Member
Joined 2009
Paid Member
Well it seems that a box design has to go to some lengths if it is to absorb the back wave without creating other problems - first off it seems that you ideally want something along the lines of the Nautilus, or at least a pipe that stretches out back behind the driver with suitable damping. Something that doesn't offer a surface just behind the cone to reflect sound back to the cone (which some floor standing TL's appear to be designed to do !) and such a pipe out the back should not be set up so that it introduces ripples into the FR - means giving up bass extension. I can't say I've seen many examples of that ?
The other options seems to be to let the back-wave out the back in an OB arrangement and deal with it in the room, either through room treatments, or by positioning the speaker away from the walls and using the back-wave as a reverberant sound source. Can we move the room treatments to the back of the speaker - like the Hartley BOFFLE - is this the ultimate solution ????
From another thread I came across the Philharmoinicaudio - kind of reminds me of the one that Cal made with an open backed mid, closed back tweeter and woofer: philharmonicaudio.com
The other options seems to be to let the back-wave out the back in an OB arrangement and deal with it in the room, either through room treatments, or by positioning the speaker away from the walls and using the back-wave as a reverberant sound source. Can we move the room treatments to the back of the speaker - like the Hartley BOFFLE - is this the ultimate solution ????
From another thread I came across the Philharmoinicaudio - kind of reminds me of the one that Cal made with an open backed mid, closed back tweeter and woofer: philharmonicaudio.com
Last edited:
Moving/attaching damping to the rear of the cabinet to attenuate the back wave ... sure you can do that. Might be preferable in certain room situations.
Well, Bigun, we've certainly wandered around on all the themes of speakers. Open Baffle, Infinite Baffle, Transmission Line, Reflex. I reckon you just do the best you can with each type, unless you can use some other room as the box!
I don't need a speaker like Troels' Poor Man's Strad design, but he ticks a lot of boxes here, down to the damped vented midrange enclosure and big baffle. This has got to be a very fine reflex freestander. The rest is down to placement and room acoustics I reckon.
Poor Man'
The drivers are not universally usable, of course. Reflex standmount can get away with drivers that will fail miserably with wallmounters due to the bigger coils you use.
I don't need a speaker like Troels' Poor Man's Strad design, but he ticks a lot of boxes here, down to the damped vented midrange enclosure and big baffle. This has got to be a very fine reflex freestander. The rest is down to placement and room acoustics I reckon.
Poor Man'
The drivers are not universally usable, of course. Reflex standmount can get away with drivers that will fail miserably with wallmounters due to the bigger coils you use.
Attachments
No I was not expecting to have it both ways. I was looking at a highly damped TL not becuase I expected any output from the terminus but because any energy that was not absrobed by the damping would be allowed to escape the box instead of reflecting towards the driver's cone.
So in short what you are saying is that if we damp the TL to the point where much of the energy is absobed and only that energy that is not absobed escapes the box; the box; in effect; is a sealed box?
I am not sure if the Neo 8 rear wave is dipole. I think the Philharmonic places the Neo 8 in a sealed box.
I have heard the Philharmonic 3 and it is a nice speaker just a tad SPL limited. The person who heard it with me (he was in the market for a new pair) eventually settled for a pair of very smooth sounding Gradient Revolution (I think the bass is OB).
Maybe a Phil 3 clone using a larger woofer (SS 26W/886x or Seas W26FX-0x), mated to a larger (B&G Neo 10) mid and RAAL tweeter (70-10D will do fine) would have satisfied but that is another topic.
I have been out of town for a couple of days and have not read all the posts so forgive me if this has already been mentioned.
But going back to the actual enclosure construction and any subsequent colouration that it contributes the following article has been very useful. It appeared in the local Oz mag' "AUSTRALIAN ELECTRONICS MONTHLY" Feb'1987 (Now deceased I'm afraid).
They came to a number of conclusions. The most useful being that subjectively, colouration was mostly dependent on decay time....the shorter the better.
And to summerize their work. The best way to achieve rapid decay times was a combination of high mass, high stiffness and high damping. The author was then in Australia but had previously worked for B&W (UK). He illustrated the findings with "waterfalls" from B&W DM1200, CelestionSL600 and then a B&W Matrix model. These represented, chipboard and pads, Aerolam and obviously a matrix enclosure.
The measurements seemed to confirm the theory very well.
Of course cynics may quote the German proverb from Gilbert Briggs book on Loudspeakers (Wharfedale 5th edition.) "When theory and practice agree....both are wrong!"
But I found it does give us a diyers a well reasoned theory to assist our planning.
Jonathan
But going back to the actual enclosure construction and any subsequent colouration that it contributes the following article has been very useful. It appeared in the local Oz mag' "AUSTRALIAN ELECTRONICS MONTHLY" Feb'1987 (Now deceased I'm afraid).
They came to a number of conclusions. The most useful being that subjectively, colouration was mostly dependent on decay time....the shorter the better.
