I'm getting keen on building one of the above-titled subwoofers (for mainly music with occasional films/movies), I believe some might call the design a sealed 'PPSL'. The intention is for it to be usable in future as a normal room subwoofer, and if successful it'll be part of a multi/distributed setup of three or more. But in the first instance my use case is for a very near-field listening situation, and an attraction is in the single/small slot area from which the sound eminates. This could be located quite neatly close to the listener's head/ears - the aim being partly for the direct sound to dominate room modes and reflections and partly to minimise SPL needed which again reduces excitation of room modes (and complaining neighbours).
I already have drivers that I want to use, similar to the classic LAB12 but in this case called the FAB12 from BK electronics over here. Looking around at LAB12's use in front loaded horns, an Sd:horn-area ratio of 2.3:1 seems to be considered a reasonable compression ratio for them to handle. A slot is not quite a horn, but I shall let that ratio guide me as to what is sensible. In terms of slot length, the shorter the better for pushing reflections up above the intended frequency band; due to driver diameter that would be about 320mm or 12.5" as a minimum. I modelled one driver firing into such a slot, with a 25L rear chamber and enough power (~600w) to reach Xmax:
Hopefully that is right; I modelled it as an offset driver in a straight sided 2-section conical 'horn', though with a slight bulge towards S2 due to the driver cones being concave. That result is without any EQ - in practice I'd want to chop off most of the top of the curve to make it more flat from about 25hz to 80-100hz (the intended range). This will be an active system with PC as a source so such adjustments need not be costly. After which, I'd be left with a maximum of >100dB with just the one driver, in a dual driver situation that would become >106dB at 1m. Way more than I need for initial purposes of close listening, at say less than 30cm/12", but good to know for other/future uses.
It would be convenient to have the sub positioned in/on shelves immediately behind the listening chair, in order to locate the slot's outlet at ear height. Again as a future option I'd like the sub to also fit in some cupboards I have, such as under my TV and front speakers (Ikea 'Besta's). As it happens, the modelled slot volume plus rear chamber volumes do conspire to fit in such a cupboard if arranged appropriately and the slot facing towards the front. Here is a rough conceptual plan to illustrate:
Reasonably large and heavy of course, to be expected with 2x 12" drivers really, but it is all relative; a lot of DIY subs are enormous. Having done some tests I believe that I could move it around okay, anyway. Probably I could also take the drivers out for house-moves. I may even make the front-baffle+slot one piece that is removable for easy access to the drivers; the remaining cabinet could then double as a storage box, too. Obviously, that would involve some attention to engineering, to be successful.
I had initially considered building this sub in two halves, perhaps to use just one half for very nearfield purposes, and somehow bolting them together for when higher SPL and opposed drivers were needed. However that adds complication and I've seen a few comments around the internet which suggest things are noticeably better (with PPSL designs) if the two drivers share the rear chamber rather than having one half-sized chamber each. I don't know if anyone here has experience of that?
cheers
Kev
I already have drivers that I want to use, similar to the classic LAB12 but in this case called the FAB12 from BK electronics over here. Looking around at LAB12's use in front loaded horns, an Sd:horn-area ratio of 2.3:1 seems to be considered a reasonable compression ratio for them to handle. A slot is not quite a horn, but I shall let that ratio guide me as to what is sensible. In terms of slot length, the shorter the better for pushing reflections up above the intended frequency band; due to driver diameter that would be about 320mm or 12.5" as a minimum. I modelled one driver firing into such a slot, with a 25L rear chamber and enough power (~600w) to reach Xmax:
Hopefully that is right; I modelled it as an offset driver in a straight sided 2-section conical 'horn', though with a slight bulge towards S2 due to the driver cones being concave. That result is without any EQ - in practice I'd want to chop off most of the top of the curve to make it more flat from about 25hz to 80-100hz (the intended range). This will be an active system with PC as a source so such adjustments need not be costly. After which, I'd be left with a maximum of >100dB with just the one driver, in a dual driver situation that would become >106dB at 1m. Way more than I need for initial purposes of close listening, at say less than 30cm/12", but good to know for other/future uses.
