It could, but it would be very hard to keep it in the right place without a whole lot of work, tuning, and oblivious to changes in weather.
So possible, but not very practical.
The set of early Klipsch Heresy i got had screwed on ⅜” unbracedply back panel. It acted like a PR and sounded really bad.
In 2 of the 3 decay graphs Juhazi posted we can se ethe strong resonance at about or just below 100Hz. This is where the BBC wants to see the resonance, and in the LS3/5A near the bottom of its LF reach.
You can hear these resonances (or at least i can).
The 3rd LS3/5A shown, at least implied by the measures, is not properly constructed to be an LS3/5.
I have heard MANY LS3/5A and i personally do not think they live up to the hype.
dave
Agreed 100% they don't, largely because the hype has little to do with the purposes of the design. The LS3/5 & LS3/5a were developed for use as nearfield monitors in small permanent studios and mobile broadcast vans, and the primary goal of that design was natural voice reproduction in those kinds of spaces. For that purpose, IMO they're still good. Not the best -they have the weight of mythology to support. But good. And within that particular design remit (minimal voice colouration in a sub-compact monitor), they were one of the first of their kind, so they are historically important for that, if nothing else.
Edit: the first pair, with screw-on backs, has attained an almost legendary status for quality. Other than the extreme pair-matching of drivers & crossover components, it's possible that this further helped break up panel modes. I haven't seen any comparative measurements of its panel behaviour as opposed to some of the others, so that's pure speculation on my part. When Sterling Broadcast were still producing different LS3/5a cabinets for people to roll their own a few years back, I seem to recall their 'reference' offering was based on that 1st pair. It may also have had a bit more of a leak-path, however minor, which may have ended up causing greater consistency in different climatic conditions. YMMV.
Edit: the first pair, with screw-on backs, has attained an almost legendary status for quality. Other than the extreme pair-matching of drivers & crossover components, it's possible that this further helped break up panel modes. I haven't seen any comparative measurements of its panel behaviour as opposed to some of the others, so that's pure speculation on my part. When Sterling Broadcast were still producing different LS3/5a cabinets for people to roll their own a few years back, I seem to recall their 'reference' offering was based on that 1st pair. It may also have had a bit more of a leak-path, however minor, which may have ended up causing greater consistency in different climatic conditions. YMMV.
Last edited:
A mechanic's stethescope much more quickly tells you where panels are vibrating, the accelerometer will give some quantitative information.
dave
If anyone here thinks that they can second-guess the research department of the BBC, good luck to them. ANY novel enclosure design built without in-depth testing and both quantitative and qualitative analysis of multiple prototypes is a finger-in-the-air guess. No more, no less, so please don't anybody kid themselves that they can get it right first time in their garage or shed...
Build what you feel is 'right' and it will probably sound good to your ears. Expectation bias is a powerful thing.
Build what you feel is 'right' and it will probably sound good to your ears. Expectation bias is a powerful thing.
This doesn't look like good reasoning. Firstly, the important low frequency resonant modes tend not to involve panel resonances which tend to occur at higher frequencies but the speaker deforming as a whole and things like the midwoofer bouncing on the front baffle. Considering the motion of individual panels rather than the motion of the speaker with drivers attached can be misleading.
Secondly, the baffle and the 5 others sides have different requirements. The rate of work put into the speaker cabinet is the product of the force from the drivers (fixed without something like isolating grommets) and the velocity of the baffle (reducible with good design). The stiffer the baffle the less energy is put into the speaker cabinet. The energy in the cabinet is then removed by sound radiation (undesirable) and the motion of the cabinet working against damping forces.
Bracing to stiffen the baffle will reduce the amount of energy entering the cabinet. Bracing to stiffen the other 5 walls raises the resonant frequencies but the SPL of the radiated sound is typically about the same (e.g. see KEF or BBC paper for evidence). It does little to nothing in terms of making the cabinet quieter despite the many posts you will see here to the contrary. Effective damping however can reduce the undamped high Q cabinet resonances by tens of dB (e.g. KEF paper for effective and BBC for useful).
1960-70's technology is past and gone. Don't you think science and loudspeaker technology have advanced in 60 years?
Rover 200TC isn't a good car by today's standards! A ride in a 1969 Rover 2000TC
Attachments
Last edited:
Materials science and technology have indeed seen a large number of developments since the '60s and '70s. However, we shouldn't get too carried away; if a given approach to mechanical engineering was functionally effective 50 years ago, it hasn't mysteriously ceased to be so in the intervening decades. The fact was, the BBC's damped thin-wall approach was very effective to its target goals then (when properly implemented), and it remains so now. We just have more alternative options realisitically available to us to select from than there were when it was developed.
