I would like to build an external crossover for my Reference 3 META. The idea is to leave the internal intact (hard to get out, no room) and make a separate, external, high-quality crossovers. Given the amazing drivers and construction of the speaker's enclosure, this should be a worthwhile undertaking.
Does anyone have access to the schematics of the Reference 3 META? Hours of searching google returned nothing.
Thanks,
Michael
Does anyone have access to the schematics of the Reference 3 META? Hours of searching google returned nothing.
Thanks,
Michael
Have you asked KEF for it? They will tell you it will void the warranty but I can think of no obvious reason not to supply it.
Indeed but audiophiles that are into boutique passive crossover components want to use them for their speakers.
The Q series is a budget speaker making it debatable how financially worthwhile improving a basic budget crossover might be but, of course, it could make a fun DIY project. The R series bookshelf before this one (or perhaps before that - not really keeping up) did seem to have something a bit off with the crossover when I auditioned them prompting me to buy a pair of the coaxial drivers rather than the full speakers. Haven't heard the current R series but the measurements suggest it may be close to the best compromise achievable with passive components and not far off what could be done with an active crossover.
I guarantee KEF will not give you the schematics they put a lot of R/D into their products they are not just going to give those schematics away. Your only choice will be to reach into your cabinet with your phone take a good detailed picture of the crossover then un-solder the components or worse clip them out with wire cutters if they do not have there values printed on them you will have to measure their values then build a replacement. I am 100% sure this could be done you could even make the replacement fit inside the cabinet so you did not have to mount them externally ruining the look of a $15000 speaker. But let me ask you this why do you want to? Please do not get offended but do you really think you can improve the crossover? Lets say you get the schematics or the values of the components and you spend $1000 buying the best metal film capacitors larger gauge inductors with less resistance in attempt to get better bass and the best resistors money can buy and put it all together and not hear any improvement at all that is more than likely exactly what will happen. Or maybe the expensive metal film caps make the high frequency's sound a bit harsh and you will wish you just stayed with the stock polypropylene caps. In my opinion your wasting your time trying to improve the crossover. Here is a link you can read about the extensive R / D KEF put into these crossovers if this does not change your mind then good luck I wish you the best.
https://assets.kef.com/pdf_doc/REF/REF-White-Paper-201219-LR.pdf
https://assets.kef.com/pdf_doc/REF/REF-White-Paper-201219-LR.pdf
The old schematic is the one thing I don't need when redesigning a crossover. However if KEF did the good job everyone believes they did, you might be better getting an equaliser and voicing them, then branch out into room placement.
Get the x/o board out so we can make our own judgement of its parts build quality (anything else needs a different kind of an evaluation). If you were selling an "upgrade" to a client of yours, I could understand the drive to "improve" on it.
Hi Michael,
i like the idea of a try of improve something that seems to be already almost perfect. Looking of what he is doing here http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm the most promising improvement could be the full activation of the woofer to coax section with a 2 way plate amp or the semi activation of the woofer only with a 1 way plate amp. Class D is good for the woofer up to 450 Hz,, this is the crossover frequency as of the tech specs for this model on the KEF website. If you just want to rebuild 1 to 1 the passive KEF crossover it's another story, what do you want more - a complete redesign of the speaker or just a clone of the original crossover with more expensive (hard to say if this will be really better...) passive crossover components. I think the most expensive part of the project is the measurement equipment, i strongly believe that without measurement equipment even the clone of the KEF crossover could be what we in the German speaking area call "ein Schuß in die Eierkiste" a shot in a basket full of eggs. With my master degree in electrical engineering and some years of work experience in analog and digital hardware development my dream for this would be to make the use a Hypex plate amp or a miniDSP 2 X 4 HD for the measurement of the activation and then to build the crossover with Papa's analog crossover circuits to avoid the disadvantage of the analog to digital and digital to analog back conversion. There is a famous developer of studio monitors in Switzerland that makes some statements about the disadvantages of DSP crossover designs - https://strauss-elektroakustik.de/passive-monitore/ - Jürgen Strauss of Strauss Elektroakustik. If you want to go for DSP crossover, than you have to think about the adoption of FIR filter, but here i am out - my master degree is too old for that expertise.
i like the idea of a try of improve something that seems to be already almost perfect. Looking of what he is doing here http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm the most promising improvement could be the full activation of the woofer to coax section with a 2 way plate amp or the semi activation of the woofer only with a 1 way plate amp. Class D is good for the woofer up to 450 Hz,, this is the crossover frequency as of the tech specs for this model on the KEF website. If you just want to rebuild 1 to 1 the passive KEF crossover it's another story, what do you want more - a complete redesign of the speaker or just a clone of the original crossover with more expensive (hard to say if this will be really better...) passive crossover components. I think the most expensive part of the project is the measurement equipment, i strongly believe that without measurement equipment even the clone of the KEF crossover could be what we in the German speaking area call "ein Schuß in die Eierkiste" a shot in a basket full of eggs. With my master degree in electrical engineering and some years of work experience in analog and digital hardware development my dream for this would be to make the use a Hypex plate amp or a miniDSP 2 X 4 HD for the measurement of the activation and then to build the crossover with Papa's analog crossover circuits to avoid the disadvantage of the analog to digital and digital to analog back conversion. There is a famous developer of studio monitors in Switzerland that makes some statements about the disadvantages of DSP crossover designs - https://strauss-elektroakustik.de/passive-monitore/ - Jürgen Strauss of Strauss Elektroakustik. If you want to go for DSP crossover, than you have to think about the adoption of FIR filter, but here i am out - my master degree is too old for that expertise.