It would be convenient to have the sub positioned in/on shelves immediately behind the listening chair, in order to locate the slot's outlet at ear height. Again as a future option I'd like the sub to also fit in some cupboards I have, such as under my TV and front speakers (Ikea 'Besta's). As it happens, the modelled slot volume plus rear chamber volumes do conspire to fit in such a cupboard if arranged appropriately and the slot facing towards the front. Here is a rough conceptual plan to illustrate:
Reasonably large and heavy of course, to be expected with 2x 12" drivers really, but it is all relative; a lot of DIY subs are enormous. Having done some tests I believe that I could move it around okay, anyway. Probably I could also take the drivers out for house-moves. I may even make the front-baffle+slot one piece that is removable for easy access to the drivers; the remaining cabinet could then double as a storage box, too. Obviously, that would involve some attention to engineering, to be successful.
I had initially considered building this sub in two halves, perhaps to use just one half for very nearfield purposes, and somehow bolting them together for when higher SPL and opposed drivers were needed. However that adds complication and I've seen a few comments around the internet which suggest things are noticeably better (with PPSL designs) if the two drivers share the rear chamber rather than having one half-sized chamber each. I don't know if anyone here has experience of that?
cheers
Kev
Many "PPSL" are push-pull (the front of one woofer facing the plenum, the rear of the other facing the plenum), sharing a back chamber makes construction simpler, as the chambers would otherwise need to be made different size due to the chamber volume occupied by the driver being negative for the "pull" side, and positive for the "push" side.
I have made push-push (like your proposal) using separate chambers, which are better in terms of box stiffness.
The "compression ratio" is not a problem, the slot "horn" has unloaded well above the subwoofer pass-band.
The plenum can be made as small as possible without exceeding a turbulent velocity, somewhere around 17 meters per second (5% of the speed of sound).
The extra volume this affords in the rear chamber reduces the power needed to reach Xmax, reducing power compression.
In your simulations, "ATC" (Area of Throat Chamber) can be used to account for the concavity of the cones.
Superb, thank you, Art! Yes, a smaller slot would be very handy, so if it isn't an issue then I'll certainly reduce its cross-sectional area. That could be really quite a useful reduction in total volume.
I had started out with push-pull but of course that made the slot even bigger, too big for what is supposed to be compact box, so my thread title is already out of date; I suppose mine would now be called push-push, if direction is related to the driver orientation. I've seen suggestions that a reversed driver can help reduce some slot artefacts and in PA type use it can help cool one of the motors, but this is for home use and my drivers have quite significant vents which could make noise; so pointing them both into the rear chamber instead of the slot could be good in that respect.
I also like your thoughts on how this affects symmetry, with respect to sharing or dividing the rear chamber. I couldn't think of a convincing reason why two half chambers should be significantly different to one big one, but by that point I was no longer reversing a driver; I'll bet the comments I saw would relate to where a driver is reversed. As mine will be made identical, dividing the chamber shouldn't matter; the cones should push symmetrically. So I could actually build this as two separate boxes which could be used together (opposing) or individually. More complexity, but also easier to move around and offers more flexibility - a key aim of this little project.
The tip on 'ATC' is also very useful; I'm not yet very experienced with hornresp, so all such pointers help. I find it a lot less intuitive than WinISD but mostly because it is far more capable and complex (not to mention well maintained and less buggy). It really is very good so far, so time to move on and learn more about it.
Thanks again, this project is going to be better as a result!
Kev
I had started out with push-pull but of course that made the slot even bigger, too big for what is supposed to be compact box, so my thread title is already out of date; I suppose mine would now be called push-push, if direction is related to the driver orientation. I've seen suggestions that a reversed driver can help reduce some slot artefacts and in PA type use it can help cool one of the motors, but this is for home use and my drivers have quite significant vents which could make noise; so pointing them both into the rear chamber instead of the slot could be good in that respect.
