Photos would be great, i'll manage anyway, for the test i just placed the components without much thought. I'll try not to diverge from the schematic's values, still I would like to experiment. I've been toying with XSim using FRD/ZMA files from speakers that have noting to do with B139/B110/T27 trying to bring up frecuencies between 200-400hz... I imagine this is some kind of heresy but I'm a layman regarding crossovers, it was just to get an idea of all this
Last edited:
That's great, so we'll all learn from what you test. I am curious!
I tried different makes of caps, as I describe in the original post #1. I found that lower ESR caps in parallel (LL, "plain foil" electrolytics, polyester or polypropylene) degrade the sound. this is contrary to what Falcon does, and I am still asking why they go that route, as it clearly changes the balance of the crossover to the worse.
I also experimented while I re-capped the crossovers of my Reference 105. In those old KEF crossovers I'd go with normal, but good quality electrolytics in all positions in parallel to the speakers. In Germany, Mundorf "rough foil" caps with a medium voltage do the job. The same is valid for Fischer & Tausche "rough foil" (ATBI).
I believe that F&T actually supply Mundorf – and they also supply Falcon, at least part of their new ALCAP series. The printing on some of those caps is clearly done in the same way as is on the original F&T and also says ATBI.
What concerns caps in series, some can be low ESR. as I wrote, the input cap for the mid section is of particular importance. so I put a LL ("plain foil") Mundorf there. You'd probably fare even better with a 30uF polypropylene, but the size is far to big to mount it on the original board.
so the baseline is: the idea to swap electrolytics for polypropylenes in an indiscriminate way will throw the crossover off balance. The ESR is part of the circuit, and KEF factored it in. There are some threads here which discuss the fact that you can't easily change a vintage crossover for the better.
So I found that the parallel 7uF (6.8uF in my case) needs to be high-quality but not low ESR – F&T ATBI. An equally mid-ESR Audyn cap sounded much worse. You might counter the lowering of ESR when you use a LL or poly cap if you raise the series resistance in this position. But ESR is not resistance... so this will not a perfect mitigation.
one more thing:
as I wrote before, I have also experience with the Kenwood KA-7300 amp. It is the one I am using most at the moment in one of my systems, together with my KEF 105's.
I found, however, that this amp suffers easily from dirty switches (and it has a lot of them), including the pre-out––main-in switch. Dirty switches reduce bass, among other things.
If you want to judge your speakers on a KA-7300 you should thoroughly clean all switches, ideally buy opening them and cleaning all the contacts. You can also dry to drown them in De-Oxit. But this will work only short term and is not a good way to really cleanse contacts.
One other weak spot in the KA-7300 is the 0.47uF polyester input cap into the main amp. It will reduce bass and introduce phase shifts. This cap is easy to access. I'd swap it against a 1uF (at least) polypropylene. I used 2 0.47 WIMA MKP4's in parallel. Or use a 2uF Panasonic Polypropylene (Mouser) -- they come with long leads, so easy to fit into place.
You might give bridging all the switches in the preamp a thought, and using dedicated cables for pre-out--main-in. You'd then have still the 3-step loudness knob for adding bass. I can tell you how I did it if you wish advice on this.
I found that with these changes the KA-7300 became a very, very potent amp, pretty powerful and extremely clean, but not "cold" at all. Just neutral. Hard to beat.
as I wrote before, I have also experience with the Kenwood KA-7300 amp. It is the one I am using most at the moment in one of my systems, together with my KEF 105's.
I found, however, that this amp suffers easily from dirty switches (and it has a lot of them), including the pre-out––main-in switch. Dirty switches reduce bass, among other things.
If you want to judge your speakers on a KA-7300 you should thoroughly clean all switches, ideally buy opening them and cleaning all the contacts. You can also dry to drown them in De-Oxit. But this will work only short term and is not a good way to really cleanse contacts.
