations....
one, mechanical vibration making noise coupling to floor, or with main speaker stacked on sub
Two, sub is trying to walk around. Common with subs with real balls,.... when crankin.
Slot loaded, imho, is not worth the bother other than it makes for a great way to protect drivers with a very simple grill. Pet proof, inadvertent puncture protection.
I do not think slot load helps with reducing distortion.
And must say, a slot-loaded sub is probably one of the least favorable designs, if trying to take a CD all the way down to the sub. Simply due to a slot-loaded's limited upward frequency extension.
The idea of a ripole's, front and back slots' dipole radiation, is interesting however...
In combination with room, unfortunately the 15's response was not better. In situ, the FR of the Sub, was better than the 15". The design is based around an ear height tweeter, and full spectrum mains. The flaw is tied to floor, ceiling bounce and at least one of the cancellations won't change due to the height of the mid, which is at where its supposed to be.... So if it can't live where its supposed to, it cannot live. The floor level woofer has the best FR. I can't cover 200-20hz with the 15" so no reason to "reposition" it there. Theres another dip around 80hz from the ceiling, I think, for the floor level sub.
I never had a part of the designing process that was aimed at designing driver heights according to crossovers versus predictable floor/ceiling reflections... Thats my fault.
It ends up being this; I can't have a woofer positioned where a null is passing close by or through, the crossband.
Thank you for sharing that link, these results prove what I have said to be true. The height of my sub woofers put a cancellation well out of the passband
I just needed to move the null, that was the only issue, nothing else is an issue in the current design except there a null in the middle of the crossband band. The "peak" is not an issue, even in the Horn resp sim, the decay times where not an issue, in particular, after designing a HF roll off. After that, the decay is no existent. There is a GD spike, at that peak, but after EQing out the peak the GD is gone with it.
There is another approach I haven't entertained. Damping material in the slot, very simple in HR but in real life? Never tried.
I'm not sure what number to use in Fr, but I think you get the jist.
Last edited:
You found boundary reflection cancellations at 54"... What about the one meter (39") listening location you wrote you had decided upon?
So it seems that unlike your avatar picture, you separated the woofer and "tweeter" (Axi2050/horn), putting the tweeter at (seated) ear height, the woofer on the floor and decided that's where its supposed to live?
As your sim indicated, damping material in the slot will reduce the depth of the dip and peak.
Mineral wool (Rockwool, GeoWool etc.) would be fairly easy to cut and friction fit in the slot.
Attachments
In the picture, The top of the sub cabinet is 29" Desired HF axis height is 48"... This is ball park scale accurate (+/- a few inches)
A tweeter height of 42"-48" would create a distance of 33"-39" CTC.
Its either that or raise the woofers off the floor. Lose the good mode performance or lose the 1/4WL spacing. If I raise the Tweeter up to said height, allow it to rest the same way as it rest's on the sub, the tweeter is slightly angled up. I could enjoy a slightly lower tweeter, this way, other wise I have to point the tweeter by raising the rear of the horn some how. I imagined the pictured configuration and thought.... There's room for another woofer. I wasn't thinking about how I got to this point, nor did I keep this cancellation business in mind, at that time.
There is a floor cancellation that I need to avoid. I can see 42-48" being the tweeter height. The tweeter angle needs to be adjustable.
So, in the first instance of how I ended up with up you see in the unedited picture. I did not consider the accumulated output. Then when I thought about where I really wanted the tweeter to be, and the open space between, the horn and the driver, I thought, I could put a driver there, it will help to bridge the distance to the lowest woofer. Then there is the accumulated FR that I didn't study enough.
The real issue I was seeing is that bandwidth is actually a topic of SQ. I think one should cover as much bandwidth as possible, per acoustical axis, while high passing the areas that would cause issues with constant directivity. If I move to 3 acoustical points, TMM, I'm suggesting that the middle axis should cover as much Bandwidth as the lowest, in my case. The lowest woofer is covering, 200hz-Sub, no issues. So the middle woofer should compliment is what I am getting at. My two lower Axis should be able to maintain 1/4WL ctc spacing, through the passband, it would just be a matter of execution at this point. If I can keep both my lower Axi's within 1/4WL of each other, I could treat them as a single source, they are crossed over at a low enough KA to keep things minimum phase even at the most extreme angle, had from the lowest woofer.
Thats my thoughts so far... I think about 36inches sounds right for box height.
The above is to show what happens when I fill the slot to the middle of the current 18"s. I looked at this before building the box, just never took a chance to see how real life would play out, yet. Keeping the filling towards the back, leaves the lf un attenuated according to the sim.
This, below, looks like the performance I might have in the basement from the midwoofer. Somewhere between 100hz-200hz there will be a null, as it were upstairs. I can get a little better result if I finish the baffle sim in the program but really, is the two way version that bad>? Seems to have potential for a better FR at least. There is always the sacrifice of keeping the horn as I had it previously, sitting on the 26" box putting the horn center at about 36" slightly pitched upwards
A tweeter height of 42"-48" would create a distance of 33"-39" CTC.