I also like your thoughts on how this affects symmetry, with respect to sharing or dividing the rear chamber. I couldn't think of a convincing reason why two half chambers should be significantly different to one big one, but by that point I was no longer reversing a driver; I'll bet the comments I saw would relate to where a driver is reversed. As mine will be made identical, dividing the chamber shouldn't matter; the cones should push symmetrically. So I could actually build this as two separate boxes which could be used together (opposing) or individually. More complexity, but also easier to move around and offers more flexibility - a key aim of this little project.
The tip on 'ATC' is also very useful; I'm not yet very experienced with hornresp, so all such pointers help. I find it a lot less intuitive than WinISD but mostly because it is far more capable and complex (not to mention well maintained and less buggy). It really is very good so far, so time to move on and learn more about it.
Thanks again, this project is going to be better as a result!
Kev
So, now thinking about 'chuffing' being the limiting factor for slot dimensions. Which begs the question of how loud is satisfactory. I am helped here by the fact that slot wind noise will rise and fall in step with sound output; i.e. it will diminish greatly during quiet, low excursion periods. So whilst people might aim for very low static noise floors, I believe slot noise could be considered more like a kind of relative distortion where perhaps 45-50dB below the music might be considered a safe target. I'm also not going to be routinely listening at anywhere near Xmax (that is for e.g. parties and explosions in movies) so things like air turbulence around sharp slot edges are going to be less audible when listening normally/critically.
At Xmax at 30hz, I can expect around 105dB output at 1m from a driver. 45-50dB quieter would therefore suggest a maximum of 55-60dB at 1m for slot wind noise. The Engineering Toolbox has a page which can help to quantify airflow velocity and the noise it generates in different size ducts; at the kind of slot size I'm looking at, their graph suggests this level of noise would equate to around 16-20m/s air velocity. Which actually ties in quite well with the rules of thumb people tend to use for reflex ports, which is reassuring; I may be on reasonably safe ground, then.
Hornresp can predict velocity at the mouth of the slot. 16-20m/s can be reached (for one FAB12) with slot-mouth areas of 150-120 square-cm. The latter is almost half what I'd originally intended, which is a useful saving. It may be possible to make further space savings by making the slot tapered into a narrow conical horn; not a lot of savings with an already narrow slot, but it also pushes up the unwanted high frequency peak slightly further from the desired frequency band so might be a small improvement if I can be bothered to cope with an angled baffle.
At Xmax at 30hz, I can expect around 105dB output at 1m from a driver. 45-50dB quieter would therefore suggest a maximum of 55-60dB at 1m for slot wind noise. The Engineering Toolbox has a page which can help to quantify airflow velocity and the noise it generates in different size ducts; at the kind of slot size I'm looking at, their graph suggests this level of noise would equate to around 16-20m/s air velocity. Which actually ties in quite well with the rules of thumb people tend to use for reflex ports, which is reassuring; I may be on reasonably safe ground, then.
Hornresp can predict velocity at the mouth of the slot. 16-20m/s can be reached (for one FAB12) with slot-mouth areas of 150-120 square-cm. The latter is almost half what I'd originally intended, which is a useful saving. It may be possible to make further space savings by making the slot tapered into a narrow conical horn; not a lot of savings with an already narrow slot, but it also pushes up the unwanted high frequency peak slightly further from the desired frequency band so might be a small improvement if I can be bothered to cope with an angled baffle.
Slot noise is a minor concern if at all. I have dual 15s in and IB SLPP with 2.5:1 compression and no slot noise. I also have JBL 5532 with a 7:1 compression ration without noise. I do believe the compression increases velocity improving perceived "hit" sound of these types of designs. There'll be a reduction in second harmonic distortion too. My adhoc measurements of the IB showed less than 2% THD at loud levels. They have very nice bass sound.
If you could afford the space and go bass reflex there would be serious room pressurization.
If you could afford the space and go bass reflex there would be serious room pressurization.