One other weak spot in the KA-7300 is the 0.47uF polyester input cap into the main amp. It will reduce bass and introduce phase shifts. This cap is easy to access. I'd swap it against a 1uF (at least) polypropylene. I used 2 0.47 WIMA MKP4's in parallel. Or use a 2uF Panasonic Polypropylene (Mouser) -- they come with long leads, so easy to fit into place.
You might give bridging all the switches in the preamp a thought, and using dedicated cables for pre-out--main-in. You'd then have still the 3-step loudness knob for adding bass. I can tell you how I did it if you wish advice on this.
I found that with these changes the KA-7300 became a very, very potent amp, pretty powerful and extremely clean, but not "cold" at all. Just neutral. Hard to beat.
now some pics. I relocated the new air core inductor for the tweeter section as the caps of the aB section were to many and to big, so needed all the space in that row. I also put some elements on the solder side of the xover.
it is still to "thick". but as the textile on my grilles is worn out anyway and needs to be replaced I didn't bother with it being a bit stretched. So cosmetically with the grilles on the whole thing is a bit tight.
And please bear in mind that I turned the connectors 180 degrees (the wires are long enough but you'd need to proceed with care).
It pays to thoroughly clean the connectors and the contacts on the boards. I used stiff cardboard drenched in De-Oxit for the connectors.
It might be interesting to know if bass becomes more or better defined if you lower the series resistance of the woofer circuit (LL series cap or even polypropylene –– very, very big can ––, bigger gauge inductor, different wires). I pondered these things but found them too intrusive. In the place where I use these speakers at the moment together with a restored-modified NAD 3140 the lowest, low and mid bass is not a problem at all, to the contrary.
it is still to "thick". but as the textile on my grilles is worn out anyway and needs to be replaced I didn't bother with it being a bit stretched. So cosmetically with the grilles on the whole thing is a bit tight.
And please bear in mind that I turned the connectors 180 degrees (the wires are long enough but you'd need to proceed with care).
It pays to thoroughly clean the connectors and the contacts on the boards. I used stiff cardboard drenched in De-Oxit for the connectors.
It might be interesting to know if bass becomes more or better defined if you lower the series resistance of the woofer circuit (LL series cap or even polypropylene –– very, very big can ––, bigger gauge inductor, different wires). I pondered these things but found them too intrusive. In the place where I use these speakers at the moment together with a restored-modified NAD 3140 the lowest, low and mid bass is not a problem at all, to the contrary.
Attachments
Last edited:
Thanks for sharing your experiences! Very informative.
It's interesting you found electrolytics in parallel to actually perform better than film in this case. Probably you've come across this thread where poster Retrovert swears by film caps against electros in crossoversCROSSOVER CAPACITOR QUESTION ? | Page 2 | Audiokarma Home Audio Stereo Discussion Forums
Today I've been able to test 2 Jantzen Standard 15uf caps (making a 30uf cap) against the old Elcaps in the crossover I'm testing with your (approximate) circuit in the B110 position... the difference was subtle, hard to tell at first, but yes, Jantzens sounded better, but then Elcaps must be +30yrs old! According to the bad rap Retrovert gives to electrolytics I expected a more dramatic difference. I imagine upgrades in several components must add up ending up in more noticiable improvements. I'm not so worried about the size of crossovers as I have them inside the cabinets, so I might even make a new board and glue the components if i end up getting bigger inductors and a film cap for the B110 and maybe another for the B139... anyway I need time to ponder all this before I order the parts... I wanna be sure as with lower gauge inductors and film caps the grand total goes through the roof easily!
