With that in mind I've modelled a conical slot with almost no rear wall area and sufficient mouth size to keep port velocity below 17m/s at Xmax. For the sake of starting somewhere, the rear chamber is 27L which should add less than 5% air-spring distortion even at Xmax.
The resonant peak has been pushed further out of band as a result, though is still shown as fairly high. The strategies mentioned previously will help me to acoustically lower the peak greatly (even without EQ). Though I gather Hornresp might sometimes overestimate such things, and they are quite sensitive to small changes too. So, probably for someone of my limited ability, testing and measurement would be the best way to play sround with the possibilities.
This is a drawing of one possible way in which this might come together; shown are two subwoofer boxes:
The intention now is that these will simply be individual sealed boxes per driver, just with a slightly angled baffle.
They could then either be used as individual sealed boxes or they could be fixed together for dual opposed configuration. I suppose as individual boxes, they could also be pushed against a wall or have panel added to make a slot, but I increasingly think it unlikely that I'll want to use a slot if they stand individually. For me, the slot is still mainly just a way to make two opposing drivers both fire forward.
The resonant peak has been pushed further out of band as a result, though is still shown as fairly high. The strategies mentioned previously will help me to acoustically lower the peak greatly (even without EQ). Though I gather Hornresp might sometimes overestimate such things, and they are quite sensitive to small changes too. So, probably for someone of my limited ability, testing and measurement would be the best way to play sround with the possibilities.
This is a drawing of one possible way in which this might come together; shown are two subwoofer boxes:
The intention now is that these will simply be individual sealed boxes per driver, just with a slightly angled baffle.
They could then either be used as individual sealed boxes or they could be fixed together for dual opposed configuration. I suppose as individual boxes, they could also be pushed against a wall or have panel added to make a slot, but I increasingly think it unlikely that I'll want to use a slot if they stand individually. For me, the slot is still mainly just a way to make two opposing drivers both fire forward.
Last edited:
5% would put the 40Hz second harmonic ~26dB below the fundamental, but since the majority of low frequency musical notes have harmonics with levels near or greater than the fundamental, the disparity in perception makes sense.
36% at 30Hz, 3% at 40Hz in the 36Hz Fb box with ~35volts.
Sealed, 11% at 30Hz, 8% at 40Hz.
The sealed box was equal in output to the ported at 30Hz, -6.4dB at 40Hz.
I'd expect so, adjusted for voltage vs. impedance, though I'd expect the larger voice coil FABs to have a lot less power compression which can be an issue with the LABs in such tiny cabs.
At 49volts, the LABs were ~16% at 30Hz in ~100L, you might sim what the impedance is at that frequency.
You could just fix rectangular boxes together with a "V" slot, just angle the connector piece rather than the baffles and cabinet sides.
Fewer angle pieces makes construction easier!
Art
A familiar sight... It crosses my train of thought, with it being so close to perfect opposition and 0 resulting mechanical forces, I have stuck with the normal slot since the end solution in my case always leads to damping material. In the end, you can either have a FR like the one you've, shown, as a best case scenario, another best would be the rear baffle of the slot cupping the driver. Burying half the woofer is another option, but they all lead to FR's that need TLC from an EQ at the very least. Damping half the rear gets me here. I wonder about the scenario that Transients and Harmonics could cause an resonant air column to resonate. With Damping material causing a FR like below, not much to think about.
There is also the option of some type of phase plug.
Last edited:
Notice! Kev06 has plans for not aPush-Pull slot loaded subwoofer
Weltersys linked his old measurements that very well prove the reduction of even order harmonics. The problem is that odd order is more easily heard and it did not change.
So, did you hear a difference between those?
However with high spl in-room There are other mechanisms that can cause more distortion than the loudspeaker itself (Other ojects resonating). Here my PPSL 2x15" sub measured in-room. Nearfield measurements show typical profile unless the UMIK-1 starts clipping
Weltersys linked his old measurements that very well prove the reduction of even order harmonics. The problem is that odd order is more easily heard and it did not change.
So, did you hear a difference between those?
