SpeakerTalk

[audiolabtower]

p.s. can't see your pictures

[T.O. Chef]

It would seem to me that if the board had not been modified with the cut in the trace then the circuit would have been : 30uF series, 0.028mH to ground, then 3.3uF series, 0.35uF to ground, 7uF series, zobel 2.2uF / 47R, then series fuse and T52+. The aB section included also.

But as Guru pointed out, a 30uF series on the HF input makes no sense.
_________________
Music is food for the soul.

[audiolabtower]

If you mean 0.35mH then it does not make any sense since a second order filter followed by a third order filter plus the tweeter rolloff would not be the linkwitz design of a total design fourth order rolloff. All I can confirm is my pcbs are laid out in the configuration shown in the picture I posted.

[speakerguru]

[speakerguru]

[audiolabtower]

Guru, the green wire photo (which my browser can't see) is not mine but T.O.Chef's, and are his modified crossovers, nothing to do with my crossovers or posts.

I can only state that my original pcbs have a large space for C9 (30uF) across the 28uH, as well as C11 (the 0.22uF across the third order caps). These are unpopulated on mine since had the normal 3.3/3.3 uF filter rather than 3.3/7uF shown in the circuit diagram.

I have never seen a 105 with the aB crossover, (which as I said earlier does not mean much) just the circuit diagram I found, so have no way of knowing whether the 30uF and 0.22uF (and 7uF) are the correct values. However my genuine pcb does follow that circuit topology exactly.

I put forward the theory that since the 104aB crossover had an extra high value capacitor in series with the tweeter, it has similar topology as the 105 circuit, the difference is the 104aB has 10uF across 0.3 ohm in series on the T27 side of the 3rd order filter, instead of the 30uF and 28uH on the input side of the 3rd order filter shown in the T52 circuit.

But if the aB refers only to the small value cap across the 3rd order caps, then these extra series high value caps are just normal shaping and not needed for the aB function?

[audiolabtower]

Chef - I can now see your pics, having disabled Malwarebytes. The SP1100 labelled pcb in the pics is nothing like mine, which are the original type mounted behind the B110 enclosure, with a large hole in the pcb for the vertical angle head screw, and fuses at the bottom with flying leads for the B300 connections. Mine are wider, more professionally laid out, and a completely different arrangement of components on top. However the circuit topology (component connections) of mine are as shown in the circuit diagram I posted.

[T.O. Chef]

Yes. I know exactly what you're talking about with the 105 crossover mounting and component layout.
What you describe is precisely the same as the 2 pairs of 105's I have had the pleasure of restoring.

This pair is different in many ways. The plastic back panel is instead a pair of flat metal covers screwed in place (my own screws showing). The pcb is horizontal instead of vertical. And there's that crazy component distribution.

BUT. The topology is exactly as described in your post except there is an aB circuit, a Zobel, and the strange 30uF in parallel at C5 on my board with the 0.028 L on the tweeter input.
I have meticulously sketched it out. There is no doubt in my mind at all the components are the same as yours in value and location other than as noted..

https://nofile.io/f/RyQPeaFKv9X

https://nofile.io/f/jco8knknS3u

Is (or was I should say) the pcb in your labeled SP1100 as is this one?
_________________
Music is food for the soul.

[T.O. Chef]

BTW .... Jerry (Falcon) seems to be aware of this configuration.
The cap set he sent me was an exact duplicate of what's shownin on the board as found (by me(.
As you can see from the photo.

_________________
Music is food for the soul.

[T.O. Chef]

Gentlemen, you know infinitely more than I about crossover design and particularly KEF.
I can't thank you enough for your patience and generosity taking the time to both assist and educate me.
It's a relief to know that the filters installed in my 105's have indeed been poorly modified by the addition of this additional "tank circuit" and some crude wiring.

Guru, I had already done what you suggested and removed that 30uF cap, left the aB circuit in place, and tested it with my ears. It sounds as it should in that configuration. I only put things back as you have seen the photo's for another test and to show you what it looks like the way I found them.

With all the components in and out repeatedly I'm afraid I've made quite a mess of the pcb's. Though they will work properly it will forever bug me just knowing it's the way it is.
I may have an ace up my sleeve however.
These filters are mine:

https://nofile.io/f/uNvQcm9aOaE

Removed from some 105.2's (and purchased by me at a cost I'd prefer to forget) they are now recapped and in seemingly good condition. I have tested them in my 105's. They did work, but I was not overjoyed with the bass response. To my ears the original 105 bass section is fuller and better pairs with the rest. The 105.2's bass only seems to come to life with higher spl's than I would generally listen at.

I am considering using these 105.2 filters for their B-110 / T52 sections only (bypassing the 480uF section) while retaining the original 105 bass filters.
Any thoughts on this?

In that my original filters are aB fitted I assume those T-52's benefit from it.
I note from schematics I've seen posted here that 105.2 capacitor values seem to match the original 105's but the inductor values differ. So I'm not sure if making modifications are advisable.
My second question is, if I could fit a 0.22uF aB cap in place and change C11 to 7uF, then add a 2.2uF into the empty C12 Zobel location, would it mess up the design?

Please. If you think this a crazy mans idea I won't be offended by your saying so. I defer to your judgement .... most appreciatively.
_________________
Music is food for the soul.

[T.O. Chef]

If anyone here should be interested, here's what the nearly complete restoration of my 105's looks like.
I do have the original full size cloth covered shrouds in good condition, but somehow I prefer the look with the heads showing. The jury is not out with a verdict on the sonic differences. ~

The top cover is made from 1/2" ply with 2 layers of different density foam atop, then covered with grill fabric.
The front grills are simple 1/2" ply, cloth covered.
The heads needed deep dents and broken off edges repaired. They were then painted with 3 coats of Rustoleum spray having a "hammered metal look" finish. I'm not happy with this look and will redo it with a flat finish then clear-coat overlay.