That is useful to hear, thanks. I'll check that things aren't getting extreme, but otherwise not worry too much about it. Unfortunately I don't really want the size of a bass reflex; it isn't (of course) needed for the initial very near-field situation but if/when this design ever gets used for room-sized applications I'd lean towards having three or more smallish distributed subs rather than one or two louder big ones, so compact and sealed seems to fit my situation better.
So to progress, I changed the plan/sketch to reflect what can be achieved with a smaller slot, and it has indeed been useful; saving about 10cm or 4" off the height of my previous version, losing about 1/5th of the total volume I'd intended, which is significant. The new dimensions shown below were simulated and should give a reasonable 18.5m/s maximum (slot-mouth) air velocity, yet now with quite a modest slot.
In fact there is really only one dimension left in which the slot could be sensibly reduced any further (i.e in width), and that would only change the overall box size a tiny bit now, so probably isn't worthwhile.
The reduction in size is enough that I'm coming back to making this one dual-opposed cabinet, rather than two separate halves that might be used singly or together. The drivers have nice rigid cast baskets so I think the force-cancelling effect would take benefit from a very solidly constructed slot/plenum directly coupling the two, and that would be easier in one piece.
So to progress, I changed the plan/sketch to reflect what can be achieved with a smaller slot, and it has indeed been useful; saving about 10cm or 4" off the height of my previous version, losing about 1/5th of the total volume I'd intended, which is significant. The new dimensions shown below were simulated and should give a reasonable 18.5m/s maximum (slot-mouth) air velocity, yet now with quite a modest slot.
In fact there is really only one dimension left in which the slot could be sensibly reduced any further (i.e in width), and that would only change the overall box size a tiny bit now, so probably isn't worthwhile.
The reduction in size is enough that I'm coming back to making this one dual-opposed cabinet, rather than two separate halves that might be used singly or together. The drivers have nice rigid cast baskets so I think the force-cancelling effect would take benefit from a very solidly constructed slot/plenum directly coupling the two, and that would be easier in one piece.
There is also some possibility of making the rear chamber smaller, that I want to consider next. The drivers are rated at 500W but that is RMS, and I'm only using 450W even to reach the peaks, so the RMS power the drivers will currently see is much less than they're rated for. So in theory a smaller rear chamber, requiring more power to reach Xmax, would be possible.
That said, being in a small sealed box will mean much more heat build-up than in bigger or vented designs; of course I don't want to cook them but below that I'd rather avoid significant power compression. So there will be limits to the power that can be applied, though I'm not entirely sure how to work out what these are.
Though my amplifier is able to offer 1550W (775W per driver) if they're in series at 8ohms total, or one channel can do 1200w (600W per driver) if I connect the drivers in parallel at 2ohms total. So that is at least one limit. It might be a sensible one to design right up to, given that in all these designs I'm considering peaks (excursion/power) not RMS, which should be considerably lower - i.e. for music in a small home situation. I don't really expect the drivers will be asked to reach Xmax except quite briefly and on infrequent occasions.
That said, being in a small sealed box will mean much more heat build-up than in bigger or vented designs; of course I don't want to cook them but below that I'd rather avoid significant power compression. So there will be limits to the power that can be applied, though I'm not entirely sure how to work out what these are.
Though my amplifier is able to offer 1550W (775W per driver) if they're in series at 8ohms total, or one channel can do 1200w (600W per driver) if I connect the drivers in parallel at 2ohms total. So that is at least one limit. It might be a sensible one to design right up to, given that in all these designs I'm considering peaks (excursion/power) not RMS, which should be considerably lower - i.e. for music in a small home situation. I don't really expect the drivers will be asked to reach Xmax except quite briefly and on infrequent occasions.
Last edited:
Or, on reflection perhaps the power issue isn't actually hard to decide, after all...