However with high spl in-room There are other mechanisms that can cause more distortion than the loudspeaker itself (Other ojects resonating). Here my PPSL 2x15" sub measured in-room. Nearfield measurements show typical profile unless the UMIK-1 starts clipping
Last edited:
Yes, plans quickly changed so the title of this thread became a little misleading, unfortunately. I slill like the idea of opposed drivers, when excursion is going to be large, but am now thinking push-push instead. That is partly to reduce slot size (given my limited space) and partly because the drivers have a significant vent at the magnet end, and I'd prefer any vent-related noises to be on the inside of the box.
Art, that is most helpful, thank you. All those real-world examples with similar drivers will certainly help to guide my targets and intentions here.
Yes, I think perception must be partly behind the wide range in opinions on what distortion is allowable. It seems that tests have used different material (everything from sine waves and single tones to a variety of music styles), at all sorts of SPLs and in different conditions etc etc. It has even been suggested that many listeners (to reproduced or reinforced music) are so used to hearing distortion that they don't particularly recognise it even at very noticeable levels.
For some context I've had a look at distortion measurements of different popular (home/home-theatre) subs, and they vary widely too. Some look to be pretty awful e.g. >60% THD at 20hz, and still over 20% up by 40hz (in fact some can't realistically reach down very low at all, so are even worse). But interestingly some are considerably better, like 15% THD at 20hz and 2% or so by 40Hz.
Of course, a lot relates to size/excursion/headroom relative to the SPL they're being tested at, and mine are quite modest. But I need to remember that in the case of any future larger rooms I'd be using four distributed subs, not just one, and (due to headroom requirements for movies) my normal intended listening levels would be at much less than their maximum output, too. So I think/hope I should be able to achieve something towards the better end of subwoofer distortion.
Yes, I think perception must be partly behind the wide range in opinions on what distortion is allowable. It seems that tests have used different material (everything from sine waves and single tones to a variety of music styles), at all sorts of SPLs and in different conditions etc etc. It has even been suggested that many listeners (to reproduced or reinforced music) are so used to hearing distortion that they don't particularly recognise it even at very noticeable levels.
For some context I've had a look at distortion measurements of different popular (home/home-theatre) subs, and they vary widely too. Some look to be pretty awful e.g. >60% THD at 20hz, and still over 20% up by 40hz (in fact some can't realistically reach down very low at all, so are even worse). But interestingly some are considerably better, like 15% THD at 20hz and 2% or so by 40Hz.
Of course, a lot relates to size/excursion/headroom relative to the SPL they're being tested at, and mine are quite modest. But I need to remember that in the case of any future larger rooms I'd be using four distributed subs, not just one, and (due to headroom requirements for movies) my normal intended listening levels would be at much less than their maximum output, too. So I think/hope I should be able to achieve something towards the better end of subwoofer distortion.
Thank you, it is good to be reminded of other possibilities.
The peak has been pushed reasonably high out of band, with the V-shape to the slot. But yes, I am also thinking of reducing the peak's magnitude (acoustically, not just electronically). A U-shape would spread it upwards in frequency and lower in magnitude, so seems sensible and relatively easy.
I'd thought that e.g. 400Hz would be quite low in frequency for damping to take much effect, but you have got very good results there, so that is definately very interesting to see. It could be better and perhaps even simpler than shaped slots. I can envisage a wedge of damping material in a kind of phase-plug shape to bring it forward without blocking the drivers, or (as their diaphagms are concave) maybe a lens-shape could be put in the middle of the slot between the drivers.
I really need to get something built and start real-world testing!
Thanks again,
Kev
The depth of the slot strongly affects the peak. a U shape cupping the drivers, if thats what you mean, is the shortest slot one could make with starting to cover the driver. I've thought about covering the driver but then you create another resonance between the driver and baffle covering it. @weltersys didn't seem fond of it lol.
I have thought about this, a triangle creating the conical like profile on each side... I dont think there is a penalty for covering half the driver with damping material but I agree there may be other routes of applying damping to good effect without covering the driver. If you go with a slot that creates no null, then I think you'd have a better chance since the null is the worst characteristic.
A proper phase plug would seek to take up all open space, I think.... It would try to occupy the space inside the cone, only allowing enough room for the cone to reach xlim, and then only enough CSA to not cause velocity related issues. Right Art?
Looking more into this, I found one of your threads, Camplo. If that was the same build, it seems that you used bigger drivers than me, with consequently a deeper slot, so the damping needed to cope with a slightly lower frequency peak. Yet, from your chart above, it has still worked extremely well, which is very encouraging.