And to summerize their work. The best way to achieve rapid decay times was a combination of high mass, high stiffness and high damping. The author was then in Australia but had previously worked for B&W (UK). He illustrated the findings with "waterfalls" from B&W DM1200, CelestionSL600 and then a B&W Matrix model. These represented, chipboard and pads, Aerolam and obviously a matrix enclosure.
The measurements seemed to confirm the theory very well.
Of course cynics may quote the German proverb from Gilbert Briggs book on Loudspeakers (Wharfedale 5th edition.) "When theory and practice agree....both are wrong!"
But I found it does give us a diyers a well reasoned theory to assist our planning.
Jonathan
Yes, I follow that, Jonathan! 🙂
Aerolam Reinforced Aluminium panels are probably a bit exotic for most of us. Here's a Steen Duelund "Matrix" type closed box design made of MDF I suppose.
One of Steen's hobbyhorses is the glue holding the matrix must be a slightly soft type to give a path for energy to dissipate. But for sure, the matrix holds the wadding in a way that encourages energy absorbtion. I don't think I'd go quite this far, but the idea is useful. Notice the empty cavity in the middle. Discontinuities like that encourage scatttering.
Aerolam Reinforced Aluminium panels are probably a bit exotic for most of us. Here's a Steen Duelund "Matrix" type closed box design made of MDF I suppose.
One of Steen's hobbyhorses is the glue holding the matrix must be a slightly soft type to give a path for energy to dissipate. But for sure, the matrix holds the wadding in a way that encourages energy absorbtion. I don't think I'd go quite this far, but the idea is useful. Notice the empty cavity in the middle. Discontinuities like that encourage scatttering.
Attachments
a well stuffed TL is NOT a closed box though.
i hate sweeping generalistation even though i have been guilty of it! The flattening of impedance is a worthwhile boon, alone. My ideal would be an unfolded aperiodic TL, extending in the depth dimension. Only a closed box of the same dimensions AND correct volume, would approach this. 6ft of stuffing makes ALOT of difference. You could always make a closed box like this, of course, but it would have to be non optimal to flatten Z, with the corresponding overdamping of LF. I know which i would choose.
i hate sweeping generalistation even though i have been guilty of it! The flattening of impedance is a worthwhile boon, alone. My ideal would be an unfolded aperiodic TL, extending in the depth dimension. Only a closed box of the same dimensions AND correct volume, would approach this. 6ft of stuffing makes ALOT of difference. You could always make a closed box like this, of course, but it would have to be non optimal to flatten Z, with the corresponding overdamping of LF. I know which i would choose.
Most of the TL data that I have seen shows non-flat impedence curves. With light stuffing (and heavy mid band ripple) you can see a typical vented box splitting of the free air resonance in two. As soon as you move towards the amount of stuffing needed to get mid band ripple less than 2 dB (a reasonable target) the impedance tends back towards a sealed box single resonance.
This of course makes sense. The only way for the line to have an effect on the woofer impedance is with a fairly strong reflection from the end point. The only way to have smooth response is with line vent output reduced 10 to 15dB in the midrange. These are mutually exclussive conditions.
Who needs a flat impedance curve anyway?
David S.
This of course makes sense. The only way for the line to have an effect on the woofer impedance is with a fairly strong reflection from the end point. The only way to have smooth response is with line vent output reduced 10 to 15dB in the midrange. These are mutually exclussive conditions.
Who needs a flat impedance curve anyway?
David S.
Member
Joined 2009
Paid Member
I'm starting to think seriously about a convertible. I don't just mean an Aston Martin either 😀
I'm thinking of designing an OB to which I can attach a box at the back and so make it possible to switch the design on-the-fly so to speak. What I'm not sure about is how feasible this is - if it's limited to a mechanical design challenge or if making such changes would screw-up the XO design. I want a fixed XO design and have the option of only changing the back. I could 'bolt on' a) a closed sealed box, b) a vented box, c) a boffle type back.
Maybe it would be more than just an experiment, it could be that I would change the back depending on where I was using the speaker (outside on the deck verses indoors) etc.
I don't think I'd build this particular design, but I like the approach which gives a broad and curved front baffle. I think this is something I might try to incorporate into a future design.
I'm thinking that you'll be stuck here concerning the xo. The actual frequency wouldn't change ... just all the other stuff (BSC, OB gain/attenuation, LF roll off, .. etc) associated with the different enclosure alignments. Don't see how one xo approach could offer satisfaction for several design choices. ... but then, you'll have to choose a set of drivers that like each variation. Not gonna happen.
Why not just build an OB and see for yourself if it's something you'd like? They're stupid simple from a construction POV.
Last edited:
Member
Joined 2009
Paid Member
I did build an OB a few years ago so I know that I like the sound: http://www.diyaudio.com/forums/full-range/174632-apollo-ob.html
However, I was also impressed by the listening impressions reported by people who built the InConcert Miles bass-reflex. See post 62 of this thread. And so I have to ask myself if I wouldn't be better off building a proven design such as that. I do want a high efficiency speaker for a low power SET.
Why not build the Miles with removable back ?
