Serious question - would it "compromise" the cabinet - or just be a different cabinet - i.e. not a replica LS3/5a? Would a stiffer cabinet sound inferior? Is cabinet resonance an important part of the LS3/5a?
Sorry, I am WELL out of my depth here. What do you mean by "The B110/B139 support a lower crossover than typically used" and "Line Level crossover"?
MDF is a bad material for building speakers. The BBC did a whole lot of research coming up with the box they did. Still, i believe a better cab could be built, but not with MDF.
Sorry, I am WELL out of my depth here. What do you mean by "The B110/B139 support a lower crossover than typically used" and "Line Level crossover"?[/QUOTE]
Typically XOed at about 400 Hz, the B139 would like to be XOed lower to further roll-out the 1 Khz resonance, and the B110 easily goes down to 100 Hz. A line level XO (active or passive) and bi-amping at or a bit below the bafllestep makes a lot of sense (in the neighborhood of 200-250 Hz)
dave
Line level crossover is another way of saying 'active', ie the power amps connected directly to the drive units, one amp per driver. Arguably if you did that with a KEF 105 clone you might have something better than the original.
It's a good idea to build the B110/T27 enclosure as a separate unit as you can optimise it. The bass box could sit underneath or alongside or, as in the case of the Linkwitz, hidden. A 100Hz crossover, as Dave suggests, would facilitate this. I can't off-hand recall how large the Linkwitz bass box was but use two halves to sit under the mid-treble box.
If building a LS3/5a clone don't be tempted to change the spec too much. The BBC found that thicker enclosure walls didn't improve the sound. The thin wall is emphatically not about adding resonance but damping it and getting rid of it. I have a pair of LS3/5a 9mm wall cabinets here fitted with Jordan Eikona units and it's the most solid and 'quietest' cabinet I've used.
It's a good idea to build the B110/T27 enclosure as a separate unit as you can optimise it. The bass box could sit underneath or alongside or, as in the case of the Linkwitz, hidden. A 100Hz crossover, as Dave suggests, would facilitate this. I can't off-hand recall how large the Linkwitz bass box was but use two halves to sit under the mid-treble box.
If building a LS3/5a clone don't be tempted to change the spec too much. The BBC found that thicker enclosure walls didn't improve the sound. The thin wall is emphatically not about adding resonance but damping it and getting rid of it. I have a pair of LS3/5a 9mm wall cabinets here fitted with Jordan Eikona units and it's the most solid and 'quietest' cabinet I've used.
I'm sure you are right about MDF. However, it seems to be used a lot, even in supposedly decent speakers and I suspect that it was not available as a material when the BBC researched and designed the LS3/5a?
Sadly, I am once again well out of my depth but if the B110 could sensibly take over from the B139 at a lower frequency, surely the people who used that combination in the past (Kef, Rogers, IMF, et al) would have followed that route?
Now you have really lost me - sorry 🙁 I don't think I really understand the difference between active and passive crossovers (actually I'm sure I don't) and nor do I really understand the effect or efficacy of bi-amping - I would speculate that that relates to power and the effect of some sort of interaction between the Amp & Speaker(s) and is really getting into almost inaudible, esoteric differences?
Please don't get me wrong, I find your comments stimulating and I do appreciate them; it's just that I am way behind your understanding and clearly have a LOT to understand and learn - so thanks for them 🙂
Many thanks for that, much as I had suspected, like a sub-woofer in a TV sound system then which has to be connected to mains.
As a matter of interest, I understand that bass is pretty non-directional. In a domestic setting, how reasonable would it be to have a single bass cabinet, combining the low frequency signal from the left and right channel?
To be honest, I would not be too concerned about building a LS3/5a clone, just a reasonable utilisation of the B139, B110, T27 and 4001G in two or possibly more domestically acceptable enclosures giving a clean, faithful sound.