I added 'Bumpers" of hard plastic as Vacuum cleaner protection.
The castors are new, black roller ball low profile type, and the bass cabinet bottoms have been reinforced with a glued and screwed 1/2" ply painted black.

https://nofile.io/f/PenTttNe610

https://nofile.io/f/syaD7VOTMe6
_________________
Music is food for the soul.

[audiolabtower]

Yes my boards are most certainly screen printed as SP1100. I have seen pictures of that screw back head arrangement but never seen one. I did find a picture of a crossover "kit" which is similar but with a more traditional mid/treble pcb layout:

https://photos.app.goo.gl/wNQXZUeTFvYz5jL69

You can see in this picture the 0.22uF aB cap is o/c and also the 30uF across the 28uH.

Interestingly, here is a crop of the original 105 printed brochure. The bass crossover obviously had the layout changed in production, but the picture of the mid/treble crossover is accurate, and here the aB component and the tweeter extra 30uF series capacitor C9 and C11 are populated.

https://photos.app.goo.gl/3ZYshaJGeZP9BuNk7

[T.O. Chef]

Curiousuer and Curiouser.

How so very interesting there should be so many incarnations.
Just tweaking along the way, or do you think because of varying characteristics between driver production runs?
Or Both!

Now I really start to wonder what filters were in my 105's from the factory! Perhaps different boards altogether

_________________
Music is food for the soul.

[T.O. Chef]

Now that I'm as far as I think I can go using the filters that came with this 105 set (following Guru's advise, Thank You Sir) I am pleased as punch with what they can do with good material.

I have a Thorens TD-160mkii table tracking a Stanton 681eee cart. There's a good assortment of LP's here, a lot of Jazz and classic rock. I use internet radio (RP in FLAC), and I have a huge 660 GB digital collection in lossless format.
Lots and lots of music.

The verdict is: In my small work area the bass needs help. I add up to 6db at 100 Hz. It sounds fabulous, though it nearly lifts the floor above as it's very easy to keep turning that volume knob clockwise. It provokes a huge grin as it flaps the trousers. I think in a larger room this bass boost my not be required but I cannot be sure.
I am tempted to pad down the mid and HF input to get a better balance, if that's advisable

My question here is why does the sp1100 pcb have provision for a resistor at R2? It would seem to pad down the tweeter input. Has anyone ever seen that used, and was it for some other KEF system using the same pcb?
_________________
Music is food for the soul.

[habrune]

Sorry to revive this old thread; but to confirm:

I recently acquired old 105-I heads with XO boards, and these were also completely populated. This was under discussion.
So WITH C9 (30uF) parallel to L6 (0.028mH), and the complete aB layout and the Zobel C13+R3. Both are genuine KEF boards, installed as usual on the back of the heads, in good order, with functioning protection + window circuits. Boards SN 3093 and 3175. KEF inspection stickers, etc.
All caps showed to be the known black Elcaps with red ends, so to me it seems that there was no later messing with the boards, or adding to the boards. It all looked as installed in the factory.

Most caps measured values reasonably close to the printed value on them (according to the scheme from Audiolabtower, see his link above).
But as T.O.Chief also found in his XO-s: C13's in the T52 Zobel were way off (strange). Futhermore: L7 = 0,26mH, not 0,35mH; C10 both close to 5.3uF, not the stated 3,3uF (have they really aged like monozygotic twins?).

So I am puzzled by the denial from Speakerguru that KEF didn't issue boards with the debated 30uF cap in the T52 lead, under these visible proofs, if I may take my KEF boards as proof adding to the earlier presented ones.
But I cannot deny his much valued experience, and do agree to his affirmation that such a 30uF cap in the lead of the T52 is quite odd, to quote him mildly.

E.g., I ran a simulation in Xsim with a T52 graph found in an old "Lautsprecherjahrbuch"(Germany) taken in an anechoic room from HiFiSound, (Münster, Raimund Saerbeck), which shows a small dip at 5 kHz in the T52 response. In the simulation this dip is of course only augmented by the debated C9/L6 suckout filter in the tweeter lead, not compensated for.

I have to recap first, after that I probably will post some audio comment on this C9, although I am inclined to omit C9 and leave it at that...

Measured values my XO's:
C5 32,7 / 32,5 uF (30**)
C6 11,6 / 11,4 uF (10**)
C7 114 / 134 uF (100**)
C8 8,2 / 9,1 uF (7**)
C9 34,6 / 41,2 uF (30**)
C10 5,2 / 5,4 uF (!)(3.3**)
C11 0,21/ 0,21 uF (0,22**)
C12 7,9 / 8,8 uF (7**)
C13 16,0/ 19,4 uF (!) (2.2**)
L3 1,34 / 1,37 mH (1,3*) R = 0,3 Ohm
L4 2,02 / 2,09 mH (2,0*) R = 0,4 Ohm
L5 0,264/0,261mH (0,25*) R = 0,20 Ohm
L6 38 / 39 uH (!) (28*) R = 0,12 Ohm
L7 0,258/ 0,255 mH(!) (0,35*) R = 0,25 Ohm
R2 = 10,3 / 10,5 Ohm (10**)
R3 = 46,3 / 48,3 Ohm (47**)

** Value in XO-scheme in link provided by Audiolabtower AND printed on component
* Value in XO-scheme in link provided by Audiolabtower (unmarked component)