I'm just starting out discovering the capabilities of the KEF B110 & T27 drivers having recently been given a pair of KEF Concerto Speakers.
Fist impressions are positive given that the speakers are all original and 43 years old!
Work on the speakers so far has been restricted to a deep clean and getting the crossover connectors to work.
A pair of Falcon Acoustics 5ab crossovers have been ordered
KEF CONCERTO UPGRADE CROSSOVER No.5aB MF and HF Recommended Upgrade
The crossover upgrade is interesting as it has a 18dB roll off for the Top end of the B110 and (QUOTE) a simplified version of the acoustic Butterworth section for the T27
Not quite sure what that involves yet!
Amplifier is my Rauch DVT50 rebuilt by Ben Duncan Research.
This is firmly supported on a wall mounted equipment rack to minimise any microphonic effects. The idea that electronics are not effected by microphony is definitely something that I would completely disagree with.
Fist impressions are positive given that the speakers are all original and 43 years old!
Work on the speakers so far has been restricted to a deep clean and getting the crossover connectors to work.
A pair of Falcon Acoustics 5ab crossovers have been ordered
KEF CONCERTO UPGRADE CROSSOVER No.5aB MF and HF Recommended Upgrade
The crossover upgrade is interesting as it has a 18dB roll off for the Top end of the B110 and (QUOTE) a simplified version of the acoustic Butterworth section for the T27
Not quite sure what that involves yet!
Amplifier is my Rauch DVT50 rebuilt by Ben Duncan Research.
This is firmly supported on a wall mounted equipment rack to minimise any microphonic effects. The idea that electronics are not effected by microphony is definitely something that I would completely disagree with.
Surv1v0r - The simplest option might be to contact Falcon Acoustics.
They have a lot of experience with KEF drivers and have recently started re manufacturing the B110. Give them the details of what cabinet you're using (especially baffle size) and see what options they suggest. They may have a simpler option than the LS3/5A crossover and can be relied on to offer sensible options regarding component quality.
They have a lot of experience with KEF drivers and have recently started re manufacturing the B110. Give them the details of what cabinet you're using (especially baffle size) and see what options they suggest. They may have a simpler option than the LS3/5A crossover and can be relied on to offer sensible options regarding component quality.
Last edited:
Re active electronics in speakers, this link may be of interest
Q. Is it worth isolating my speakers and other equipment?
Q. Is it worth isolating my speakers and other equipment?
The B110 and T27 are appropriate drivers for people that want to DIY 40 year old designs like the LS3/5A, Daline, etc... They are not appropriate drivers for better sounding designs because after 40 years they are expensive and don't perform as well as modern drivers. Not wildly but sufficiently that nobody capable of designing a better sounding speaker would opt to use them today.
If you lack the skills to design a speaker and the desire to acquire them then I think you are probably stuck with using your drivers in 40 year old designs or selling them and using modern drivers in a modern design to get a better sounding speaker.
I must admit I am somewhat confused by the turn my thread has taken.No way. 100W is a LOT! You're not going to run a serious HiFi amplifier off a little 12V or 24V laptop or phone wall wart.
<snip>
Quite what this has to do with BBC mini-monitors, I can't remember.
However, to clarify, on this thread previously . . .
- I am considering placing the B110s & T27s from recently acquired KefKit 3 DIY speakers in Atkinson MiniLine Transmission Line cabinets (43 cms high x 22.5 cms wide x 25 cms deep external - 12 mm ply walls - line length approx 120 cms).
- I asked for advice on crossovers, the original design used the Isophon KK8 rather than theT27 tweeter and the crossover seems pretty basic; I wondered whether to use the crossover design recommended by Kef on the B110 & T27 spec sheets or perhaps a LS3/5a crossover sourced from Falcon Acoustics.
- It was suggested that since I was planning to start from scratch I should consider active crossovers. I sought advice & guidance from the worthy burghers of DIYaudio.
- Much confusion later I am pretty much convinced that I should continue down the passive crossover route.
Since I will be dismantling KefKit 3 based speakers which include the DN12 SP1004 crossover, as an interim measure would it be possible to take the output destined for both the B139 & the B110 and route it to the B110 -or- could I remove some component(s) to eliminate the bottom end cut-off for the B110?