For the rest -don't forget we're all of us standing on the shoulders of Olson, Beranek and the pioneers from W.E., Bell, RCA, Stromberg, Jensen etc. who did much of the groundwork for the rest of us back in the '30s - '50s, and created some designs which still equal (or outperform) subsequent offerings. Look at the LC-1A or some of the compression + horn offerings. These were some of the most capable engineers around, with essentially unlimited budgets for R&D. About the only things they didn't create were digital (and they laid some of the very early groundwork for that), and some modern materials.
Agreed about the 2000TC though. Ghastly. A Low-Drag Lightweight E-type (very important it's that particular permuation) on the other hand...
For the rest -don't forget we're all of us standing on the shoulders of Olson, Beranek and the pioneers from W.E., Bell, RCA, Stromberg, Jensen etc. who did much of the groundwork for the rest of us back in the '30s - '50s, and created some designs which still equal (or outperform) subsequent offerings. Look at the LC-1A or some of the compression + horn offerings. These were some of the most capable engineers around, with essentially unlimited budgets for R&D. About the only things they didn't create were digital (and they laid some of the very early groundwork for that), and some modern materials.
Agreed about the 2000TC though. Ghastly. A Low-Drag Lightweight E-type (very important it's that particular permuation) on the other hand...
Last edited:
To expand on a good point. The BBC approach was never intended to be high performance but just good enough to make the cabinet resonances inaudible when things were done as specified. Their goals were not that well aligned with typical home speaker DIYers today. The largish monitors had to be light enough to lugg around reasonably comfortably which is not a requirement for many here. They had to be relatively cheap and easy to manufacture for the low tech speaker manufacturers of 50 years ago. If you read their papers they were aware of superior materials and techniques but didn't pursue them for this reason.
So yes the engineering knowledge the approach was based on is just as valid today as it was then but the readily available materials, manufacture and goals of most here are sufficiently different that no one with a modest grasp of how a speaker cabinet behaves would follow it for a high performance design today. The manufacturers that claim to do so today (Spendor, Harbeth, etc...) are using it for marketing reasons and/or to build speakers for the retro market. These are perfectly valid reasons but they are not going to be well aligned with the objectives of many here that are more likely to be seeking engineering for high technical performance (i.e. low levels of sound radiation from the cabinet).
Well, no it has not. Most loudspeakers still use a cone of cardboard flapping around in some kind of wooden box... What truly new speaker technology did you have in mind?
Agreed on all points, Except the E-type drag coefficient, which is actually higher than a 1969 VW van... 😱
I once made a pair of test boxes for some Audax HM130MO which were a squatter 'homage' to the LS3/5A, which i grew up with in the family home.
I attempted to take the thin wall concept to a somewhat different level. They were made with 6mm MDF and glued using Liquid Nails, during to the limited area for gluing a butt joint, with Aluminium angle for the panel joins.
They were very solid, but not rigid enough (the baffle and rear panel were 12mm material.
In the end, I lined the interior with 12mm plasterboard, cut very closely to size, and generously caulked with the Liquid nails adhesive.
They ended up very rigid, and damped - but with another flaw, the lining was very reflective!
It took a layer of wool carpet and heavy felt to reduce the reflections back through the paper cone, to an acceptable level, but I was never really happy with the finish cosmetically, or the haphazard construction.
But the dissimilar laminate approach, definitely worked to dissipate panel vibration, as did the very rigid corner construction.
If I were to repeat the experiment, I'd probably choose something like 4mm ply, 4mm aluminium plate, with something like PU plate/Nylon66/EPA foam sandwiched in the middle.
Then the problem remains, coupling at panel edges, defeating the whole concept, or....just accepting a rubbish looking exposed panel edge.
I attempted to take the thin wall concept to a somewhat different level. They were made with 6mm MDF and glued using Liquid Nails, during to the limited area for gluing a butt joint, with Aluminium angle for the panel joins.
They were very solid, but not rigid enough (the baffle and rear panel were 12mm material.
In the end, I lined the interior with 12mm plasterboard, cut very closely to size, and generously caulked with the Liquid nails adhesive.
They ended up very rigid, and damped - but with another flaw, the lining was very reflective!
It took a layer of wool carpet and heavy felt to reduce the reflections back through the paper cone, to an acceptable level, but I was never really happy with the finish cosmetically, or the haphazard construction.
But the dissimilar laminate approach, definitely worked to dissipate panel vibration, as did the very rigid corner construction.
If I were to repeat the experiment, I'd probably choose something like 4mm ply, 4mm aluminium plate, with something like PU plate/Nylon66/EPA foam sandwiched in the middle.
Then the problem remains, coupling at panel edges, defeating the whole concept, or....just accepting a rubbish looking exposed panel edge.
