SpeakerTalk

[eschenborn]

hi to everybody,

this is my first post here. thanks for letting me in!

I would like to dIy a 105.2/CS9 clone, though with T33 (SP1074) tweeters (as in the 105.4). I have a spare pair here, and the T52s are hard to find. The B300 bass units (SP1071) are also here, a pair of B110 (SP1057) should arrive in a couple of days.

Does this sound like a doable concept? What would be needed to be done to the original 105.2 crossover? Substitute the 105.2 HF section with that of the 105.4? Or are other modifications needed?

Another question is the choice of caps and inductors. I could go the "film and air coil" way, or stay as close to the material used by KEF (i.e. try to get the electros from Alcap, and use the same style of inductors), to mirror KEF's calculation of ESR and coil resistance. What would you do?

any input will be greatly appreciated!
that will be a slowly evolving project, but somewhere it needs to start.
thanks!

[audiolabtower]

Falcon do kits for the 105 crossovers replacing the alcaps. If they can supply the coil values too you would be well on the way. I replaced all my caps years ago but if you went for alcaps now they should easily do for a decade or more.

The idea of the different sections for the tweeter is what I would do, if the spec of the T33 is more efficient then it would work more or less, if you felt it was too bright a series resistor could be used. I would compare the 2 impedance graphs of the networks for T52 and T33 ( in 105 and 105.4), if they are similar around the crossover point it might be ok.

Now if your T33 is 4 ohm it could be more problematical, you could have a very low impedance dip around crossover which would need a good current amp to survive, but most good ones nowadays should be fine, and the power spectrum of most music will be falling above these frequencies.

[eschenborn]

hi audiolabtower,

thanks for your reply!

The T33 is 8 ohms, so that should not be a problem. and, yes, good idea to ask Falcon directly. I have heard they are pretty helpful.

Anybody experiences with a film & air coil clone of the crossover here? I plan to go the as-close-to-the-original-as-possible-route, but I am curious about other approaches.

[audiolabtower]

Air core would be very much more resistance and probably diverge from the design if substituted just on values alone. Kef used iron dust core ferrites in the reference series which had lower distortion and higher saturation than standard ferrites that other manufacturers used. A high power ferrite type should be fine.

I replaced all my alcaps many years ago with polyesters and polypropelenes when I had access to professional measuring equipment for exact values. The result was clearer but brighter (which may easily have made the clearer!) My crossovers are external at the amplifier terminals so I put a board in for set on test resistors for mid and treble. I ended up with adding 1.1 ohm 20W in series with the midrange and 1 ohm 10W in series with the tweeter to compensate for the lower ESR of solid cores caps.

[eschenborn]

hi audiolabtower,

thanks so much for the additional clarifications and for sharing your experience. that's exactly what I was looking for.

so I will be on the lookout for high quality iron(dust) cores, and I will go the alcap route, as I planned anyway.

If anybody could supply a little additional expertise on using the HT section of the 105.4 instead of the one of the 105.2 here with the T33 I want to fit in, that would be fantastic. thanks!

[speakerguru]

[eschenborn]

thanks speakerguru!

to give some clarification: I am asking myself if I can "implant" the HF section of the 105.4 into the 105.2's crossover (keeping the circuit for the mids and lows as it is in the 105.2 because I will be using the B110 and the B300) without further modifications... or will this cause imbalances?

I could also speak of thinking of implanting the B300 circuit in the schematics of the 105.4. too put it the other way round.

I am too much an amateur to know if the sections (tweeter, mid, bass) can be treated as modular and can be freely combined, or if the crossover is an integral design. I will start answering my own question by comparing the schematics of both speakers.

in that case I should probably start hunting for T52's. but maybe that kind person way behind the switchboard in maidstone, kent, has access to a schematic of some attempts KEF did with the combination I am aiming at. in case this exists, that is.

[eschenborn]

[audiolabtower]

There's no way you could put a B300 in a 105.4 bass section because it was designed for two B200s (in parallel perhaps?), but with one certainly rolled off before the other for a single unit ranged into the mid driver low end. Since the crossovers incorporate response shaping and are in effect parallel sections for bass. mid, treble connected to the amp, the safest route would be the one I took and use multiple amps. The T33 and crossover connected to one amp will have no interaction with the other crossover connected to mid and bass and another amp. Balance is then easily achieved by adjusting the input levels to the amps.

Personally an alternative I would try - get the 105.4 circuit and build the head unit in an LS3/5a type box and use the 105.4 mid and treble sections with your T33s and SP1057s. The crossover would be designed for the dispersion of the original head unit so this would be closer than mounting the units on a large baffle, and drive these with one amp. The B300 could then be put in a 70 litre box underneath and use the 105 bass section crossover (or CS9 bass section) and another amp. Again balance would be fairly straightforward.

[speakerguru]

[eschenborn]

thanks speakerguru, thanks audiolabtower,

so it sounds I should try the T33 route. in any case T33, B110 and B300 are here, and falcon seems to be able to supply the neeeded crossover elements, including the right kind of inductors.

I'll post any progress on that project here. I am curious how it will work out. I have been dreaming of owning a KEF system with a B300 since I bought my CS3 speaker kit back when I was 16.

thanks so much for the nice reception and the informations shared. awesome place here.