Last edited:
You could either modify the SP1004 by bypassing the 30µ cap and dissconnect the 4 mH coil.
http://www.hifiloudspeakers.info/Anatomy/SpeakerSystems/Concerto/ConcertoCrossoverSchematic.jpg
An other option would be to build the 18 dB leg to the T27 and the 12 dB filter to the B110
I have plans to build the MiniLine but then with the cabinet hight extended and the depth increased to accomodate 12mm MDF or plywood in all parts including the interal dividers. If I ahde the KEFkit 3 I would be tempted to build IMF style transmission line speakers if I had the space.
http://www.hifiloudspeakers.info/Anatomy/SpeakerSystems/Concerto/ConcertoCrossoverSchematic.jpg
An other option would be to build the 18 dB leg to the T27 and the 12 dB filter to the B110
I have plans to build the MiniLine but then with the cabinet hight extended and the depth increased to accomodate 12mm MDF or plywood in all parts including the interal dividers. If I ahde the KEFkit 3 I would be tempted to build IMF style transmission line speakers if I had the space.
Can you elaborate on the 18dB T27 & 12 dB B110 changes please?
If you have a cutting plan for the MiniLine I would love to see a copy of that. Did you make it double skinned as the original design?
I can't get away with building a large speaker or I would build the B J Webb design - also, doesn't the IMF design use HF1300s and/or HF2000s?
To get a rough idea how the Atkinson will sound with a B110/T27 based xover from a different design, you need to determine the difference in diffraction between the Atkinson cabinet's baffle and the baffle of the design you take the xover from.
The Atkinson baffle:
Atkinson Mini TL HFN Nov78
will be much larger than almost any other design and so will have more on axis output in the upper bass to lower mids, on axis, than the xover intended.
Similarly, the tweeter diffraction will differ and that will have an impact too.
Use this to quickly model both cabinets and then look at the difference for a rough indication of the impact
Home of the Edge
This doesn't tell you dB for dB how the sound will change since the on axis doesn't tell you the whole story.
It will probably indicate though that to get the same sort of response intended by the xover for the B110, you'll need to make the inductor in series with the B110 smaller and then adjust the other components around it to get a similar roll off.
Using the existing crossover designed for a different tweeter will give you indeterminate unintended results, unless the two tweeters are very similar in acoustic response and electrical impedance.
But it sounds like you aren't comfortable modeling the xover in a free sw program. If you can't do this, you'll be blind and it will just be guess work.
The Atkinson baffle:
Atkinson Mini TL HFN Nov78
will be much larger than almost any other design and so will have more on axis output in the upper bass to lower mids, on axis, than the xover intended.
Similarly, the tweeter diffraction will differ and that will have an impact too.
Use this to quickly model both cabinets and then look at the difference for a rough indication of the impact
Home of the Edge
This doesn't tell you dB for dB how the sound will change since the on axis doesn't tell you the whole story.
It will probably indicate though that to get the same sort of response intended by the xover for the B110, you'll need to make the inductor in series with the B110 smaller and then adjust the other components around it to get a similar roll off.
Using the existing crossover designed for a different tweeter will give you indeterminate unintended results, unless the two tweeters are very similar in acoustic response and electrical impedance.
But it sounds like you aren't comfortable modeling the xover in a free sw program. If you can't do this, you'll be blind and it will just be guess work.
With regard to the 18 and 12 dB legs. The tweeter uses a third order filter (cap-coil-cap). I would use plastic film caps and a standard coil of the right inductance the serial resistance is not that critical.
The woofer uses a second order (Coil-Cap). Also here nothing fancy just keep the coil resistance reasonably low.
Yes I know that the coil resistances affect the Q of the high or low pass fliter but the driver is probably changed due to 40 years of aging so any filter is interacting with a different driver anyhow. Four coils and six caps will not cost you much. If you think that the drivers and cabinet show promise you can start thinking about more elaborate crossovers.
No I have not built the TLS so far and I have at least two other projects to finish before I take on the mini TLS. If you have not heard and liked the drivers do try them with a decent proven crossover before going all over board with complicated expensive crossovers.
The woofer uses a second order (Coil-Cap). Also here nothing fancy just keep the coil resistance reasonably low.
Yes I know that the coil resistances affect the Q of the high or low pass fliter but the driver is probably changed due to 40 years of aging so any filter is interacting with a different driver anyhow. Four coils and six caps will not cost you much. If you think that the drivers and cabinet show promise you can start thinking about more elaborate crossovers.
No I have not built the TLS so far and I have at least two other projects to finish before I take on the mini TLS. If you have not heard and liked the drivers do try them with a decent proven crossover before going all over board with complicated expensive crossovers.