It's good to know you can get the Kenwood to another level bridging the preamp but I think I will leave it as is for now as I'm actually using it's preamp, according to the manual when tone/presence controls are not engaged they're out of the circuit (I guess the rest remain in: attenuator, loudness, stereo-mono modes, etc)
It's interesting you found electrolytics in parallel to actually perform better than film in this case. Probably you've come across this thread where poster Retrovert swears by film caps against electros in crossoversCROSSOVER CAPACITOR QUESTION ? | Page 2 | Audiokarma Home Audio Stereo Discussion Forums
Today I've been able to test 2 Jantzen Standard 15uf caps (making a 30uf cap) against the old Elcaps in the crossover I'm testing with your (approximate) circuit in the B110 position... the difference was subtle, hard to tell at first, but yes, Jantzens sounded better, but then Elcaps must be +30yrs old! According to the bad rap Retrovert gives to electrolytics I expected a more dramatic difference. I imagine upgrades in several components must add up ending up in more noticiable improvements. I'm not so worried about the size of crossovers as I have them inside the cabinets, so I might even make a new board and glue the components if i end up getting bigger inductors and a film cap for the B110 and maybe another for the B139... anyway I need time to ponder all this before I order the parts... I wanna be sure as with lower gauge inductors and film caps the grand total goes through the roof easily!
It's good to know you can get the Kenwood to another level bridging the preamp but I think I will leave it as is for now as I'm actually using it's preamp, according to the manual when tone/presence controls are not engaged they're out of the circuit (I guess the rest remain in: attenuator, loudness, stereo-mono modes, etc)
Thanks for sharing your test. I am curious of the subsequent results. Great that you have no space limits, as the crossovers are inside the cabinets!
I'd recommend to try my "version" of the B139-B110-T27 crossover with all fresh caps, and with the cap styles I marked in the schematics (and with the exact inductor values). Not all cap styles can be substituted for others. To my experience film caps as substitutes in British vintage crossovers are often making things worse – often somewhat more "trebly" or "middy", which might at first be perceived as more detail.
A lot of testing and explanation has been done here by alan-1-b, among others:
Celestion 66 needs mid-range
Even if you switch out the tone network in the Kenwood the signal still must pass through the switches, that's where the problem is. What main amp are you using for your tests right now? The Yaqin? Tubes and KEF Bextrene drivers (which are not very efficient) might be a slightly difficult marriage (I am trying to find more reasons for your experienced lack of upper bass).
I'd recommend to try my "version" of the B139-B110-T27 crossover with all fresh caps, and with the cap styles I marked in the schematics (and with the exact inductor values). Not all cap styles can be substituted for others. To my experience film caps as substitutes in British vintage crossovers are often making things worse – often somewhat more "trebly" or "middy", which might at first be perceived as more detail.
A lot of testing and explanation has been done here by alan-1-b, among others:
Celestion 66 needs mid-range
Even if you switch out the tone network in the Kenwood the signal still must pass through the switches, that's where the problem is. What main amp are you using for your tests right now? The Yaqin? Tubes and KEF Bextrene drivers (which are not very efficient) might be a slightly difficult marriage (I am trying to find more reasons for your experienced lack of upper bass).
Thanks, I will keep you posted!
It's being fun as I'm a newby and I'm beggining to understand the tip of the iceberg on how components affect sound. I've been comparing again the Jantzen Standard film caps (30uf) in series with the B110 and now I notice the difference more than yesterday, again, the electros are old Elcaps, so not a fair comparison.
For the tests I'm using the Kenwood, it's handy having the balance control and comparing one speaker to the other or listening just to one channel, although I also change the crossover boards on the same speaker. Today I tested a 4mh inductor from some Wharfedale XP2 I inherited, and liked what I heard, I think there was some increase in lower mids. I take into account your advice of using the exact values you posted, but I'm leaning towards getting something in between, maybe about 4.5mh...
It's being fun as I'm a newby and I'm beggining to understand the tip of the iceberg on how components affect sound. I've been comparing again the Jantzen Standard film caps (30uf) in series with the B110 and now I notice the difference more than yesterday, again, the electros are old Elcaps, so not a fair comparison.
For the tests I'm using the Kenwood, it's handy having the balance control and comparing one speaker to the other or listening just to one channel, although I also change the crossover boards on the same speaker. Today I tested a 4mh inductor from some Wharfedale XP2 I inherited, and liked what I heard, I think there was some increase in lower mids. I take into account your advice of using the exact values you posted, but I'm leaning towards getting something in between, maybe about 4.5mh...