I'm admittedly less keen to use a nominal 2ohm setup because it puts more stress on the amps (Crown XLS2502), and probably increases distortion a bit. However, they are specified for 2ohms per channel so should cope, really. It would effectively double the channels at my disposal, which could be quite valuable if in future I ever reach my wish for multiple distributed subs. The bottom row in this table:
So with that in mind, I shall at least preserve the option to do this, by designing the (dual-driver) sub to be able to reach X-max on one channel alone if necessary. That means no more than 1200W per dual-driver sub or 600W per driver, which in turn means a rear chamber of at least 40L (20L per driver). As a bonus, this decision should avoid too much worry about heat, since 600W peak should mean a much lower RMS power than the driver rating.
I'm admittedly less keen to use a nominal 2ohm setup because it puts more stress on the amps (Crown XLS2502), and probably increases distortion a bit. However, they are specified for 2ohms per channel so should cope, really. It would effectively double the channels at my disposal, which could be quite valuable if in future I ever reach my wish for multiple distributed subs. The bottom row in this table:
So with that in mind, I shall at least preserve the option to do this, by designing the (dual-driver) sub to be able to reach X-max on one channel alone if necessary. That means no more than 1200W per dual-driver sub or 600W per driver, which in turn means a rear chamber of at least 40L (20L per driver). As a bonus, this decision should avoid too much worry about heat, since 600W peak should mean a much lower RMS power than the driver rating.
Sorry for the repeated posts, but I find that if I don't record the design process, I forget what was tried/considered and end up asking the same questions again. Anyway, I think I now have what is probably going to be the final plan for a prototype, though it needs refining before any wood is cut. It is dual-opposed, in a single box although for simplicity/rigidity of construction each driver now has its own rear chamber. It needs 600W per driver to reach Xmax within the frequency band of interest and has a maximum slot-mouth air velocity of 23m/s (which is about the most I'm currently comfortable with). So basically, the volumes involved are about the minimum for my particular circumstances.
As it happens, if I use my cupboard width as a maximum horizontal measurement, the volumes make a box with inside dimensions barely deeper than the drivers and not much taller. So that works out rather nicely and (pleasingly) is about what I'd been hoping for before doing any simulations or calculations. So I'll not make any more attempts to make it smaller. Though it might get bigger again when/if I find out why it doesn't work!
The response is similar to before, and should be identical once it is truncated/flattened from about 25hz to 80-100hz where there'll be a low-pass filter.
As it happens, if I use my cupboard width as a maximum horizontal measurement, the volumes make a box with inside dimensions barely deeper than the drivers and not much taller. So that works out rather nicely and (pleasingly) is about what I'd been hoping for before doing any simulations or calculations. So I'll not make any more attempts to make it smaller. Though it might get bigger again when/if I find out why it doesn't work!
The response is similar to before, and should be identical once it is truncated/flattened from about 25hz to 80-100hz where there'll be a low-pass filter.
That is rather like AMT (Air Motion Transformer) tweeters, hmmm? Kinda sorta?
@Kev06 can you simulate stuff like putting a U-shaped insert into the cavity, with the bottom of the U on the inside? To remove/spread some rectangular resonance? And/or a triangle/wedge between the woofers to "direct" the air flow but also to reduce (spread) resonance)
I don't personally yet know how to simulate those exact shapes in hornresp, just their affect on slot area. But I think it is possible to predict that wedges or a U-shape at the back of the slot will vary the effective slot depth, reducing it towards top and bottom. This will spread out the (1/4 wavelength-related) peak upwards and also lower its magnitude. I'm sure that I've seen this mentioned on other slot-loaded threads somewhere, but just to illustrate here is a crude sketch of a cross-section through the slot:
(I'm not sure if the dotted arrow is relevant, but the wedges would possibly introduce some other dimensions too).
I think if I were to consider shallow triangles/wedges 'between' the woofers, I'd instead prefer to angle the slot itself into a narrow triangular horn; this also pushes up the resonance a bit, so no bad thing.