A few years back, I redesigned a speaker with 40 yr old B110s. DrBoar is dead on about it ageing. Compliance aged to about half its advertise value (see below)
I was recommending an inductor change for a different reason, and that was baffle size. For the same driver, speakers designed for nominally flattish response on axis require larger low pass series inductor for smaller baffles, to compensate more more baffle diffraction. This invariably then calls out a different cap for a given xover point. Of course, the cap/inductor combo sets the Q so baffle size becomes a driving element in limiting xover corner shapes. You can get around this by adding resistors in the cap leg, or traps here and there but its well beyond the capabilities of the requester
Measured Data
"Method: Box-loaded (7.50 liters)"
"DCR mode: Measure (-0.10 ohms)"
"Area (Sd): 92.00 sq cm"
"Series resistance: 75.75 ohms"
"Stimulus level: 0.49 volts"
0.550 "RMSE-free Ohms"
50.228 "Fs Hz"
6.817 "Re Ohms[dc]"
110.308 "Res Ohms"
6.733 "Qms "
0.416 "Qes "
0.392 "Qts "
0.511 "L1 mH"
0.828 "L2 mH"
2.989 "R2 Ohms"
0.502 "RMSE-load Ohms"
11.801 "Vas liters"
10.114 "Mms(Sd) grams"
992.738 "Cms(Sd) uM/Newton"
7.231 "Bl(Sd) Tesla-M"
87.385 "SPLref dB"
0.006 "Rub-index "
Advertised Parameters
38 +/-2"Fs Hz"
6.7 +/-0.2"Re Ohms[dc]"
6.7 "Qms "
0.33 "Qes "
0.31 "Qts "
0.45 "L1 mH"
23.6 "Vas liters"
10.5 "Mms(Sd) grams"
1800 "Cms(Sd) uM/Newton"
7.1 "Bl(Sd) Tesla-M"
I was recommending an inductor change for a different reason, and that was baffle size. For the same driver, speakers designed for nominally flattish response on axis require larger low pass series inductor for smaller baffles, to compensate more more baffle diffraction. This invariably then calls out a different cap for a given xover point. Of course, the cap/inductor combo sets the Q so baffle size becomes a driving element in limiting xover corner shapes. You can get around this by adding resistors in the cap leg, or traps here and there but its well beyond the capabilities of the requester
Measured Data
"Method: Box-loaded (7.50 liters)"
"DCR mode: Measure (-0.10 ohms)"
"Area (Sd): 92.00 sq cm"
"Series resistance: 75.75 ohms"
"Stimulus level: 0.49 volts"
0.550 "RMSE-free Ohms"
50.228 "Fs Hz"
6.817 "Re Ohms[dc]"
110.308 "Res Ohms"
6.733 "Qms "
0.416 "Qes "
0.392 "Qts "
0.511 "L1 mH"
0.828 "L2 mH"
2.989 "R2 Ohms"
0.502 "RMSE-load Ohms"
11.801 "Vas liters"
10.114 "Mms(Sd) grams"
992.738 "Cms(Sd) uM/Newton"
7.231 "Bl(Sd) Tesla-M"
87.385 "SPLref dB"
0.006 "Rub-index "
Advertised Parameters
38 +/-2"Fs Hz"
6.7 +/-0.2"Re Ohms[dc]"
6.7 "Qms "
0.33 "Qes "
0.31 "Qts "
0.45 "L1 mH"
23.6 "Vas liters"
10.5 "Mms(Sd) grams"
1800 "Cms(Sd) uM/Newton"
7.1 "Bl(Sd) Tesla-M"
But has this "helpful" input got you anywhere?I do genuinely appreciate the comprehensible, helpful input from the likes of Colin & System7
Currently you want to use the wrong tweeter in a 40 year old design which therefore requires a new crossover designed for the purpose to sound good. But you are not going to do this. Instead you intend to cobble together something from crossovers for different speakers that happen to use your drivers. Since the cabinet among other things determines the values needed for a reasonable crossover the quality of the result is not difficult to predict.
Perhaps sound quality that follows from using an appropriate crossover for the drivers and cabinet is not particularly important. Perhaps what is more important is having a bit of fun with a 40 year old design using 40 year old parts and being happy with it not sounding bad rather than it sounding good? If this were known to be the case then I expect most of the "unhelpful" posts like this one that are pointing out that you are not going to meet stated objectives like better than the LS3/5A will disappear.
