Perhaps the rub. Optimum construction techniques leveraging modern capabilities might be a more fruitful discussion than revisiting roofing materials and wood panels. The latter feels like debating the optimum approach for shipping ice over long distances.
Yes, I missed your post, nice work by the way!
I took a look at your link there and noticed that there are two crossovers
the CS1 derivative of a KEF model and the CS1a derivative of the LS3/5a.
Just noticed that they kept the rear woofer mounting when I thought that
they had modernized it a bit and went with front mounting.
Here are the CS1 details with the B110 front mounted, with a different crossover:
Item Description: KEF Electronics Brochure - Part No SL353 EN01
I wonder how they compare sonically?
200 years ago we solved the problem of shipping ice using nothing but a waste byproduct that every saw mill had plenty of, sawdust.
Roofing materials like felt paper, bitumin and butyl rubber are used due to their known dampening properties. The plus side of this is low cost, put the word "acoustic" next to any product and the price goes up 10x fold. There are other products which have excellent dampening properties, Green Glue comes to mind along with a few latex based caulking compounds which are mighty expensive. Due to the complexity, expense, let alone the difficulty in measuring said cld panels where we would all mutually agree in the setup and equipement used. That's a nearly impossible proposition, don't you think?
If each of us actually contributed to solving the problem, instead of say, buliding yet another simple wood box everytime we would certainly be ahead of the game.
Currently am working on these issues. Do believe the inherent strength of a curved panel greatly helps. Multiple thin layers with at least two cld regions, a mass region and a decoupling region which can be made from vinyl mat or felt. Internal bracing done cld, seperate from the cld motor mount and isolated (yes cld) baffle mount. Clay comes in handy for this. Another interesting facet is that externally dampening a cabnet wall with natural fiber carpet helps lower the audibility of the higher resonances. Problem is I won't do fuzzy car boxes in my livingroom. Cats love them for that too. If you have this issue, on their first inspection of that thing that begs to be climbed, scratched or pissed on, with the flick of a switch blast them with 110dB of white noise. That'll cure them for life... once removed from the ceiling
Forgot to add, each layer is in the thickness range of upto 1/2". With one vinyl layer that's 3mm bonded between two 3mm thick layers of cardboard. Have some BB in 1/4", 3/8", 1/2" and 1/4" mdf. As well as some GG and butyl sheeting that goes into the mix.
The mass layer I'm contimplating might end up being screen door screen with lead pellets filling the holes in a pattern (?), spray glued (headliner rated), another layer of screen and pellets in a differrent alternating pattern, glued and bonded to a layer of butyl. The GG is used for isolating the outer shell.
Dave, I often find mass loaded cld sheets used in dampening auto panels is in the 1lbs/ft range. Am wondering if this was choosen due to the typical guage of steel used in doors is optimal? Typical thicknesses of either 3mm or 6mm, foil faced or not.
The mass layer I'm contimplating might end up being screen door screen with lead pellets filling the holes in a pattern (?), spray glued (headliner rated), another layer of screen and pellets in a differrent alternating pattern, glued and bonded to a layer of butyl. The GG is used for isolating the outer shell.
Dave, I often find mass loaded cld sheets used in dampening auto panels is in the 1lbs/ft range. Am wondering if this was choosen due to the typical guage of steel used in doors is optimal? Typical thicknesses of either 3mm or 6mm, foil faced or not.
My experience with resonant structures is that even if you try and drive them without exciting the modes, they pick up some stray energy and are eventually in "full song". With speaker cabinets most modes seem to be full cabinet modes. They couple through the edges and some panels will be moving inwards while others are moving outwards. I don't think, in practice, you will find one side active and an adjacent side still.
David S
In the end, the popular materials will always be the ones that the basement constructer can work with his table saw. If they aren't wood they will be "wood like" (as MDF should really be described). Even interesting materials such as Corian are less popular due to difficulty of working.
The greatest promise for the diy enthusiast are probably building up layers in a CLD sandwich. The core material becomes unimportant and you can layer to your desired thickness. Two part polymers (we used Polyoll for some applications at KEF) are interesting. Dropping ceramic tiles into the polymer will allow you to add mass without excessive stiffness. Neoprene rubber between layers of MDF is a good solution also. A lot of the systems I launched at Snell had a sandwich baffle built with that technique.
Roofing materials and wood might still be near the top of the heap. It isn't about new or old, but rather the intrinsic material properties. 3D printed cabinets are not the solution!