However, I'm not 'too' bothered about introducing more complication; the peak shown in my previous post is already around 1.5 to 2 octaves above the intended range, so shouldn't be difficult to simply EQ out. So I might go with shaping the back of the slot because it is easy and just leave it there. This is perhaps one bonus of using (relatively) modest 12" drivers (and using active EQ). Had the drivers been 18" then the greater slot depth might make a resonance down around 180hz, or 21" drivers might see it around 160Hz, a little harder to tame without affecting the desired band.
(I'm not sure if the dotted arrow is relevant, but the wedges would possibly introduce some other dimensions too).
I think if I were to consider shallow triangles/wedges 'between' the woofers, I'd instead prefer to angle the slot itself into a narrow triangular horn; this also pushes up the resonance a bit, so no bad thing.
However, I'm not 'too' bothered about introducing more complication; the peak shown in my previous post is already around 1.5 to 2 octaves above the intended range, so shouldn't be difficult to simply EQ out. So I might go with shaping the back of the slot because it is easy and just leave it there. This is perhaps one bonus of using (relatively) modest 12" drivers (and using active EQ). Had the drivers been 18" then the greater slot depth might make a resonance down around 180hz, or 21" drivers might see it around 160Hz, a little harder to tame without affecting the desired band.
Last edited:
I have built one PPSL sub with 15" woofers. It is so large that I can't do outdoor measurements...
But I have been wondering about this "loading" bandpass effect... What is the virtual mic distance in Hornresp? Is loading only a nearfield effect? Anyway in-room measurements won't tell much and I haven't seen comparison to double front baffle woofers vs. PPSL measured outdoors
But I have been wondering about this "loading" bandpass effect... What is the virtual mic distance in Hornresp? Is loading only a nearfield effect? Anyway in-room measurements won't tell much and I haven't seen comparison to double front baffle woofers vs. PPSL measured outdoors
I'm only using the slot here as a way to bring two opposed drivers together into offering one small mouth, so haven't really gone into the loading thing. But I wasn't expecting any, and in fact haven't seen much effect of the slot in my simulations until above about 45-50hz (where a small low-pass effect begins but is then dominated by the 1/4-wave peak). The lower end looks identical to the response of a sealed box of the same rear volume (i.e. the slot appears to be doing nothing down there). Which is what I'd expected really, knowing something about the sizes that horns need to be in order to load down to low frequencies.
There is a slightly different thing with open SLOBs, where they are (I think) interested in creating some kind of differential between front and rear output. But that isn't going to be relevant in a sealed version like mine. Obviously I'll see an increased mouth air speed due to the reduced mouth area vs driver Sd, but if that translates neatly to acoustical power (which I doubt) hornresp isn't showing it - not in the way I have simulated things, anyway.
I suppose with a deep enough slot, one could move the 1/4 wave resonance down to boost lower frequencies. But I'm not personally looking for greater SPL at the cost of greater dimensions, or the downsides of resonant designs (especially if done crudely). Otherwise I would be looking at bandpass or bass reflex designs etc.
There is a slightly different thing with open SLOBs, where they are (I think) interested in creating some kind of differential between front and rear output. But that isn't going to be relevant in a sealed version like mine. Obviously I'll see an increased mouth air speed due to the reduced mouth area vs driver Sd, but if that translates neatly to acoustical power (which I doubt) hornresp isn't showing it - not in the way I have simulated things, anyway.
I suppose with a deep enough slot, one could move the 1/4 wave resonance down to boost lower frequencies. But I'm not personally looking for greater SPL at the cost of greater dimensions, or the downsides of resonant designs (especially if done crudely). Otherwise I would be looking at bandpass or bass reflex designs etc.
Last edited:
Anyway, I went out to start cutting some ply and have given up. My fingers were numb with cold before even getting the bench set up; oh for a workshop! I've got this week off and had high hopes of doing something, but the weather is making that look optimistic. So, more drawing..