It's been an interesting thread and there has been a lot of very helpful input from a number of people. Andy and DDR are spot-on about what can be achieved with the B110/T27. DDR's measured parameters are particularly revealing.
The classic design using these drivers is the LS3/5a. It won't be the cheapest option (the crossover is complex and expensive) but will produce the best results in the circumstances. What won't be entirely predictable is any change in the driver parameters and how the crossover might need adjusting for that.
There are re-manufactured B110/T27 units available from Falcon and there are better drive units available, both for the LS3/5A clones and for other use. Stirling Broadcast, for instance, sells a certified LS3/5A without a KEF driver in sight.
It's really a case of how much you want to use those drive units. As original units, they might sell well enough to cover the cost of more modern, easier to use units which require a less complex crossover and ultimately achieve better results.
Drive units have improved a lot of the last 40 years.
Doing something vintage will be fun and quite satisfying but the results could fall below what is achievable today with more modern drive units.
The classic design using these drivers is the LS3/5a. It won't be the cheapest option (the crossover is complex and expensive) but will produce the best results in the circumstances. What won't be entirely predictable is any change in the driver parameters and how the crossover might need adjusting for that.
There are re-manufactured B110/T27 units available from Falcon and there are better drive units available, both for the LS3/5A clones and for other use. Stirling Broadcast, for instance, sells a certified LS3/5A without a KEF driver in sight.
It's really a case of how much you want to use those drive units. As original units, they might sell well enough to cover the cost of more modern, easier to use units which require a less complex crossover and ultimately achieve better results.
Drive units have improved a lot of the last 40 years.
Doing something vintage will be fun and quite satisfying but the results could fall below what is achievable today with more modern drive units.
Yes, absolutely.
You say "Since the cabinet among other things determines the values needed for a reasonable crossover . . ."; perhaps you can help me to understand the significant, discernible ways in which a cabinet impacts on the tweeter side of a crossover? I think I asked something along these lines a while back and didn't get any response.
I hope that I have made it fairly apparent that I do not understand either cabinet or crossover design.
- I have a KefKit 3 in a pretty shabby cabinet.
- Based on the tweeter and bass/mid-range units from this I would like (need) to build something smaller that would sound good.
- The LS3/5a seemed the obvious choice but people suggest that they lack bass and are somewhat tiring to listen to.
- Falcon Acoustics who specialise in the LS3/5a suggest that a slightly larger cabinet would make better use of the B110 & T27.
One of these suggestions was that I might have a look at the Atkinson Mini Transmission Line speakers; they seem to have potential in that they are an acceptable sort of size, were designed my someone who is reputed to have known what he was doing and use my B110. Since I am using different but well respected tweeters which I happen to have readily available I would like any advice I can get on passive crossovers I could use with them. Based on recommendations here I have considered and researched active crossovers which might well be more flexible but I will almost certainly not employ them because they add a whole new layer of complexity and confusion.
I am truly sorry if my ignorant approach offends the sensibilities of those with greater knowledge and experience but I am still grateful for any help I get from anyone more sympathetic - OK?
You need to model a speaker on different sized baffles to get an idea what that does to the bass response. Most of the effects are below 1kHz. Generally speaking, bigger baffles need less bafflestep correction, which means smaller bass coils in the filter.
You also get some tweeter diffraction effects which are controlled by offset placement on a baffle.
Hence some merit in this design:
Overall, I can't see the LS3/5A had much going for it beyond being exactly what it was designed to be, a small accurate monitor for BBC broadcast vans.
Serious HiFi buffs preferred something heftier. I had a lot of good use out of the similar 3 way Chartwell PM400 monitors. I see no reason why a KefKit 3 couldn't be modified for a different bass unit if the baffle is removable. But a serious project. KEF tweeter filters improved with time too.
You also get some tweeter diffraction effects which are controlled by offset placement on a baffle.
Hence some merit in this design:
Overall, I can't see the LS3/5A had much going for it beyond being exactly what it was designed to be, a small accurate monitor for BBC broadcast vans.
Serious HiFi buffs preferred something heftier. I had a lot of good use out of the similar 3 way Chartwell PM400 monitors. I see no reason why a KefKit 3 couldn't be modified for a different bass unit if the baffle is removable. But a serious project. KEF tweeter filters improved with time too.