Thanks for the thoughtful response. Just a few points
- For a given mass and damping, mechanical q is proportional to the square root of the stiffness;
https://www.easycalculation.com/engineering/mechanical/learn-mechanical-q-factor.php
- amplitude of vibration increases 6 dB per halving of damping, but does acceleration as well? Given your early comment that SPL is proportional to acceleration, this is probably more applicable. I haven't worked with these formulas in many years so I don't know the answer
- What differentiates the Harwood studies from the others and perhaps make it a bit of date is the negative slope vs freq of the q audibility curve is quite a bit less in the more modern studies. But I do agree his analysis was excellent and very useful within the context given, its just a fairly narrow context
Dave
Precisely, this feels a bit like debating the relative merits of oak versus ash dust in a finite element analysis world. Your last paragraph sounds like the promising approach.
Not really a requirement. Almost no DIY builder performs the sort of confirmation measurements described, going on faith that even the wood available at Home Depot or Lowes has much in common with the samples in those old BBC studies. Noise control and associated products have become much more common in the last forty years. Home builders have also become much more comfortable with alternatives to monkey coffins. It's nearly incomprehensible that a half-century old construction technique based on common materials of the day still represents the best performance/value mix today.
Agree entirely on CLD as an approach. As I mentioned earlier, the JR149's thick bitumen on thin aluminum was not impressive. The entire cabinet vibrated. CLD made of 1/8" linoleum floor tile and thin aluminum bonded with polyurethane glue was almost enough to tame a poorly considered 1/4" welded steel cabinet experiment. I suspect it, or similar construction using plexi, etc., would handily trump plain bitumen and still be garage friendly.
Yes, both systems based on the different models of the B110.
The CS1 has the B110SP1057 (B-model with a little bit stronger magnetic values) front mounted and CS1a the B110SP1003 (A-model) rear mounted.
With the combination of T27 and B110A KEF had had a lot of small famous speakers like the Coda.
I think, that KEF tried to build a simplier and cheaper derivative for diy.
Don't know, I only have the A-model and never listened to the other one.
At least I am wondering why this simple thread where a member just asked to build a speaker system with T27 and B100 is misused for discussion about thickness of different materials
.
Last edited:
DDF said:
IMHO in a mechanical system "Q" cannot be traded vs "level" in terms of audibility. If you raise the Q you raise the level.
David pointed out, that Fryer's study was based on constant level but variable Q conditions using a parametric equalizer. But this is not how speaker cabinets act, because here level is dependent on Q.
Stating that higher Q resonances are more unlikely to be audible in a mechanical system is simply a misconception based on Fryer's data, which has spreaded around itself more sucessfully than the correct interpretation:
The transfer step from
"parametric equalizer having constant level" to "passive system having level depending on Q"
is often omitted.
I am a bit surprised, that also my mentioning dispersion of bending waves and coincidence frequency does not seem to be considered. The efficiency of radiation from a cabinet's walls depends on the propagation speed of bending waves.
As stiffness goes cubic with thickness a thicker wall has lower coincidence frequency. This factor alone gives a clear upper limit for a wall thickness to be useful, especially when using no damping.
The coincidence frequency tells when a cabinet wall e.g. made from MDF solely crosses the line from "unessessary thick" to "daffy thick".
A cabinet's walls are "daffy thick" if the coincidence frequency is within the used bandwith of the cabinet or just closely above.
E.g. a panel of 20mm thickness made from plywood - dependent on E-module and mass/area - may have a coincidence frequency Fc around 900Hz.
http://www.sga-ssa.ch/pdf/events/Buetikofer_Abgestrahlter_Koerperschall.pdf
See "Figur 6", page 5: Efficiency of radiation rises near coincidence.
http://mech.vub.ac.be/avrg/Courses/Acoustics/Acoustics lesson 3.pdf
See pages 12,13 for coincidence effect.
Last edited:
This is about some practical composite Materials:
HAUPT
http://www.picosound.de/GehMatAll.png
I interpret the highest peaks as being close to coincidence frequency.
The peaks would possibly be even higher if the microphone were not placed in front of the middle of the panels. The max. radition from a panel at coincidence is to the sides: Lobes occur in parallel to the plane of the panel.
Above coincidence there may also be radiation lobes in other directions.
HAUPT
http://www.picosound.de/GehMatAll.png
I interpret the highest peaks as being close to coincidence frequency.
The peaks would possibly be even higher if the microphone were not placed in front of the middle of the panels. The max. radition from a panel at coincidence is to the sides: Lobes occur in parallel to the plane of the panel.