In thinking about a convertible. The XO in a 2-way is going to be up above the bass, and the enclosure mostly affects the bass (apart from colourations that is) so the enclosure is not affecting the XO region - hence why wouldn't I be able to set up the XO for, say 900Hz, and then leave it alone ?
However, I was also impressed by the listening impressions reported by people who built the InConcert Miles bass-reflex. See post 62 of this thread. And so I have to ask myself if I wouldn't be better off building a proven design such as that. I do want a high efficiency speaker for a low power SET.
Why not build the Miles with removable back ?
In thinking about a convertible. The XO in a 2-way is going to be up above the bass, and the enclosure mostly affects the bass (apart from colourations that is) so the enclosure is not affecting the XO region - hence why wouldn't I be able to set up the XO for, say 900Hz, and then leave it alone ?
Last edited:
A "U" frame will boost the LF output .. as in Miles without a back. ... and still leave a guy with nice articulate bass. Personally, I prefer the framed driver(s) vs a pure flat baffle in that it allows more flexibility in the overall sound (tone/output) ... it's my cup of tea.
Only speaking in general terms about the convertible option ... the frequency can stay the same, it's the other conditions that change when removing an enclosure back that didn't exist when the back was on ... many being OB step loss/gain which could be adjusted w/eq or within the xo itself (components added to the passive circuit to contract the differing enclosure conditions ... not a frequency change).
With a xo point of 900hz I'm fairly certain that you'll be adjusting driver output in the 200hz region (or so) getting the midrange to sound good. Something that the Miles (as is) wouldn't have an issue with (haven't heard them so .. just sayin'). OB or it's variants are a bunch easier to deal with using an active processing approach. They do take a considerable amount of fiddling with, especially when enclosure design is in flux. (speaking from my limited experiences with them)
Only speaking in general terms about the convertible option ... the frequency can stay the same, it's the other conditions that change when removing an enclosure back that didn't exist when the back was on ... many being OB step loss/gain which could be adjusted w/eq or within the xo itself (components added to the passive circuit to contract the differing enclosure conditions ... not a frequency change).
With a xo point of 900hz I'm fairly certain that you'll be adjusting driver output in the 200hz region (or so) getting the midrange to sound good. Something that the Miles (as is) wouldn't have an issue with (haven't heard them so .. just sayin'). OB or it's variants are a bunch easier to deal with using an active processing approach. They do take a considerable amount of fiddling with, especially when enclosure design is in flux. (speaking from my limited experiences with them)
That's prettty much how I build everything, and have for ages. Midranges get their own long tapering tube a la Nautilus. Works very well.
Member
Joined 2009
Paid Member
Something that occurred to me later was the possibility that edge diffraction might in fact cause a small amount of ringing - if you have a baffle with a sharp edge the wave that travels along the baffle and reaches an edge will diffract in all directions - including back the way it came, which would then pass back across the baffle and diffract from the opposite side, a portion of which would then diffract back the other way and so on... hence a portion of the energy travelling back and forth from one edge to another.
However unlike a reflection between two flat walls where the loss per reflection can be very low thus leading to a high Q resonance I would imagine that only a very small percentage (how much ?) of the total diffracted signal will turn 180 degrees and travel back the way it came so the loss per "reflection" would be very high. This would lead to energy of the resonance being dissipated very rapidly and any efforts to reduce diffraction by rounding the edges etc would cause it to dissipate even more rapidly, so it might be more of a theoretical consideration than a practical concern.
On the other hand a flat open baffle is going to have by far the worst diffraction effects since there is a full 180 degree angle at the point of diffraction, so it might have some relevance there. Does a standing wave forming between diffraction edges explain the large dipole peak that occurs before rolloff, or is that more to do with phase shift between front and rear radiation causing in phase addition at some frequency ?
I'm not sure I understand both of those impulses - in the first are you saying that there is a single reflection with a half cycle delay ? If so I don't see how the ongoing ringing is occurring... perhaps some software artefact rather than a real effect ?
In the second one when you say multiple reflections that makes sense - since a standing wave resonance is just multiple equally spaced reflections...
Member
Joined 2009
Paid Member
so when you said that's how you build 'everything and have for ages' you mean you haven't ever finished anything or you mean the special pipe out the back for the mids is a new adventure for you and haven't got anything to share yet - I'm confused
No, I've built lots of boxes this way (matrix bracing), including the last lot of PA speakers and a couple with the mid tube. My mains are Unitys and don't need the mid tube and I won't bother with it for the surrounds now. The design I have for the bedroom speakers will use if for the MB and MF, but that is last project and receiving less effort now as I collect the parts slowly as budget allows. However, my design will probably not translate to other designs, so you'd be far better off to visualise how to apply the principle to what you want to do. My bed is very tall, so the speakers will be too, so they bend through 90* and head towards the floor so can be longer than most people could use this way.
I am not much of a photographer and didn't even own a camera until recently and didn't feel the need to document stuff for other people, so I never took pics. Because I'm still not much of a photographer and have decided on what I'm doing with the current designs, I will post it all when I'm done rather than discuss it as I go as I find that distracting.
Member
Joined 2009
Paid Member