Its either that or raise the woofers off the floor. Lose the good mode performance or lose the 1/4WL spacing. If I raise the Tweeter up to said height, allow it to rest the same way as it rest's on the sub, the tweeter is slightly angled up. I could enjoy a slightly lower tweeter, this way, other wise I have to point the tweeter by raising the rear of the horn some how. I imagined the pictured configuration and thought.... There's room for another woofer. I wasn't thinking about how I got to this point, nor did I keep this cancellation business in mind, at that time.
There is a floor cancellation that I need to avoid. I can see 42-48" being the tweeter height. The tweeter angle needs to be adjustable.
So, in the first instance of how I ended up with up you see in the unedited picture. I did not consider the accumulated output. Then when I thought about where I really wanted the tweeter to be, and the open space between, the horn and the driver, I thought, I could put a driver there, it will help to bridge the distance to the lowest woofer. Then there is the accumulated FR that I didn't study enough.
The real issue I was seeing is that bandwidth is actually a topic of SQ. I think one should cover as much bandwidth as possible, per acoustical axis, while high passing the areas that would cause issues with constant directivity. If I move to 3 acoustical points, TMM, I'm suggesting that the middle axis should cover as much Bandwidth as the lowest, in my case. The lowest woofer is covering, 200hz-Sub, no issues. So the middle woofer should compliment is what I am getting at. My two lower Axis should be able to maintain 1/4WL ctc spacing, through the passband, it would just be a matter of execution at this point. If I can keep both my lower Axi's within 1/4WL of each other, I could treat them as a single source, they are crossed over at a low enough KA to keep things minimum phase even at the most extreme angle, had from the lowest woofer.
Thats my thoughts so far... I think about 36inches sounds right for box height.
The above is to show what happens when I fill the slot to the middle of the current 18"s. I looked at this before building the box, just never took a chance to see how real life would play out, yet. Keeping the filling towards the back, leaves the lf un attenuated according to the sim.
This, below, looks like the performance I might have in the basement from the midwoofer. Somewhere between 100hz-200hz there will be a null, as it were upstairs. I can get a little better result if I finish the baffle sim in the program but really, is the two way version that bad>? Seems to have potential for a better FR at least. There is always the sacrifice of keeping the horn as I had it previously, sitting on the 26" box putting the horn center at about 36" slightly pitched upwards
Last edited:
https://www.ebay.com/itm/233172783962
This one has low Qes
https://loudspeakerdatabase.com/18Sound/21LW1400
This one has low Qes
https://loudspeakerdatabase.com/18Sound/21LW1400
Last edited:
Hi Rob, I'm not at issue with CTC, the large sizes create the headroom and the low crossover makes for good ctc, so far.
Looking at floor and ceiling reflections versus dimensions in my place, it appears that at my desired listening distance there are predictable nulls in the FR. It seems that there is no place between the floor woofer and the horn the doesn't have this null between about 100hz and 200hz but then the horn and floor woofer avoid either null zone via crossover or by placement on the baffle.
Im not sure that I need or want to place another driver on the baffle because of this aspect.
I need to experiment more, but, in the name of smooth FR, I can choose to not add another driver, leave the horn lower than initially desired or raise the horn to high ctc spacing than desired.
In my basement, with a low ceiling, the issue is worse, I think... I have not gotten that far, but the Vituixcad sim says predicts similar behavior. The issue with the floor wouldn't change anyway, I guess.
Eventually I'll start to look into what can room treatment and corrective eq do for the issue.
Looking at floor and ceiling reflections versus dimensions in my place, it appears that at my desired listening distance there are predictable nulls in the FR. It seems that there is no place between the floor woofer and the horn the doesn't have this null between about 100hz and 200hz but then the horn and floor woofer avoid either null zone via crossover or by placement on the baffle.
Im not sure that I need or want to place another driver on the baffle because of this aspect.
I need to experiment more, but, in the name of smooth FR, I can choose to not add another driver, leave the horn lower than initially desired or raise the horn to high ctc spacing than desired.
In my basement, with a low ceiling, the issue is worse, I think... I have not gotten that far, but the Vituixcad sim says predicts similar behavior. The issue with the floor wouldn't change anyway, I guess.
Eventually I'll start to look into what can room treatment and corrective eq do for the issue.
This is the small print for xmax on the above driver
@b_force @GM
I haven't gotten that book yet but I do have a small knowledge base on this and along with the forums help, I should be able to pick some woofers, without reading the book first. I started comparing BL to mass, this is when I seen a new perspective as I noticed the BL to mass ratio was highest on certain woofers, like the M2 Woofer.... 18H+ is up there as well. So from there I was able to start organizing the loudspeaker database in a better way to expose this trait, and thats how I found the Oberton. I think Qes, Bl to mass ratio, and Le affect transient response, either can be a bottleneck. A driver like the JBL where the BL to mass ratio is high but the Qes isn't low confuses me..
