SpeakerTalk

[eschenborn]

oh that is crazy. and pretty sad. somebody should have downloaded it.

well, I am still revelling in the intermediate result I obtained with my mix of caps right now. every piece of music sounds great you'll know the feeling.

[speakerguru]

[eschenborn]

hi speakerguru, thanks for this info.

I actually thought a bit about this already. there will be so much data lost because of incombatibility between evolving computers.

at least there is a great substitute in the form of this forum.

another practical idea would have been a sheet of paper glued into the bass cab with a schematics and all the actual values my old sansui TR-707 receiver has a schematics stuck to the inside of its metal lid.

[speakerguru]

[eschenborn]

True! I'd actually have loved too, in a way. Or at least do a little time journey.

After my enthusiastic last posts I have had the well-known frustration dip after changing something in a circuit for the "better". I still found the (upper mid-treble) balance not right.

So I pulled out the crossovers again from the heads of my 105.1. This time I changed the 3.3uF caps at C10 back to electrolytics. As Falcon had supplied polyester films, I used Mundorf ECaps (smooth). I also used the rather big Mundorf 33uF ECaps for C5. I left the WIMA MKP4 0,22 at C11. now some days to listen... or only hours if it turns sour.

It would be better to make just one swap at a time, but that would double or triple the work and the required time.

In the meantime I have read the "Celestion 66 needs mid-range" and "crossover nightmare" threads over at DIY audio. there is a member (alan-1-b) which seems to be pretty much in the know.

I get that the newer alcaps supplied by falcon might not be electrically identical to the older elcaps. maybe this has been discussed here already. Any opinions? If this is the case, the crossover will never be the same – neither with the original elcaps (drifted, higher ESR), nor with the new alcaps. Nor with any caps, probably.

Or might serendipity have it that the actual mundorf ECaps are quite a bit similar electrically to the old Elcaps? The Mundorfs are also rather big physically.

The suggestion at DIY audio is to change everything to polypropylene with a lot of ESR-simulating resistors.

@audiolabtower, do you still have the documentation of your original values and the ESR-simulating resistors? you made the swap to film early on. Would be interesting to know what you have used. They give the values there for the 66.

Yet this would not take into account the matching between L variation and caps done by KEF. And it would mean big bulky tinkering. And an exterior crossover. Pros and cons. Maybe one day. Or if... try to go active? Maybe still another day

I'd love to keep the speakers aesthetically original.

So I feel one must kind of search one's way to a nice sound balance by tiny variations, which might take time. at this moment of writing, listening, it feels ok... a bit less spectacular again, but nicely integrated and a bit warmer than with the films at C10 – be it polyester or polypropylene.

but there might be the frustration dip around the corner. I'll tell.

What I find strange is that Falcon put the 3.3uF polyester into the replacement kit, although to my view it changes the balance. How did they come to that? To my eyes the Falcon people are kind of the heirs of the 1970-1980s KEF know-how (but they ain't got the computer tape records

Another question I am asking myself:

If one replaced the Zobel cap across the tweeter (C13, 2.2uF) with a film, this should not cause a lot of unbalance due to ESR variations, as ESR would be in series with a 47 ohm resistor, and not change the total resistance substantially. maybe this would even be valid for C6. Any opinions on this? Is this an important cap sound-wise? It somewhat seems so, as in later designs, the CS3/CS5 kit version for instance, KEF used a film – 1uF – with a 22R resistor in series across the tweeter.

Still another question: As several of us have recorded that this Zobel cap originally was way off (22uF instead of 2.2 in my case), and as the cap used was physically huge – did KEF intentionally use 22uF, and maybe one batch was mislabelled at the factory as 2.2?

Final question: Which might be the most sensitive caps for a change of ESR?

And as to the tank network: the C9-L6 combination at the input of the high pass is indeed in parallel, at least according to my tracing of the crossover tracks. I have it in place.

Thanks to all for following!