Glad you find it helful. In my case it was totally worthwhile, but i guess it was because at comparing them with other speakers I felt something amiss...
Here are the results ->https://www.diyaudio.com/community/threads/kef-concerto-or-kef-reference-104-2.390903/post-7171463
By the way, forgot to say thanks to Eschenborn!
Here are the results ->https://www.diyaudio.com/community/threads/kef-concerto-or-kef-reference-104-2.390903/post-7171463
By the way, forgot to say thanks to Eschenborn!
Hi ! Slightly resurrecting this thread to ask your opinion on the concertos vs ditton 66, w. Original crossovers and upgraded versions ? Looking for a nice three way speakers w. balanced yet laid back sound, and just a bit of sparkle. I have a pair of Chorale w. Falcon crossovers. The T27 is detailed but a bit harsh on the vocals, and the bass is there but not that well defined.
Thnx !
Thnx !
Hi @Zeccar , thanks for asking. I never heard the Dittons. There is a great thread with many comments and replies by alan-1-b to the Ditton-recap question: https://www.diyaudio.com/community/threads/crossover-nightmare.152995/
The Concertos with my changed crossover are very competent and as I wrote not very different from the Ref 105s. I like them more than the (also recapped) Cantatas.
The Falcon version of the Chorale crossover is their number 37k. Did you use this? It has the aB section – as stated above I also implemented the aB circuit for the T27 tweeter in the Concertos. With the aB section in the crossover the T27 is very good and I never hear it as harsh (there is another thread somewhere here which shows what effects the KEF aB section has cascading down the frequency spectrum which the speaker transmits).
The T27 with aB is this good that it can reveal upgrade possibilities in your amplification. The same goes for the bass – the Chorale has a surprisingly deep and clean bass already with the original KEF crossover. That's my experience (I own a pair... recapped with a Falcon kit but not upgraded for aB, to have a reference for the original circuit which is a bit more crisp in the highs).
The Concertos with my changed crossover are very competent and as I wrote not very different from the Ref 105s. I like them more than the (also recapped) Cantatas.
The Falcon version of the Chorale crossover is their number 37k. Did you use this? It has the aB section – as stated above I also implemented the aB circuit for the T27 tweeter in the Concertos. With the aB section in the crossover the T27 is very good and I never hear it as harsh (there is another thread somewhere here which shows what effects the KEF aB section has cascading down the frequency spectrum which the speaker transmits).
The T27 with aB is this good that it can reveal upgrade possibilities in your amplification. The same goes for the bass – the Chorale has a surprisingly deep and clean bass already with the original KEF crossover. That's my experience (I own a pair... recapped with a Falcon kit but not upgraded for aB, to have a reference for the original circuit which is a bit more crisp in the highs).
PS Falcon's aB section for old KEF speakers works well, but the caps they offer for the bass sections can change the sound towards "thinner-harsher". I stopped using the Falcon kits because I found them not really tailored to reproduce the original sound.
What I would first try is to put a 1R or slightly higher 5Watt resistor in series with the parallel cap in the bass circuit of your Chorales or replace this cap with a new one with a relatively high ESR (Audyn e.g.). I am planning to look into this same thing with my pair of Falcon-recapped Calindas this winter.
What I would first try is to put a 1R or slightly higher 5Watt resistor in series with the parallel cap in the bass circuit of your Chorales or replace this cap with a new one with a relatively high ESR (Audyn e.g.). I am planning to look into this same thing with my pair of Falcon-recapped Calindas this winter.
Hi @eschenborn
Thnx for the insight, then the culprit could be this nice little kenwood fellow, as I haven’t changed its old electrolytics yet. Also, I normally use the chorales in a small 9m2 room as a second set of nearfield monitors (the other is a pair of Yamaha HS8). Both speakers get boomy in there if I crank up the volume, but the warm signature of the Chorales’ works quite well at low amplitude.