[Nimo_jon]

Hi eschenborn

Wonder how has your project progressed. I am doing a very similar exercise now. I miss my KEF 105.2 after the tweeters died 20 years ago (I sold the pair away ). I have gatherrf a lot of information on the cabinet dimensions from this forum. This is my current work:

1. Reproduce the base cabinet using MDF but increased the internal volume to 90L The cabinet will be deeper and with internal brace supports
2. Reproduce the head units using MDF (the tweeters will be placed in the same position as the middle of the B110 cone center for time alignment.
3. Using active crossover Linkwitz-Riley variable from 100-400Hz for the B110. The adjustable bass will be a very useful feature
4. Bi-amp using two Quad 405 amplifiers

Modern technologies have given new drivers that are comparable to the original KEF units.
1. Morel MDT32s to emulate the T52
I like the Morels very much and have made kits from Wilmslow audio using the MDT29 and MDT33 units
2. Monaco SPH 135C to emulate the B110
The new Monaco drivers may be more musical than the B110?
3. Peerless SLS-315 to emulate the B300
Subwooder technology has improved a lot due to car audio

What are your thoughts? SpeakerGuru Sir?

This may not have the same sound as the good old KEF 105.2 . I should complete the pair by Sep 2018.

[SaSi]

Speaking of "emulating" the design, there can be much more flexibility in the design.
One aspect that one should consider is that Falcon acoustics offer brand new B1110 SP1003 and T27 SP1032 drivers. Much better options than 40 year old drivers pulled out of speaker units.
One additional option is that they have also brought out a brand new version of the B139, close to SP1044 specs with 3 times the power handling (and about twice the excursion for the cone. This could be a very good alternative for a B300.

For high quality reproduction, one should also consider going active. The benefits of active crossover vs passive are extensive - if not countless - and the two most important ones are:
1. Much easier to source three low power amplifiers to drive individual drivers with the woofer driven by an easy to find 2x80-2x100WRMS power amplifier. Less distortion to the mids and highs from the inevitable compression/clipping in the low frequencies in transients

2. Get done with the shortcomings and compromises of the passive crossover, energy wasted in there, cost of high quality inductors and capacitors (they can easily add up to the cost of a power amplifier for the tweeters) and much less flexibility (i.e. hassle) to adjust the performance.

Cost isn't the plus in going active in a three way system as you need an active crossover in the equation but the results will be awsome.

I tried to mess around (can't consider my experiment as design work) using a B139 SP1044 mounted in a KEF Calinda with the B200 opening blanked with pieces of wood - and the T27 opening closed using blown tweeters. I placed a pair of CS1 inspired enclosures in the shape of the 105 pods, using B110 and T27.

I used a Behringer DCX active crossover and three amplifiers that were available to me:
One Technics SU-V10X as pre-amp and driving the bass units - fed the complete signal to the DCX from the pre-out output and the return low pass signal back to the power amp in. Gives out more than 100WRMS and was plenty.

Two Technics SE-9021 power amps (2x60WRMS) for the midrange and tweeter. Much more than enough for the tweeter so signal for the high pass was attenuated. These power amps have meters for the power output and it was very handy for visual on-the-spot inspection of output power.

The one thing I was impressed with this exercise was the clean sound produced. Especially in the midrange. The bass was a bit less than desired but the box volume was clearly the culprit. One full experiment I am planning for this summer is to mate a B139 with two PRs in a 90lt box and see how that works.

[Nimo_jon]

Hello Sasi,

Thanks and nice to read your sharing. You are too right to say that the passive components L and C can add up to the cost of a decent amplifier, and provide little flexibility for tuning.

( I am currently restoring a pair of KEF 104aB which is over 40 years old. The cabinet is somewhat very tired looking. I have a professional carpenter who loves KEF to refurbish the cabinets. Real wood veneered on the exterior, black faced front panel. This will become KEF 104 activeBiamped.

Ebay offers high quality 2 way Linkwitz–Riley active cross over unit (at a cost of below USD80). The cross over frequency can be set at 3khz. Ebay also offers high quality built up stereo LM3886 amplifier module at below USD70. I can have an active cross over with 4 channels below USD300.

I have the T27 recoiled and added fresh ferrofluid. The B200 surrounds replaced with new rubber surrounds from SimplySpeakers. voic coiled realigned. The B139 are two brand new passive radiator. I managed to get 3 pairs of NOS B139 a few years back from England.

I will run the KEF biamped and cross the frequency at 3khz. I can adjust the gain of the amps and also adjust up or down the cross over Fc. I can measure the summed output. I have a good pair of KEF 104aB with passive crossover as reference. The originals has rather small ferrite coils.

Not sure how will the biamped version sound. The vertical polar response from the 3rd order BW will tilt upwards but the LW 4th order will be horizontal; wire the T27 in phase). The phasing of the cross over need to be reviewed. Maybe need to wire the cross over as BW 3rd.

Back to the KEF 105.2. The woofer, B110 and T52 are all aligned in a slope of about 80 degree from the horizontal. This time alignment of 3 drivers are also seen on many modern designs.

KEF put in a lot of expertise on the time alignment and unit positioning to give the 105.2 an amazing sound stage. The T52 center is at the center of the B110 cone. The distance between the two drivers units is critical.

[audiolabtower]