I designed an LS3/5A homage using 9mm BB plywood for all sides except front and had a white oak internal frame of 5/8in square battens inside that were fastened to the panels with strong PL Premium at the corners but compliant latex caulk along the lengths of the battens. The removable front baffle is fastened to the ledger of battens with foam rubber weatherstripping to absorb vibration. The panels are all lined with foil covered mass loaded butyl sheets to dampen vibrations and covered with eggcrate foam. The whole cabinet is filled with polyfill wool to extend the bass deeper by increasing the virtual volume. The design details were inspired by what the BBC did based on their white paper. The Young’s modulus of white oak is similar to beech. I could not obtain beech readily here. Anyhow, I have to say that it works wonderfully well. These speakers use very basic and readily available Dayton drivers and my own asymmetric crossover that resembles a passive Harsch at 5500Hz with an inverted tweeter.
RST28F and DC130A Foamcore Homage to LS3/5A
Here is detail of internal construction with 9mm BB panels (mitered corner construction), oak battens, Noico foil, and eggcrate foam, here Vunce was trying to see where to mount the XO board borrowed from another project but worked out well with a few mods of jumpers and placement.
RST28F and DC130A Foamcore Homage to LS3/5A
Here is detail of internal construction with 9mm BB panels (mitered corner construction), oak battens, Noico foil, and eggcrate foam, here Vunce was trying to see where to mount the XO board borrowed from another project but worked out well with a few mods of jumpers and placement.
Attachments
Last edited:
A point of contention:
European Beech was used for its combination of attributes, strength and flexibility can indeed be approximated In other woods - but its stability is one of the less obvious reasons it was used, and the real reason that Euro Beech is priced so highly.
I cannot speak for the stability of US white oak in comparison, but yellow oak is IIRC less stable, especially with humidity changes.
Maple would likely be a better choice, but it may not matter, if you don't experience any change in humidity (this side of the pond, it might be an issue)
European Beech was used for its combination of attributes, strength and flexibility can indeed be approximated In other woods - but its stability is one of the less obvious reasons it was used, and the real reason that Euro Beech is priced so highly.
I cannot speak for the stability of US white oak in comparison, but yellow oak is IIRC less stable, especially with humidity changes.
Maple would likely be a better choice, but it may not matter, if you don't experience any change in humidity (this side of the pond, it might be an issue)
George,
Do you have any info on how cabinets with 3 layered walls including one "soft" layer are constructed?
One plan of action can be:
Step-1: Make the exterior cabinet (with 16 mm MDF as was used in your test)
Step-2: Line the inside with bitumen (4 mm)
But when it comes to Step-3, how to fix the inside panels (4 mm ply) to the bitumen?
I loved that car. Years ahead of its time…
A buddy had an E-Type… not my favorite.
Taste for cars is as widely varying as are speakers ;^D
dave
@Juhazi I tried to use the formula for calculating panel resonance that You linked to but I got odd results, it might be I have messed with orders of magnitude as I got 5,95 Hz for the front and back panels and 9,01Hz for side panels. 595 Hz and 901 Hz would make a lot more sense 🙂 Could You post what coefficients did You use?
I confirmed that my local supply has alder ply, not beech, as beech is pretty hard to get. I found some measurements online regarding modulus of elasticity for alder ply, which I assumed at some 7500 N/mm2 or 1,09 mili psi. I also had problem with Poisson number, I googled that someone used 0.3 for ply in general. Its not very accurate but I just need a ballpark number to now if I should try to push resonances below or above midwoofer's passband.
How to account for damping in this formula?
Edit: I recalculated resonances for beech ply and they are some 15% higher in frequency compared to alder ply.
I confirmed that my local supply has alder ply, not beech, as beech is pretty hard to get. I found some measurements online regarding modulus of elasticity for alder ply, which I assumed at some 7500 N/mm2 or 1,09 mili psi. I also had problem with Poisson number, I googled that someone used 0.3 for ply in general. Its not very accurate but I just need a ballpark number to now if I should try to push resonances below or above midwoofer's passband.
How to account for damping in this formula?
Edit: I recalculated resonances for beech ply and they are some 15% higher in frequency compared to alder ply.
Last edited:
It was, but I prefer the landcrabs. Especially the Wolseley variation.
Nor mine. But it's not any old E-type I was talking about: it was the Lightweight Low-Drag. Only two were ever built for privateer racing. For anyone who hasn't had the pleasure -attached. And their CD was down at around 0.3 once Malcolm Sayer had finished with them, although it was the aerodynamic stability that was their party-piece.
Attachments
I had a land crab. My dear brother drove it over a median and wrote it off. Sort of like a Tardis.
dave
dave