I am looking for either ditton 66 or concertos for the living room (25-30m2). Chorales do work in there but they perform better in the other room, especially the bass repsonse. I also suspect the thin walls’ material to be quite reverberating.
T27 :
I think I’m used to the laid back treble of the Harbeth P3esr. The T27 brings more detail forward, and while amazing on trumpet for instance, the vocals are a bit harsh. It’s like hearing someone talk at 1m distance vs having them whisper in your ear. Mb it comes down the tweeter’s signature and I should look for the LS3/5A T27 grilles to tame this harshness a bit. (Maybe EQ too)
Crossover :
I do have the 37k crossover in the Chorales, and it’s amazing when compared to the original x-over with out-of-spec electrolytics. It seems falcon discontinued those vintage kef x-over upgraded boards so custom ones are the way to go regarding concertos if I get a pair.
I tinkered a bit with boxsim and afaik x-overs should be strictly designed according to a specific enclosure, so a bit skeptical on casual transplants, but if I get the concertos, I’ll follow your design for sure.
I haven’t heard the ref 105, my only references are the speakers mentioned in my reply.
The concertos go for 200-250€ here in Paris, the ditton 66 are less rare but usually cost x2 to x3 and more. Also considering a pair of DM2A…
Thnx for the insight, then the culprit could be this nice little kenwood fellow, as I haven’t changed its old electrolytics yet. Also, I normally use the chorales in a small 9m2 room as a second set of nearfield monitors (the other is a pair of Yamaha HS8). Both speakers get boomy in there if I crank up the volume, but the warm signature of the Chorales’ works quite well at low amplitude.
I am looking for either ditton 66 or concertos for the living room (25-30m2). Chorales do work in there but they perform better in the other room, especially the bass repsonse. I also suspect the thin walls’ material to be quite reverberating.
T27 :
I think I’m used to the laid back treble of the Harbeth P3esr. The T27 brings more detail forward, and while amazing on trumpet for instance, the vocals are a bit harsh. It’s like hearing someone talk at 1m distance vs having them whisper in your ear. Mb it comes down the tweeter’s signature and I should look for the LS3/5A T27 grilles to tame this harshness a bit. (Maybe EQ too)
Crossover :
I do have the 37k crossover in the Chorales, and it’s amazing when compared to the original x-over with out-of-spec electrolytics. It seems falcon discontinued those vintage kef x-over upgraded boards so custom ones are the way to go regarding concertos if I get a pair.
I tinkered a bit with boxsim and afaik x-overs should be strictly designed according to a specific enclosure, so a bit skeptical on casual transplants, but if I get the concertos, I’ll follow your design for sure.
I haven’t heard the ref 105, my only references are the speakers mentioned in my reply.
The concertos go for 200-250€ here in Paris, the ditton 66 are less rare but usually cost x2 to x3 and more. Also considering a pair of DM2A…
Hi @eschenborn 
Thank you for sharing all of your experiments and findings on this crossover rebuild.
I unfortunately have no technical knowledge/skills and can't follow along, but am going to look for a local technician to rebuild my KEF concerto crossovers to your spec.
I have 3 questions:
1. If I provide a technician with these schematics and reference images,
is that enough to understand what's been done, or would a summary of written instructions be necessary?
2. I copied this list of parts from your diagram, is that everything I will need? I want to save the tech the job of ordering them if I can.
(I think you said I won't need to replace inductors, as they don't expire - not 100% sure which parts they are, but have put in brackets with "?" )
5 mH (Inducter ?)
Audyn raw 100V 82µF
------------------------------------
Mundorf E-Cap plain 70VDC 33µF
1 mH (Inducter ?)
2.2 Ω (Resistor ?)
4 mH (Inducter ?)
F&T ATBI 100V 6.8µF
1 Ω (Resistor ?)
10 Ω (Resistor ?)