If I step away from the BL to mass idea, this Celestion is decent, pretty much a cleaned up version of the Dayton 21"
The Lase is best bargain but is this real? I mean like, is this a trusted brand?
@b_force @GM
I haven't gotten that book yet but I do have a small knowledge base on this and along with the forums help, I should be able to pick some woofers, without reading the book first. I started comparing BL to mass, this is when I seen a new perspective as I noticed the BL to mass ratio was highest on certain woofers, like the M2 Woofer.... 18H+ is up there as well. So from there I was able to start organizing the loudspeaker database in a better way to expose this trait, and thats how I found the Oberton. I think Qes, Bl to mass ratio, and Le affect transient response, either can be a bottleneck. A driver like the JBL where the BL to mass ratio is high but the Qes isn't low confuses me..
If I step away from the BL to mass idea, this Celestion is decent, pretty much a cleaned up version of the Dayton 21"
The Lase is best bargain but is this real? I mean like, is this a trusted brand?
I don't see how its missed in this study, this huge overhang. I can't see GD in the IR but... in my experience, low motor control increases GD in bass or where ever resistance is and has worse Decay performance.
If you can add rigidity without the mass, I think performance is better. Increasing mass without increasing BL, worsens transient performance. The issue is creating a cone that is rigid enough to hold shape at high SPL. This is the reason why we see lower distortion at high SPL with more mass. If you null out cone Material, I believe this will hold true. By creating variance in cone material and size, as you compare you lose the relationship of mass to rigidity so the obvious is not obvious anymore. Using the same motor, cone material and size, as the cone gets thicker, it increases rigidity. At higher spl the cone is less likely to misshapen. At lower velocities the higher energy storage will lose to the lighter cone. Nulling as many variables as possible exposes the truth but when comparing cones that are made from different materials and geometry, thats complicating everything.
Last edited:
@GM, please allow to respectfully play devils advocate.
Can you elaborate? The above graph shows that adding mass literally distorts the transient, more.....
Why are we ignoring the overhang? Transient performance and HF performance may be connected but aren't literally the same thing. Just because something has no issue taking off from a resting position doesn't take away from the what it goes into motion. Impulse test are inconclusive evidence for this topic.
If I simulate this in horn resp, GD will be higher in the LF
Above is a driver in a sealed box and on the right I have increases mass from 74g to 300g
What I see is that anywhere there is resonance, transient performance is worse. The areas above and below resonance will have better transient performance.
The poorer the performance of the carrier signal, the worse the modulating signals will be.
Le= low pass filter
Mass= low pass filter
The only difference is that mass can be counteracted by motor force, but back to this image, the Blue Impulse is worse, than the Red one. Take this same test at 80db and another at 115db and then we will really see the truth, even more exaggerated. I don't believe some are interpreting the data correctly. velocity affects transient performance.... The test isn't even volume matched.... The Green and the Blue are worse performers than Red.
The initial attack from rest, is identical of the Red and Blue, at this volume, but I am saying that the overhang will be an issue for an modulating frequency that resides witrh the cycle time of the overhang. Clean transients not exist outside of the over hang. The worse the overhang, the worse the symptoms. Adding mass increases overhang.
Can you elaborate? The above graph shows that adding mass literally distorts the transient, more.....
Why are we ignoring the overhang? Transient performance and HF performance may be connected but aren't literally the same thing. Just because something has no issue taking off from a resting position doesn't take away from the what it goes into motion. Impulse test are inconclusive evidence for this topic.
If I simulate this in horn resp, GD will be higher in the LF
Above is a driver in a sealed box and on the right I have increases mass from 74g to 300g
What I see is that anywhere there is resonance, transient performance is worse. The areas above and below resonance will have better transient performance.
The poorer the performance of the carrier signal, the worse the modulating signals will be.
isn't this true for resonances sources else where as well, like from the cabinet? It would seem so.
Le= low pass filter
Mass= low pass filter
The only difference is that mass can be counteracted by motor force, but back to this image, the Blue Impulse is worse, than the Red one. Take this same test at 80db and another at 115db and then we will really see the truth, even more exaggerated. I don't believe some are interpreting the data correctly. velocity affects transient performance.... The test isn't even volume matched.... The Green and the Blue are worse performers than Red.
The initial attack from rest, is identical of the Red and Blue, at this volume, but I am saying that the overhang will be an issue for an modulating frequency that resides witrh the cycle time of the overhang. Clean transients not exist outside of the over hang. The worse the overhang, the worse the symptoms. Adding mass increases overhang.
Last edited:
I'm starting to feel like looking at Fs as a function of Q manipulating the roll off of the True system resonance, which is where ever sensitivity has peaked.
Here, Q has changed but LF below Fs has not. In this view it is obvious that the mass has lowered the HF sensitivity, which is what we also see with raising Le.
( same graphs from above superimposed, 74g and 300g everything else the same)
Here, Q has changed but LF below Fs has not. In this view it is obvious that the mass has lowered the HF sensitivity, which is what we also see with raising Le.
( same graphs from above superimposed, 74g and 300g everything else the same)