As I understand it, the LS3/5a was designed as a speaker to be used in Outside Broadcast vans and in cramped studios. It had to be small and light(ish) and based on the components and materials available at the time. I gather that it is extensively damped internally? I have downloaded and read the BBC LS3-5a White Paper and also a BBC paper on "Factors in the design of loudspeaker cabinets". I can't say that I understood the maths but I did get some idea as to design considerations. I don't know whether the BBC have ever done any recent work on designing loudspeaker cabinets but I did notice that they were using some Yamaha speakers at the Cadogan Hall last year.
ps - all good, fascinating stuff - please keep it coming!
The BBC found that damping pads, eg bitumen, become less and less effective as panels are made thicker.
The LS3/5a design was specifically for an OB monitor but surprised the designers by performing better than expected. As David says, their research found that thicker wood panels (18mm) were no more effective than 9mm but required 18mm of damping panel, whereas the 9mm panel required 9mm. Cost and weight made it sensible to use thinner panels. MDF wasn't around then and it's possible it might work better than ply for this specific use. I've built a 9mm MDF enclosure with 3 layers of 3mm pads and it works well. Harbeth now use MDF in their BBC-inspired cabinets.
The department responsible for loudspeaker research within the BBC was closed down in the Birt era but they still have a technical department. Nowadays it is more concerned with researching codecs, surround and similar, probably finding it easier to buy in commercial units. If you want to see what they're up to, have a look at the BBC engineering blog - they've been doing some surround experiments during the recent Proms
BBC - Blogs - BBC Radio 3 - BBC Proms in 4.0
A single bass cabinet works if the crossover point is low enough. 70Hz or below certainly works here. 100Hz is probably the highest you'd want it. If done correctly, the bass appears to come from the satellite speakers. It gives freedom to place the bass units in a corner and get an increase in available volume.
Back on KEF TL designs, I found the BJ Webb design on the Falcon site which some of you may find interesting
http://www.falconacoustics.co.uk/downloads/Webb R50 Construction.pdf
I beieve the Webb design was the basis of the Cambridge Audio R50.
Here's a fun vintage guide by Raymond Cooke
http://www.kef.com/uploads/files/en/museum_pdf/60s/Loudspeaker_Enclosure_Designs.pdf
The department responsible for loudspeaker research within the BBC was closed down in the Birt era but they still have a technical department. Nowadays it is more concerned with researching codecs, surround and similar, probably finding it easier to buy in commercial units. If you want to see what they're up to, have a look at the BBC engineering blog - they've been doing some surround experiments during the recent Proms
BBC - Blogs - BBC Radio 3 - BBC Proms in 4.0
A single bass cabinet works if the crossover point is low enough. 70Hz or below certainly works here. 100Hz is probably the highest you'd want it. If done correctly, the bass appears to come from the satellite speakers. It gives freedom to place the bass units in a corner and get an increase in available volume.
Back on KEF TL designs, I found the BJ Webb design on the Falcon site which some of you may find interesting
http://www.falconacoustics.co.uk/downloads/Webb R50 Construction.pdf
I beieve the Webb design was the basis of the Cambridge Audio R50.
Here's a fun vintage guide by Raymond Cooke
http://www.kef.com/uploads/files/en/museum_pdf/60s/Loudspeaker_Enclosure_Designs.pdf
Last edited:
Passive crossover is
Preamp - Amp - crossover (passive components) - speaker system
Active is
Preamp - active (powered) crossover - amp - individual speakers in system
The big advantages of active are:
1. Greater control of each drive unit (the amplifier is directly connected so damping is better)
2. More flexibility - easier to design higher order (24dB/octave and higher) crossovers and introduce response tailoring
3. Amplifiers don't have to be as big, they're working into an easy impedance load compared to (some) passive crossovers
You'd need 3 or 4 amplifiers for the full KEF/Coles set up but the individual amplifiers could be smaller. Because you don't have the power losses in a passive crossover, it's usually reckoned that an active system of (say) 4 50 watt amplifiers (200 total) would require a 400 watt amplifier into a passive system. Plus you can tailor the size of amplifier, so bigger for bass, etc. Given the cost of high-grade passive crossover components and the availability of good quality chip amps and class-D amps, going active may work out the same or cheaper than passive for a complex system.