Mundorf E-Cap plain 10µF
--------------------------
Expotus Polyester 600µF
7W 3Ω Resistor
Expotus Polyester 3.3µF
ALCAP 100V LL 10µF
ALCAP 100V LL 10µF
0.3mH (Inducter ?)
3. Where should I be able to buy all the parts?
(I live in Australia, but don't mind paying international shipping to get the right stuff)
Sorry for the very basic questions, and thanks in advance
Thank you for sharing all of your experiments and findings on this crossover rebuild.
I unfortunately have no technical knowledge/skills and can't follow along, but am going to look for a local technician to rebuild my KEF concerto crossovers to your spec.
I have 3 questions:
1. If I provide a technician with these schematics and reference images,
is that enough to understand what's been done, or would a summary of written instructions be necessary?
2. I copied this list of parts from your diagram, is that everything I will need? I want to save the tech the job of ordering them if I can.
(I think you said I won't need to replace inductors, as they don't expire - not 100% sure which parts they are, but have put in brackets with "?" )
5 mH (Inducter ?)
Audyn raw 100V 82µF
------------------------------------
Mundorf E-Cap plain 70VDC 33µF
1 mH (Inducter ?)
2.2 Ω (Resistor ?)
4 mH (Inducter ?)
F&T ATBI 100V 6.8µF
1 Ω (Resistor ?)
10 Ω (Resistor ?)
Mundorf E-Cap plain 10µF
--------------------------
Expotus Polyester 600µF
7W 3Ω Resistor
Expotus Polyester 3.3µF
ALCAP 100V LL 10µF
ALCAP 100V LL 10µF
0.3mH (Inducter ?)
3. Where should I be able to buy all the parts?
(I live in Australia, but don't mind paying international shipping to get the right stuff)
Sorry for the very basic questions, and thanks in advance
Last edited by a moderator:
hey @Dyleet , thank you for asking. The technician should be able to do it if he knows to hold his soldering iron on the right side. I will look at the parts list again and make it a little bit more generic, so it will be easier to choose the capacitors. Inductor can remain in place. You should get all the parts at the big electronic distributors. Please allow a moment to have a second look at my parts list.
These should work:
Bipolar 100V 82µF
------------------------------------
Bipolar 100V 33µF
2.2 Ω 10W
Bipolar 100V 6.8µF
1 Ω >= 5W
10 Ω 10W
Bipolar 100V 10µF
--------------------------
Polypropylene or Polyester 0.56uF (not 600 as you have written! 0.56uF is easier to get than 0.6uF, will do)
7W 3Ω Resistor (can be more than 7W)
Polypropylene or Polyester 3.3µF
Bipolar 100V 10µF
Bipolar 100V 10µF
If you find Mundorf or Fischer & Tausche or Alcap bipolar caps, use those (an option is to order from Falcon Acoustics UK or from Darisus Germany).
Bipolar 100V 82µF
------------------------------------
Bipolar 100V 33µF
2.2 Ω 10W
Bipolar 100V 6.8µF
1 Ω >= 5W
10 Ω 10W
Bipolar 100V 10µF
--------------------------
Polypropylene or Polyester 0.56uF (not 600 as you have written! 0.56uF is easier to get than 0.6uF, will do)
7W 3Ω Resistor (can be more than 7W)
Polypropylene or Polyester 3.3µF
Bipolar 100V 10µF
Bipolar 100V 10µF
If you find Mundorf or Fischer & Tausche or Alcap bipolar caps, use those (an option is to order from Falcon Acoustics UK or from Darisus Germany).
Hi @eschenborn thanks for answering these questions. I am also going to build your crossover.
What size/RDC specification is the extra air core inductor? (Does it matter?)
What size/RDC specification is the extra air core inductor? (Does it matter?)
thanks for the interest. I think the specifications do not matter so strictly, as the other values of the tweeter filter network are the same as in the version that KEF used, e.g. for the Calinda, and they did not choose air core inductors but standard iron dust core ones. I would not worry here and just buy any generic air core inductor.