SpeakerTalk

[LWCM]

Hi, I just bought a KEF 105 mk1.

The previous owners have changed the caps before but I want to check to make sure the correct ones were changed.

I found a picture of the crossover circuit of the 105.1 on the internet. Can I check if this is correct?

Link to picture https://pix.sfly.com/8jeXR-


Also there is a distortion for some frequencies on one of the tweeter. What is the cause? Is it the capacitors?


The video is here https://youtu.be/aLyKaHta4Fs

If I need to change the tweeter, is the T52 SP1072 a direct replacement for the T52 SP1049? I know the screw holes appears different but other than that, can it be used to replace? I do note that when I compare the crossover diagrams of the 105.2 which uses the SP1072, the circuit and capacitors used are different. So I am not sure if I can swap in the SP1072 without modifying the crossovers.

Thanks for your advice.

[audiolabtower]

That is the crossover. Some versions did not need the aB circuit in the tweeter so were a bit simpler, that's why the circuit is different between some Mk1s and some Mk2s, not just between Mk1 and Mk2. If you could measure the caps it might give clue to the high frequencies, whether incorrect crossover response or mechanical issue. The 72 is a direct replacement, they just simplified the magnet and reduced the cost of production, but the cutout will need significant modification.

[LWCM]

If the circumference of the base is the same for the 72, can I not just slot it in and use new tapping screws to screw in the new holes? Has anyone done the change from 49 to 72 before?

[LWCM]

Also I checked Falcon acoustics and it seems they only sell the 5% capacitors replacement sent. Can’t find the 2%. Is 5% ok?

[SaSi]

For an MkI version, the age is such that capacitors may have drifted seriously. I've had with a Calinda and 104 pair that sounded muffled. Some capacitors had drifted by more than 50-60%.

So, since the age and materials are similar, if that has happened to the capacitors of that crossover, 5% tolerances would be much better than what you have.

In order to get something better than that, closer to what KEF did, you should be able to match values, not tolerances, between the inductors, capacitors, resistors (maybe) and drivers. That is more complex than getting lower tolerance capacitors and impossible unless you know the various value segments that were originally used AND get plenty of components and select from them. Impractical, even before considering that the drivers themselves are not brand new and performing as they originally did.