More reading here 🙂
BiAmp (Bi-Amplification - Not Quite Magic, But Close) - Part 1
(Rather confusingly, he calls it bi-amplification.)
Preamp - Amp - crossover (passive components) - speaker system
Active is
Preamp - active (powered) crossover - amp - individual speakers in system
The big advantages of active are:
1. Greater control of each drive unit (the amplifier is directly connected so damping is better)
2. More flexibility - easier to design higher order (24dB/octave and higher) crossovers and introduce response tailoring
3. Amplifiers don't have to be as big, they're working into an easy impedance load compared to (some) passive crossovers
You'd need 3 or 4 amplifiers for the full KEF/Coles set up but the individual amplifiers could be smaller. Because you don't have the power losses in a passive crossover, it's usually reckoned that an active system of (say) 4 50 watt amplifiers (200 total) would require a 400 watt amplifier into a passive system. Plus you can tailor the size of amplifier, so bigger for bass, etc. Given the cost of high-grade passive crossover components and the availability of good quality chip amps and class-D amps, going active may work out the same or cheaper than passive for a complex system.
More reading here 🙂
BiAmp (Bi-Amplification - Not Quite Magic, But Close) - Part 1
(Rather confusingly, he calls it bi-amplification.)
I discovered this interesting quote from the SpeakerTalk forum (although can't account for how accurate it is):
"B. J. Webb a retired schoolteacher (and perfect gentleman) designed the R50, which was bought a a complete design according to Stan Curtis, a former director of Cambridge Audio.
Burt's DIY version may be found in The Audio Amateur issue 1/75.
Chris Rogers version is a “copyright avoidance” design not a “proper job”!
"B. J. Webb a retired schoolteacher (and perfect gentleman) designed the R50, which was bought a a complete design according to Stan Curtis, a former director of Cambridge Audio.
Burt's DIY version may be found in The Audio Amateur issue 1/75.
Chris Rogers version is a “copyright avoidance” design not a “proper job”!
It doesn't seem entirely surprising that "damping pads become less and less effective as panels are made thicker", I would have expected that thicker panels would resonate less anyhow, especially if braced. Was this BBC research ever published and if so, do you have any idea where I might find the details?
The B J Webb design is very interesting. The B139 TL duct is more convoluted than in the R50 and I see that the B110 is placed in a sort of "mini TL". I like the idea of the angled "deflector" panels at the top and bottom of the rear panel and the fact that he advocates lots of braces to avoid resonance. I notice that he uses ¾" chipboard rather than any form of plywood.
I had a look at Wilmslow Audio's Jordan Eikona 2 based Transmission Line speaker. It works out at over £1,000 for a pair in unfinished MDF. Looking on eBay based on recent sales of Concertos, KefKit 3s & R50s, I reckon I would get £200 tops before commission, etc. for my B139, B110, T27, 4001G & Crossover. I would like to think that I could put together something along the lines of the R50 / B J Webb TL for less than £300 in total and then I could spend the difference on a few CDs 😉
The reference to the BBC Proms is interesting, I have a Gallery Season Ticket for the Proms and I have noticed two things - other than some incredible music. At about a third of the concerts there has been a substantial mixing desk at the back of the Arena and on a couple of occasions I have had the feeling that they have been "playing about" with some of the speakers in the hall giving strange effects in terms of the direction from which some instruments (percussion and wind/brass) have come. I haven't listened to any of it on iPlayer and wouldn't have a setup which would work with surround sound.
Last edited:
They don't resonate less per se, the resonant BW and amplitude are different. For a given material type, it will be shunted lower. Not necessarily a good thing sans multiple qualifications since, for e.g., this can put panel resonant modes right in the middle of a bass enclosure's functional BW, where the most energy is available to excite them.