Above coincidence there may also be radiation lobes in other directions.
Last edited:
That's right
, but I think he wanted to get pratique suggestions for a cabinet and not a philosophique discussion 
of different materials, even if the author asked for:Sorry but that is my opinion on such a thread.
Hi Line Array,
I am aware of coincident frequency effects from architectural acoustics but I'm not sure how to apply the theory to cabinets. It seems that the theory describes a large panel phenomenon, i.e. If a straight transverse wavefront moves across a large surface, the surface is transparent for some frequencies and angles.
Does it still apply to cabinets or should we just consider modal behaviour?
Regards,
David
Hi David,
coincidence does still apply even to small structures and modal vibration.
The effect applies in a slightly different way, than treated in architectural acoustics, when looking at sound going through walls of larger area.
When looking at a simplified example of a panel or a beam vibrating in such way that area is divided by a single node line into two sub-areas moving in opposite phase, it will be important how large the bending-wavelength on the panel or beam is in relation to the wavelength of the sound radiated into the air.
You can imagine varying the bending wavelength is like varying the "dipole path length" of our "modal radiator".
If wavelength of the bending waves is very small compared to wavelength of sound of the same frequency in air, positive and negative pressure zones near the panel will have "time enough" to cancel out by "pressure equalisation".
This is why there is only rather small radiation into the far field, given that bending wavelength is small compared to sound in air.
When bending wavelength is "comparable" or even larger than wavelength in air, you get an efficient radiator, which is to be avoided in a loudspeaker cabinet.
Typically at and above coincidence frequency Fc, where bending wavelength is equal to wavelength of sound in air, you will get a lobing radiation and much higher efficiency than for frequencies far below Fc, where bending wavelength is small compared to wavelength of sound in air.
To me this is a clear indicator, that resonating cabinets should not be used near or above coincidence, unless vibration is kept very very small.
And keeping coincidence out of the bandwidth would place another upper limit for the stiffness/mass relation of the cabinet walls.
Here also mass/area comes into play in lowering the propagation speed (phase speed) of bending waves on the panel and thus making the bending wavelengths smaller.
We have to look at the dispersion relation, to estimate bending wavelengths of the given material/composite in the frequency range to be used.
cf.
"Structure-Borne Sound: Structural Vibrations and Sound Radiation at Audio Frequencies",
by L. Cremer, M. Heckl, Björn A.T. Petersson
Springer Science & Business Media, 05.12.2005 - 608 Seiten
https://books.google.de/books?id=pA...0aqDXgNgB&ved=0CCYQ6AEwAA#v=onepage&q&f=false
Page 55, (3.85)
Kind regards
Last edited:
I have given up wondering . . . seems like obsessive single mindedness about esoteric materials, an inability to understand seemingly straightforward questions or simple bad manners?<snip>
At least I am wondering why this simple thread where a member just asked to build a speaker system with T27 and B100 is misused for discussion about thickness of different materials
Who cares, have a great 2015 y'all
surv1v0r said:
I wish a happy new year to all too !
... e.g. "MDF", "plywood", "polymer-bitumen" and other roofing materials seem "esoteric" to you already ?
To me it is much more "esoteric" making loudspeakers using 70's driver designs without need. So i guess you have to make the best cabinet and crossover possible ... just to lay some sense into such a project at all
http://p10hifi.net/planet10/TLS/drivers/images/T27-dataS.gif
Usually i would write posts like
"3 (or even 10) reasons why tweeters like KEF T27 are outdated".
Instead many are contributing in how to make the best of it. You may of course call that "single minded" and you are not too far off by stating that i fear ...
I mean the best outcome will be a loudspeaker with
- crossover frequency far too high and being in an auditively critical range
- a radiation pattern much narrower than necessary in upper mids to lower highs
- a more than necessary pronounced discontinuity in radiation pattern at crossover frequnency
- a rather low dynamic headroom, since linear excursion of the B110 unit is small due to e.g. outdated ( or at least unathletic) motor design http://p10hifi.net/TLS/drivers/images/B110/B110_SP1003.gif
- ...
What do you like to achieve ?
Having a cabinet virtually free of coloration and a decent crossover in line with adequate baffle design and driver's arrangement on the baffle would be the only plus points justifying such a project (except for nostalgy of course) .
Last edited:
You've got a good point there.
But I can't believe the combination of speakers, with their weaknesses, would benefit from a coloration-free cabinet. I think they benefit from a cabinet with a well chosen coloration-spectrum to hide and supplement the weaknesses.