I realised that the rear slot shaping/filling could be increased since really just the majority of the driver's cone needs to be clear, not the frame (or maybe even the rubber surround if some clearance is left for excursion). So a shape like this (in blue) would raise the resonant peak a bit more than previously:
I don't really need to go too far here, since EQ could deal with the already fairly high-frequency peaks, but it makes sense to do what is easy and helpful to make life easier.
I've also seen it said in PPSL designs that having the reversed driver's magnet in the slot reduces or eliminates the peak. I haven't got the slot-width to reverse a driver, and don't really want to anyway. But there isn't much reason why something else couldn't be placed part way down the slot to break up the longest resonances or reflections
Though that might be a step too far for my needs, especially as there could be unwanted effects of it. We aren't dealing with proper horns in which this might do weird things for HOMs, but it could cause issues maybe. Edit: similarly I could consider the drivers firing through a smaller hole or tap closer to the front of the slot and fill the whole of the back, but that would again increase complexity for modelling, and likely side effects. Probably I'd only look more into things like this if EQ wasn't easily available.
I realised that the rear slot shaping/filling could be increased since really just the majority of the driver's cone needs to be clear, not the frame (or maybe even the rubber surround if some clearance is left for excursion). So a shape like this (in blue) would raise the resonant peak a bit more than previously:
I don't really need to go too far here, since EQ could deal with the already fairly high-frequency peaks, but it makes sense to do what is easy and helpful to make life easier.
I've also seen it said in PPSL designs that having the reversed driver's magnet in the slot reduces or eliminates the peak. I haven't got the slot-width to reverse a driver, and don't really want to anyway. But there isn't much reason why something else couldn't be placed part way down the slot to break up the longest resonances or reflections
Though that might be a step too far for my needs, especially as there could be unwanted effects of it. We aren't dealing with proper horns in which this might do weird things for HOMs, but it could cause issues maybe. Edit: similarly I could consider the drivers firing through a smaller hole or tap closer to the front of the slot and fill the whole of the back, but that would again increase complexity for modelling, and likely side effects. Probably I'd only look more into things like this if EQ wasn't easily available.
Last edited:
One meter.
The additional air mass load slightly reduces the speaker's in box resonant frequency.
The velocity effect is near field, under one meter for the size of typical plenums used for 10" to 24" drivers.
The out of band "horn loading" extends into the far field.
SynAudCom
https://www.prosoundtraining.com/2010/06/28/far-field-criteria-for-loudspeaker-balloon-data/
In regard to full range frequency measurements, Pat Brown wrote :
"A working “rule-of-thumb” for determining the boundary between near-field and far-field is to make the minimum measurement distance the longest dimension of the loudspeaker multiplied by 3.
While this estimate is generally acceptable for field work, it ignores the frequency-dependency of the transition between the near and far fields."
He then writes:
"It is often thought that a remote measurement position is necessary for low frequencies since their wavelengths are long. Actually the opposite is true. It is more difficult to get into the far-field of a device at high frequencies, since the shorter wavelengths make the criteria in Item 4 more difficult to satisfy.
Item 4:
4. The distance from the source where the path length difference for wave arrivals from points on the device on the surface plane perpendicular to the point of observation are within one-quarter wavelength at the highest frequency of interest ."
An important distinction between high frequency and low frequency measurement, criteria #4 can be satisfied up to 210Hz for a subwoofer with a one square meter face measured at one meter with a measurement mic on the ground plane.
Art
Note that a few of the pioneer's early cinema/PA horns I was exposed to used an adjustable throat that was basically two boxes with either 1, 2, 4 drivers each hinged at the rear with top bottom plates added, so when Nelson premiered his box, just assumed he'd made it a fixed parallel variant.
Regardless, seems like the hinged variant should satisfy your mods' goals and then some.
I think that I understand your desicription, GM. Intriguing; I hadn't considered an adjustable or hinged version, I shall have to think about that and look for some examples.