Factors in the design of loudspeaker cabinets. - Publications - BBC R&D
Deflector panels don't do much harm in near aperiodic TLs like the ones in question here, since the back wave is being almost completely damped out. Aside from one, say, behind the driver though, they don't do much good either since LF wavelengths are too long to be significantly affected while the higher ones are rapidly attenuated by the line damping.
Bracing panels stiffens the panels, thereby raising their resonant BW -if done thoroughly, this can shunt the panel modes sufficiently high to avoid significant excitation.
Thanks for that - final paragraph on page 19 (or 25 of 29 in .PDF speak). Any idea what "Mutacell" is? Presumably something akin to Dedsheet?
Edit: Just had a quote of £280 for "18mm & 25mm MDF cut as list and two panels to be CNC routered with four holes for drive units and one port". More than I had hoped for.
Last edited:
I believe Mutacell was a propriety / brand of damping sheet material available at the time. The name has since disappeared. Essentially, the panel damping material; plenty of equivalents in terms of bitumen sheet etc. Not to be confused in function with the acoustic damping / wadding / stuffing / foam etc. which has a different job.
£280 for all panels CNC cut for a pair of speakers? Not all that bad for a one-off job like that.
£280 for all panels CNC cut for a pair of speakers? Not all that bad for a one-off job like that.
Last edited:
The cost of supplying and cutting the 18 & 25 MDF is £125, the rest (£155) is for CNC cutting and routering the various openings! Sadly there is no way I can do this work myself - I may know a man who can - I will check.
Quite a price difference there; says loads about Wilmslow Audio 🙁
Thicker panels are stiffer and shift resonance upwards, I think. This can make the resonance far more audible
The old BBC R&D reports have a wealth of information on cabinet design, room design and other acoustics topics
Is the R50 a true TL or is it a simplified and shortened design as used in later TDLs like the RTL and T-Line series?. The problem with the big IMF and TDL Studios was that they were huge
Not automatically. Depends on context. Thicker panels are also higher mass which to an extent holds the resonant BW down a little despite the increased stiffness.
Pushing panel resonant BW higher in frequency is actually quite effective for bass enclosures since it's a practical impossibility to shunt them below the operating BW of the box, where the maximum amount of energy is available to excite them. If you can push them high enough (quite practical), then the remnants of what end up being excited are quite easy to damp out. The BBC's thin-wall construction (relatively light, moderately stiff panels heavily damped) also works well, but is not always very practical for some cabinet types.
Depends what you call a 'true TL'. A true TL per se would be a line designed with the sole purpose of providing the flattest possible impedance load, zero other considerations. Very few indeed (including the line Bailey himself described) actually fall under that definition; they mostly fall into the mildly resonant category. TLs by nature tend to be large, especially if used with reasonably sized woofers since the latter's Vas, Q requirements have to be accounted for.
Materials cost and difficulty are another way that sealed boxes win out.
I don't have access to Martin King's Mathcad spreadsheets (Mac user) but have played around with Augspurger's TL tables from Speaker Builder. They're quiteinteresting because they don't go for maximum bass extension but focus on a better balance between extension, power handling and efficiency. Plus there's an 'extended' table if you want a bit more bass - the trade off being a larger cabinet. Once you've done the basics, you can design the cabinet how you like (within reason) and modify it for ease of construction.
Of course, you'd have to know which B139 …
Bass Units
I don't have access to Martin King's Mathcad spreadsheets (Mac user) but have played around with Augspurger's TL tables from Speaker Builder. They're quiteinteresting because they don't go for maximum bass extension but focus on a better balance between extension, power handling and efficiency. Plus there's an 'extended' table if you want a bit more bass - the trade off being a larger cabinet. Once you've done the basics, you can design the cabinet how you like (within reason) and modify it for ease of construction.
Of course, you'd have to know which B139 …
Bass Units
Aren't sealed boxes significantly less efficient than TLs?
I just love the Kef clarity over B139 models 😀
Depends where you measure and what kind of TL. Above the operating bandwidth of the box (ie up where you XO them, they are exactly the same efficiency.
dave
- Status
- Not open for further replies.