SpeakerTalk

[habrune]

I managed to acquire a very cheap pair of CS9 kit cabinets. Very small dent in one T52, tiny spot on one B110B with coating loss, both with no impact on sound and soundstage. And rubbing B300's. After a rigorous 180 degree turn (new holes) this pair sounds terrific. Overal smoothness, stereo soundstage despite the large front, impulse respnse and clarity of detail I never heard before in all my earlier KEFs, e.g Cantata with same mid+high.

Because the cabinets have B110+T52 quite low above the ground (reference axis 6 degrees up), furniture easily hides the soundstage. So, I want to mould them into DIY 105 rebuilds.

I have read that the 105's have trouser-flapping bass but slightly less mid clarity and detail than the 105-II's. So I think I will prefer the 105-II's. 105(-II)'s need a different crossover, due to different diffraction effects (wide front, the CS9).

Colin R published on hifiloudspeakers.info-technical area (crossovers) his analysis of the 105-II crossover august 2004, in two versions, depending on serial numbers of the 105-II.

Can anybody please comment on these two versions 105-II, is the last version 105-II the best and most refined, or is it a cheaper one due to market economics or other factors (component availability? small driver changes?) Any preferences? 105-IIa or 105-IIb?

[audiolabtower]

Guru will know the best but 105 II ran from 1979 to 1987, while CS9 ran from 1981 to 1990, so it is likely the drive units were those left over from the reference series closer tolerance pairing selection. Logically the later crossover would be due to production tolerance optimisation for whatever reason. You should have date codes on the drive units which might help.

Bear in mind the actual selected values differ from the written value, eg tweeter capacitors often have a lower value on one side of the inductor compared to the one connected to the tweeter, even though both are specced at 3.3uF. The wrong way round and frequency response and impedance will suffer.

[habrune]

Audiolabtower, thanks for your realistic answer. I get from it that the deviation between actual values needed for these specific individual drivers to achieve optimal performance and the specced values might be more relevant than what XO-version to apply.
Impossible for me to know what these optimum values are, of course. This might make the operation a fictitious improvement, good as these lumpy CS9 boxes already perform, although I am still curious what speakerguru 's comments will be.
The Constructor Series units don't show the usual inspection tag, so I can't figure out a date. Only serial no's, e.g. T52 serialn 028871. Aug 1971 can't be right....

[audiolabtower]

Kef bought in their caps and sorted them into 1% bands for the Reference series, measured and paired the drive units and wound inductors to suit based on what the computer calculated for best match middle response. I was able to measure all my alcaps when I replaced with polyesters and polyprops, but the last time I looked falcon still sold 105 alcap kits and they will give a good enough result.

https://www.falconacoustics.co.uk/replacement-capacitor-sets/replacement-capacitor-kit-kef/kef-105-2-2-replacement-capacitor-set.html

To give an example my tweeter caps (both 3.3) were 3.526/3.556 uF on the input side and 3.791/3.895 uF on the driver side.

[habrune]

Yes, thank you for the link, I am aware of the admirable KEF matching procedure and the Falcon offerings (I recapped 103's, 104's and cantata's). Did you measure 3.526/3.556 at the time when the 105-II was purchased new or years later? In the latter case the question arises how much of the deviation was deliberate and how much due to ageing..
My goal is to select the best 105_II XO version, as I am free to choose because this is not a recap job, but a rebuild in new form.
Maybe I'll ask Falcon which version they advise/prefer (Malcolm Jones still there?).

[audiolabtower]

Measured about 5 years old so drift not an issue, besides it would have been remarkable if drift had been so consistent between 2 different samples. Crossovers now external to cabinets in a wooden box directly underneath the (3) power amps:

[speakerguru]

[habrune]

Audiolabtower and Speakerguru: Very pleased by your very informative answers. Thank you, really great to hear the information that is hard to get.

Am I correct to conclude that the bottom line is w.r. to crossover values, that unless you actually own a 105(-II), and measured not too long after purchase the correct XO-values, you cannot rebuild a 105 from single B300/B110/T52 units, because they all require different XO-component values IF you want the stellar matching achievement as KEF did back then. Impossible (or very hard) to determine these values, for me without a lab. And: yes (consolation) you can build a decent set with the KEF XO's from the shelf or from Falcon. Not the top as KEF intended with the matching program but still very good. Not the top because the individual variation between units is not taken into account.

Speakerguru: Yes, in fact the little dent (hardly visible, near the rim) is a residu from a total push-in by inquisitive children. I applied some heat which worked wonderful for the majority of the dome surface, however a small dent is hardly visible but obstinate and only changes place when warmed up. I tested both (small-dented and non dented) in a mono setup and there was one very well defined single narrow soundspot, even under various dome-angles, which is for me an indication of good inter unit equality.
Some worries about your remark in another post concerning the hard domes T27 &T52: "The hot melt adhesive bond between diaphragm and damping ring was prone to separate with time." Is warming the dome OK? (is the damping ring somewhere inside the dome, or is it the outer ring?)

Audiolabtower: Impressive! Thanks for the picture. Puts new standards...
Are the composed caps on the L&R side also for the 3rd order bass roll-off? Do you recommend them over the Alcaps/elcaps, in the light of the cost? I am considering this replacement but the cost makes me think twice...
I measured my T52-section Alcaps (KEF crossover 2034/35). Nominal values 5uF and 7uF, I measured 5,37(L) and 5.39 uF(R) / 7,50 (L) and 7.43 (R) @1000Hz, quite neat I believe and in line with your findings (although mine now are >28y after production).

[habrune]

Audiolabtower, I have to apopogize because I see that you adressed in other posts your 105 recap issue etc. in detail. Didn't do my homework properly, sorry.
One question remains: you mentioned the actual cap's values for the T52 section in your 105. Do you have a list of the measured values of the other caps, and are you willing to share it?

[audiolabtower]

Hi, look at it this way, only Kef and Spendor went to the expense of selecting crossover components, or at least not many others claimed to do so, except for the hand calibrated samples specially built for reviews (in most other manufacturer cases, particularly in the Haymarket rags). So the Falcon kit will be good enough, and better than caps that have drifted more than 10% from original.

The bass caps were free at the time since I used them in a voltage doubler circuit to get a high voltage for injectors in one of the first electronic controllers for direct injection diesel engines for Volvo trucks many years ago. Some shorted the bass caps, I reckoned Kef knew what they were doing with them. Today I would just use new alcaps, they should be nowhere near as critical as the mid/treble crossover by the time the bass is doing well actually in the room.

Mine are Mk1s so the values won't help you much. They were all within 1% of each other between channels. The series bass caps added up to 372.5 and 372.6 uF, the other 79.9 and 80.2 uF. The mid caps were both 31.2, 10.8/10.9, 7.2/7.3 and 95.6/95.4 uF (I left out the other 2 decimal places!).

[habrune]

That is a reassuring reply, much appreceated, and thanks for the values, a help in the sense that tuned values are not that far off so that the Falcon kit will do.
Thanks for this illuminating thread!

[speakerguru]

[habrune]

Thanks Speakerguru. That ring is still in place, pff...(sweat..)