I'm looking forward to playing around with this now; a bit more interesting than plain sealed designs anyway. It would actually be simple and neat to still make two rectangular sealed boxes; they could be joined (as Art suggested) with a wedge slot, if I wanted a wedge. But if a damped rectangular connecting slot is just as good (or better) it would make the engineering noticeably easier, and more likely to extract the most force cancelling from opposed drivers. I don't think I shall follow the threaded rod method, after your experiences though!!
I take what you say about the nulls being harder to compensate for. I actually haven't seen much of one appearing 'in band' with my hornresp sims, so I might be lucky with the dimensions. Though the peak might be masking a close-ish one; it certainly affects/raises the upper-end roll-off, so band-width could need some attention if I manage to reduce the peak too much.
EDIT: Though just to illustrate, here is a Parallel-slot (red) vs Conical-slot (grey) response chart. Same slot mouth, and about the same mouth velocity in both cases:
Though that is without any damping or packing to the back of the slots. It seems to me that a parallel slot might accomodate more damping material than a conical one, without 'blocking' the drivers too much, and would be combined with a U-shaped back to further reduce it by spreading out the remainder upwards.
(in my case I'm intending EQ so there is flexibility to flatten/tweak some aspects of the response shape)
I'm looking forward to playing around with this now; a bit more interesting than plain sealed designs anyway. It would actually be simple and neat to still make two rectangular sealed boxes; they could be joined (as Art suggested) with a wedge slot, if I wanted a wedge. But if a damped rectangular connecting slot is just as good (or better) it would make the engineering noticeably easier, and more likely to extract the most force cancelling from opposed drivers. I don't think I shall follow the threaded rod method, after your experiences though!!
I take what you say about the nulls being harder to compensate for. I actually haven't seen much of one appearing 'in band' with my hornresp sims, so I might be lucky with the dimensions. Though the peak might be masking a close-ish one; it certainly affects/raises the upper-end roll-off, so band-width could need some attention if I manage to reduce the peak too much.
EDIT: Though just to illustrate, here is a Parallel-slot (red) vs Conical-slot (grey) response chart. Same slot mouth, and about the same mouth velocity in both cases:
Though that is without any damping or packing to the back of the slots. It seems to me that a parallel slot might accomodate more damping material than a conical one, without 'blocking' the drivers too much, and would be combined with a U-shaped back to further reduce it by spreading out the remainder upwards.
(in my case I'm intending EQ so there is flexibility to flatten/tweak some aspects of the response shape)
Last edited:
That sounds about correct.
Dave Gunness/Fulcrum Acoustic's Oculus Phase Plug is a good example of a proper phase plug for combining the output of two woofers:
The same type of phase plug could be used with drivers facing each other.
That said, no need for phase plugs with subwoofers, the offset "V" loading pushes the peak/null well out of the pass band.
Vibration cancelling could be done by simply putting the drivers on opposite sides of the cabinet without the wasted space of the slot.
I don't see why one would bother with the slot mess in the first place, the additional mass loading is not worth the drawbacks.
Last edited:
As I just wrote, vibration cancelling could be done by simply putting the drivers on opposite sides of the cabinet without the wasted space of the slot.
I don't see why one would bother with the slot mess in the first place, the additional mass loading is not worth the drawbacks.
The major reason I see for slot loading is to reduce cabinet frontal area, or facilitate horn loading, or using push-pull to potentially reduce even order distortion.
None of those apply to Kev's push-push application.
I don't see why one would bother with the slot mess in the first place, the additional mass loading is not worth the drawbacks.
The major reason I see for slot loading is to reduce cabinet frontal area, or facilitate horn loading, or using push-pull to potentially reduce even order distortion.
None of those apply to Kev's push-push application.
For the third effing time (deja vu all over again..), "vibration cancelling" could be accomplished by putting the drivers on opposite sides of the cabinet without the wasted space of the slot, if all one cares about the vibration reduction result.