I did consider making this in two (left and right) halves for transportation/storage during house moves. They could perhaps have been coupled together at an angle or with a hinge to make the slot walls adjustable. Possibly at a small cost to the rigidity of the (mechanical) coupling between opposing/force-cancelling drivers. The hornresp simulations do show a slight improvement from making the slot walls conical, rather than parallel. Perhaps, whether hinged or not, I should just bite the bullet and make it so, it is only a small bit more complex to make, after all.
Though I imagine the reality might diverge from the simulations a little, anyway. Much of the slot's sides are essentially the driver cones, which of course are concave and I haven't modelled exactly. I've simulated various shapes of horn/slot though, just wrt cross-sectional area, and it doesn't seem to be particularly sensitive or critical, which is reassuring. Perhaps thats because I'm not designing for any particular loading or horn-type gain, all the slot is really doing is bringing two drivers together.
I did consider making this in two (left and right) halves for transportation/storage during house moves. They could perhaps have been coupled together at an angle or with a hinge to make the slot walls adjustable. Possibly at a small cost to the rigidity of the (mechanical) coupling between opposing/force-cancelling drivers. The hornresp simulations do show a slight improvement from making the slot walls conical, rather than parallel. Perhaps, whether hinged or not, I should just bite the bullet and make it so, it is only a small bit more complex to make, after all.
Though I imagine the reality might diverge from the simulations a little, anyway. Much of the slot's sides are essentially the driver cones, which of course are concave and I haven't modelled exactly. I've simulated various shapes of horn/slot though, just wrt cross-sectional area, and it doesn't seem to be particularly sensitive or critical, which is reassuring. Perhaps thats because I'm not designing for any particular loading or horn-type gain, all the slot is really doing is bringing two drivers together.
Thank you for this, it caused me to look a bit more closely at what the slot is doing. I've overlayed hornresp's acoustical power predictions for 20L sealed vs 20L sealed+slot and found them to be incredibly similar in the lower frequencies, as I thought
However, that is because both have been normalised to the same driver excursion (Xmax), which is hiding some effects. If we look at the power needed to reach Xmax that does differ
The slot has reduced the power needed slightly. It has also lowered the frequency at which xmax is reached, via a slightly lower in-box resonance, as you say. So, had my acoustical power charts been made at the same electrical power, rather than the same driver excursion, there would have been a small but noticeable difference.
It probably doesn't make too much difference in my case, but it is still slightly helpful. These days I tend to consider the limits based firstly on what the driver can do, to which power and EQ will just be what they need to be. Here I don't think I'm seeing any acoustical gain from the slot, i.e. the SPL doesn't increase for a given driver excursion (edit: ignoring the above-band slot-depth resonance, anyway)
So that has been very useful to look at and discover, thank you, Art.
Last edited by a moderator:
If I wanted to remake or improve my PPSL 2x15" I would use V slot instead of U. Just instinction, no sims behind this. However the original works just fine for me, the room is more problematic. Using minidsp with delay control for both sub and mains was a huge improvement for stereo music. Movies use normal AVR 5.1 setup. Hypex DS 2.0 plate amp has fixex highpass at 15Hz.
Last edited:
So just for completeness, here is the predicted acoustical power of 20L-sealed vs 20L-sealed+slot at similar electrical power:
So if one had limited amplifier power and cannot enlarge the rear sealed chamber, then the slot does offer a little bit more output at low frequencies. It is possible/likely that this gain could be made a little larger with different slot designs, but I haven't aimed for this so haven't looked at optimising it, myself.
Though, if one has enough amplifier power, or can enlarge the rear chamber slightly, to bring the driver to similar excursion in both cases, then the slot would seem to make no difference in maximum (exursion limited) acoustical output.
So if one had limited amplifier power and cannot enlarge the rear sealed chamber, then the slot does offer a little bit more output at low frequencies. It is possible/likely that this gain could be made a little larger with different slot designs, but I haven't aimed for this so haven't looked at optimising it, myself.
Though, if one has enough amplifier power, or can enlarge the rear chamber slightly, to bring the driver to similar excursion in both cases, then the slot would seem to make no difference in maximum (exursion limited) acoustical output.