Last edited:
These are not cheap speakers, and we'll assume KEF has some serious R&D money invested in these. Finding a schematic floating around the interwebs is highly unlikely. You probably have a better chance playing the lottery. 😉
If you think you can design a better crossover, then you should have most of the fundamentals down already, and possibly have the ability to design a speaker from scratch.
jeff
Last edited:
Hi again Michael,
my quick search in the Google universe brought me back to some findings already made a few years ago. I think i remember that i found out somewhere that even in the Blade flagship speaker the KEF engineers used a kind of passive crossover like this here:
it is more or less in my findings an acoustical 2nd order Bessel filter (Q=0.58), that means the sum of the electrical filter and the behavior of the chassis itself together works like this. In the newer white papers KEF seems to make a big secret about this. For the first order crossover of the tweeter of the coax i could see it on my own, i have build my own crossover with my measurements imported to BoxSim for a KEF SP1753 mid range high range coax. The series RLC (R build in the inductance made with thin wire) circuit in parallel to the tweeter is crucial to make the series capacitor works like it should and needs to be trimmed exactly to the resonance frequency of the tweeter. Until some months ago there was a chance to get the KEF chassis as spare parts from a shop in France (setelec.fr shop), but they have unfortunately finished this business in the mean time apparently. To come closer to some practical findings you definitely need to dismount your speakers, but i do not see any screw on these cabinets, may be they use these adapter rings to hide the screws like the cheap Q series speakers. It will be tough to remove the rings without scratching the cabinet, I believe that KEF have special removal tools for these rings for servicing the speakers.
my quick search in the Google universe brought me back to some findings already made a few years ago. I think i remember that i found out somewhere that even in the Blade flagship speaker the KEF engineers used a kind of passive crossover like this here:
it is more or less in my findings an acoustical 2nd order Bessel filter (Q=0.58), that means the sum of the electrical filter and the behavior of the chassis itself together works like this. In the newer white papers KEF seems to make a big secret about this. For the first order crossover of the tweeter of the coax i could see it on my own, i have build my own crossover with my measurements imported to BoxSim for a KEF SP1753 mid range high range coax. The series RLC (R build in the inductance made with thin wire) circuit in parallel to the tweeter is crucial to make the series capacitor works like it should and needs to be trimmed exactly to the resonance frequency of the tweeter. Until some months ago there was a chance to get the KEF chassis as spare parts from a shop in France (setelec.fr shop), but they have unfortunately finished this business in the mean time apparently. To come closer to some practical findings you definitely need to dismount your speakers, but i do not see any screw on these cabinets, may be they use these adapter rings to hide the screws like the cheap Q series speakers. It will be tough to remove the rings without scratching the cabinet, I believe that KEF have special removal tools for these rings for servicing the speakers.
Last edited by a moderator:
Well, is interesting to see what goes in 15 to 25 thousand dollar speakers.
Dont know the version. Seems for the money, the LF section should
be able to keep up with the mid.
anyways seems logical they would come with redundant large magical components
people like to see.
At least I hope so.
On the simple side, open it up and trace out the circuit.
be about 15 mins of work.
Dont know the version. Seems for the money, the LF section should
be able to keep up with the mid.
anyways seems logical they would come with redundant large magical components
people like to see.
At least I hope so.
On the simple side, open it up and trace out the circuit.
be about 15 mins of work.
What would be the purpose of such a post by a bot/AI? (I'm not saying it is not just that I don't fully understand how the modern web works).
If we take the post at face value then it is from someone that is enthusiastic about the properties of boutique passive components. Nobody has responded positively with respect to this (though I think there are a few around that share this enthusiasm). Nobody has posted the schematic he asked for just ways to obtain it. From their perspective given we see things differently and cannot supply what is wanted why chat? They may also have got the information from KEF or elsewhere since it is not commercially sensitive and straightforward to obtain from inspection.
If we take the post at face value then it is from someone that is enthusiastic about the properties of boutique passive components. Nobody has responded positively with respect to this (though I think there are a few around that share this enthusiasm). Nobody has posted the schematic he asked for just ways to obtain it. From their perspective given we see things differently and cannot supply what is wanted why chat? They may also have got the information from KEF or elsewhere since it is not commercially sensitive and straightforward to obtain from inspection.