OK then its my fault, I'm not making myself clear apparently. Do you think the vibration cancellation effect of opposing woofers is worth ones time? In my situation where excursion is meant to be low, at least due to efficiency, if not due to domestic spl levels. I am wondering should this machinal vibration aspect even be on my radar. My boxes are large and heavy
For me, force cancelling is sensible when my subs are to be combined, since that will be for higher SPL and higher Xmax situatins. Keeping in mind the boxes will be relatively small, light and potentially placed in/on furniture that may rattle with vibration. In other situations, like big heavy or weighted cabinets sat on a robust floor it might not be worth the effort. If you have few issues with physical vibrations being transferred, I can't see that it would matter much.
As stated, the reason I'm looking at a slot is because the other way of opposing drivers, on opposite sides of the box, severely reduces options for placement; against furniture, in corners, in cupboards etc. The loading, such as it is, is not of interest to me. It is just a simple and compact way to make tge sub forward firing. At least, simpler and and more compact than a bandpass design, for instance. Though it has a few more complications than I'd envisaged initially; this is my first slot type project, so that is fine, I'm finding it interesting.
As stated, the reason I'm looking at a slot is because the other way of opposing drivers, on opposite sides of the box, severely reduces options for placement; against furniture, in corners, in cupboards etc. The loading, such as it is, is not of interest to me. It is just a simple and compact way to make tge sub forward firing. At least, simpler and and more compact than a bandpass design, for instance. Though it has a few more complications than I'd envisaged initially; this is my first slot type project, so that is fine, I'm finding it interesting.
Depends on what your time is worth, and if the sonic improvements outweigh the deficits.
The slot resonance is an obvious sonic degradation.
That said, I've used slot subs for the placement options they have afforded, and dealt with the resonance with damping.
Regardless of excursion, I have never found a well braced, heavy box to suffer audibly from vibrations, though their mechanical transmission of vibrations can cause objects to rattle or vibrate. That said, most spurious rattles and buzzes I've heard in a domestic situation have been from airborne, not mechanical origin- putting the sub on a vibration insulator or opposing cones makes no difference.
Yes the sonic degradation of the slot is the main concern for me; in an ideal world I would prefer the slot to have no effect at all. That is why I'm becoming so interested in the (acoustic) damping options. Substancially eliminating the resonance itself seems much better than pushing it higher or trying to offset with EQ (though these could still all be combined).
It is also why I'm keen to make the thing in two halves, which can be used without the slot when situations permit or don't need force cancelling. Okay, the two halves approach might comprimise the mechanical coupling a little, but I don't buy into the need for (e.g.) bracing magnets against each other. If the baffle is small and thick and the driver's basket good quality, you can get the 'vast' majority of force-cancelling simply transferred via the cabinet. Which does the job well enough for me, especially in the context of other things in the room resonating/vibrating at sufficient SPL.
Though it will all come down to testing, when I can build something. I have greatly enjoyed looking at the implications of things like slots and bandpasses, and have learnt a lot along the way (often thanks to people on here like Charlie and Art). But if the best answer (for my purposes) truly turns out to be a boringly simple sealed box, then that is what I shall make (four off). Any cabinet vibrations could at least be reduced by a heavy slab or something, which might be cheap enough to abandon when house-moves occur.
It is also why I'm keen to make the thing in two halves, which can be used without the slot when situations permit or don't need force cancelling. Okay, the two halves approach might comprimise the mechanical coupling a little, but I don't buy into the need for (e.g.) bracing magnets against each other. If the baffle is small and thick and the driver's basket good quality, you can get the 'vast' majority of force-cancelling simply transferred via the cabinet. Which does the job well enough for me, especially in the context of other things in the room resonating/vibrating at sufficient SPL.
Though it will all come down to testing, when I can build something. I have greatly enjoyed looking at the implications of things like slots and bandpasses, and have learnt a lot along the way (often thanks to people on here like Charlie and Art). But if the best answer (for my purposes) truly turns out to be a boringly simple sealed box, then that is what I shall make (four off). Any cabinet vibrations could at least be reduced by a heavy slab or something, which might be cheap enough to abandon when house-moves occur.
Mechanical cabinet vibrations are such a tiny fraction of the sub's excursion they are of no direct sonic consequence.
Pieces of carpet or Dynamat can decouple mechanical vibrations that may cause neighboring apartments to complain, and are a lot more move-friendly than slabs.
That said, a marble slab or thick piece of glass over a sub can make a nice coffee table.