[audiolabtower]

I replaced everything with polyesters and a couple of polyprops padded to the mean of the original values (which were within 1% between channels), even the 360uFs. In those days very large value polypropylene were not that available. I did not put extra resistance in the bass feed, this has 20A cable to the drivers. The mid and treble could do with thinner cable since resistance not so critical considering the new caps. Nevertheless I ended up with 1.1 ohm 20 W in the B110 circuit, and 1 ohm 10 W in the T52. All polyprops might need a little more?
The zobel is not that critical. Some reckon series caps are more critical for sound than parallel ones.

[speakerguru]

[eschenborn]

Thank you, Speakerguru, these informations really help me!

I so far left the Mundorf smooth Ecaps in place for C10 in my 105.1 crossovers. Still sounds good to me. According to what you say, Speakerguru, by putting two 3.3uF Polyesters in the bag of the 105.1 recap kit, Falcon allows the ESR to change at a sensitive position. I imagine that some users will hear that as "more air", or "more spectacular", as I did first. I think now that it is not such a good idea.

But according to what you further say, it should be ok to put in films for the Zobel networks, right? Is it worth the effort, i.e. audible at all? The same should be valid for a shunt like C6.

KEF uses a 2.2uF Alcap for the Zobel in the 103.2, but a film cap in the CS3. Different cap values and resistor values as well. Both crossovers have very similar circuits and all the drivers are the same, and the other caps in the tweeter section of the crossover are of the same value, too.

[eschenborn]

Just to give you all an update of my journey: I have tested the crossover with Mundorf plain ECaps at C6. As I noted, I had tried Polypropylenes there (Mundorf MCap), but I found that they did not sound good (sort of flat sound, slightly muddled). I had left them in for a couple of hours.

Wit the plain ECaps the upper mids became too prominent. As in first great, but then making you turn down the volume. I was baffled as I thought the plain ECap would have no effect on the overall electrical balance of the crossover, but it did.

So after two days of listening and wondering I switched back to the blue Alcap 100V that was in the Falcon kit. In the schematic for the speaker that is somewhere on this site there is marked 150V - 50V LL for C6. The original Elcap was 100V, as was the new blue one supplied in the kit.

I also swapped C7 which in the kit was a 50V cap against a Mundorf ECap raw 100V. I expected a slight improvement, but again it was worse. So back in the blue Alcap. These are all interesting (and time-consuming) results. To me, they show that there are many more parameters at work than what I expected.

I might experiment with the shunt in the bass cap (plain ECaps to test against the blue "raw" (etched foil) Alcap which sits there at C1 already ordered). When I had time to do it, I will report here. I hope it is of use to some.

So far I could only improve the Falcon kit by replacing the supplied "raw" C5 (which according to my research is manufactured by Fischer & Tausche in Germany, who also supplies to Mundorf) by a "plain" Mundorf E-Cap, and also by replacing the supplied Polyester at C10 by a plain Mundorf. For C10 this was rather a correction than an improvement, as the supplied polyester gave an imbalance in the highs, again interesting, but fatiguing.

I could imagine to have the 105 very much "high-end" sounding through some of the changes I explored, but I'd run away after some hours. At first listening though, these "hotter" replacements give a detail and presence kick. In my (limited) experience, many hifi-ish sounding items have prominent upper mids (the Musical Fidelity B1 amp comes to my mind).

Eschenborn

[SaSi]

Anything that is very pronounced catches attention at first but in long term are sources of fatigue.
I can remember as a young student listening with full bass and treble tone controls. Sometimes with loudness depressed as well.
Now, I prefer the natural sound that comes with a flatter response. Keeping it flat another octave in the low end is all the magic. And the R107, in a decently treated room, with an amplifier with lots of reserve power, make all the difference.

I have purchased a recap kit for the R105 as well. It's in the drawer for a couple of years now. I better use it before the new caps need newer caps...

[eschenborn]

Hi SaSi,

thanks for chiming in. I totally agree.

I'd be curious as to what Falcon has sent to you in terms of caps. If the kit is already a couple of years old, there might be other (higher voltage) Alcaps in it and a 3.3 low loss Alcap for C10. I imagine that they might supply the Polyester because they ran out of stock of the 3.3 LL type (at least it is out of stock on their site right now as of Jan 2021). I ordered my recap kit in January 2019.