Attachments
The "unhelpful" post from DDR talks about this. The diffraction of sound by the cabinet and how to work with it in a crossover is covered in speaker DIY articles such as those here. I am not sure the details are particularly relevant given you do not wish to design a crossover but understanding that a good quality crossover does more than implement text book high and low pass filters probably is given proposals like using a crossover from one speaker, or part of it, in another.
I am truly sorry if my ignorant approach offends the sensibilities of those with greater knowledge and experience but I am still grateful for any help I get from anyone more sympathetic - OK?
Encouraging and sympathetic posts might not be the most useful ones. Can I suggest browsing some of the FAQ links on the RHS of this page.
I had a look at the free software (The Edge) recommended by DDF but it was way beyond me.
The Atkinson design uses the B110 so I hope it would offer pretty good bass for the driver and cabinet size. It appears to have a pretty basic crossover although I am sure it is perfect with the original Isophon KK8 tweeter - I plan to use the T27.
Mr. dB posted details of a modified / improved crossover as used in the JR149 which also used the B110 & T27. The original Kef Coda employed the B110/T27 and seemed to have a pretty basic crossover about which I have only managed to find a few limited details. You posted details of the Kef DN13 SP1017 low pass filter section for the B110 and the high pass filter for the T27. I already have a pair of DN12 SP1004 crossovers which are used with the B139/B110/T27 drivers.
It seems to me that since I am not in a position to design passive crossovers from scratch my obvious solutions are to do one of the following:
- Build crossovers based on the Kef DN13 SP1017 specs.
- Modify my existing DN12 SP1004 crossovers to eliminate the B139 section and the lower cut-off for the B110.
- Purchase and use Falcon Acoustics LS3/5a crossovers.
I say again, I accept that the B110 & T27 are ancient, outdated, worn-out drivers that nobody in their right minds would go out and buy today - I just happen to own a pair of them and I also accept that the reputation of the BBC LS3/5a speakers is no longer justified and that they were only ever suitable for use in the back of a white van. I just want to see if rather than throwing my B110s & T27s into the nearest skip it is possible for me to to build reasonable quality, small(ish) speakers based on the drivers I already own.
As to the tweeter diffraction, I was under the impression that tweeters (which have enclosed backs) tend to be pretty directional and I had therefore assumed that the cabinet and baffle would have little effect on them? I realise that the BBC LS3/5a uses felt squares around the T27s "to remove diffraction effects" which puzzles me.
I have no wish to be difficult or ungrateful - to anyone;. I recognise that my starting point is ideal. As a complete novice I am just hoping for some informed, comprehensible on-topic help.
ps - I am serious when I say that I am grateful to you and Colin for your input, you have both helped me a lot.
The B110 gets more directional as you go up the frequency range. The dome T27 will broadcast sound in a wide pattern at the lower end of its range and only get directional at the extreme top. Domes are (generally) designed to disperse sound over a wide area.
Because of this, most modern designs mount the dome flush with the baffle and take care to eliminate sharp edges at the corner of the cabinets which can cause secondary reflections. You want the sound to diffract round the edge.
The LS3/5A had the entire front baffle recessed so the reflections were treated by placing a square of felt around the tweeter. In effect, high frequency sound to the side (top and bottom) were absorbed by the felt.
Not everyone likes the LS3/5A but some love it. It wasn't just for use in outside broadcast vans, the BBC also used them in their smaller studios. Last time I visited the World Service in their old Bush House offices, there were LS3/5As dotted all over the place.
Building one will be easier than the Atkinson Mini-Line. The LS3/5A isn't a straightforward build to get right, so there is an element of skill involved which will be quite satisfying. Falcon used to extend the size of the box when they cloned it in the 1970s, making it about 22cms deep instead of 16cms.
Because of this, most modern designs mount the dome flush with the baffle and take care to eliminate sharp edges at the corner of the cabinets which can cause secondary reflections. You want the sound to diffract round the edge.
The LS3/5A had the entire front baffle recessed so the reflections were treated by placing a square of felt around the tweeter. In effect, high frequency sound to the side (top and bottom) were absorbed by the felt.
Not everyone likes the LS3/5A but some love it. It wasn't just for use in outside broadcast vans, the BBC also used them in their smaller studios. Last time I visited the World Service in their old Bush House offices, there were LS3/5As dotted all over the place.
Building one will be easier than the Atkinson Mini-Line. The LS3/5A isn't a straightforward build to get right, so there is an element of skill involved which will be quite satisfying. Falcon used to extend the size of the box when they cloned it in the 1970s, making it about 22cms deep instead of 16cms.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.