To save labour costs, I am dismantling my 104/2 spkrs.
Four woofers will be taken to Auckland to have donuts replaced.
Thanks to MikeA for his offer, will organise a trip to visit him to have xover caps replaced.  A set of replacement caps have arrived from Falcon Acoustics, UK.  Said to be to KEF specs ands sorted to 5% tolerances.

First speaker dismantled, marked A.  Photos taken each step of the way to aid reassembly.  Took 1.25 hours.  Spkr B should take a little less.

Thanks to advice from Owen, snipped off the wires to bottom woofer and will get repairer to tidy the terminals up ready for resoldering.
It was easy to pull off the spade connectors for the top woofer from the xover and have left those wires connected to the woofer. Makes it easy to identify which woofer goes where when reassembling..

Speaker B took just under 45mins to take apart.
A big surprise awaited. Both woofers have intact donuts. They seem to be a rubber compound rather than the rotted foam on the woofers of spkr A. They must have been replaced in the not too distant past. Another surprise was that the wires to each woofer were connected by push on spades!? Not soldered. Spkr A wires were all soldered on. Since the xover end is also a push on spade, I guess this shouldn’t be an issue.

Looks like the only requirement for spkr B is to replace the xover caps. Will email the tech, though, to see what type of material the replacement donuts for spkr A are made of, and whether it will make a difference in sound quality if the donuts on one spkr are of a different material to the other.
The bottom photo shows remains of the rotted foam donut - very different material.

Another surprise on spkr B. There seems to have been two non standard caps used as replacements on the xover. See photo. A on the R and B on the L. At bottom just to L of centre, there are two black caps (in parallel?) on A, one a bit smaller than the other. On B there is a grey cap and a white ceramic, presumably a cap of some sort?

Looks like repairs have been made using non standard parts?  What is the name of the round brown coloured device that appears to be joining two black caps, one smaller than the other, together on the R xover board?  There isn’t one supplied in the replacement pack from Falcon, so I will have to buy one from Jaycar.





upload image

The brown disc on the right crossover looks like a polyswitch, which is a type of resettable fuse. If the current gets too high if you’re playing music loud the fuse material melts inside it and increases resistance. It solidifies after the current is removed, hence it's resettable. The number on it will be RXE...(a number) which tells you the current rating. It looks like it’s been added in series with the two parallel capacitors. It might typically be on the tweeter circuit. Is this the speaker that had less volume? Perhaps the polyswitch was interfering. It isn’t standard. If you measure it it should be less than half an ohm resistance. I’d make both crossovers the same and eliminate the polyswitch. As long as you don’t play the speaker super loud! If you’re worried about that you could use 2 new polyswitches of the same rating on both speakers.
As for the woofers I’d ideally get the same material on the donuts on both speakers. If the new crossover caps are matched to within 5% it would make sense to me to have the woofers close in construction to each other as well.

AMR-iFi R&D

Yes, the poly switch is on the crossover of the speaker that was down in volume compared to the other, in fact the first issue was this speaker shutting down with no sound at all.   Seems both crossovers have had some modifications!
Thanks Colin.

Here is a screen shot of a 104/2 xover which shows the polyswitch is not standard

Been doing lots of searching for information about these speakers, in particular -

1. trying to find out approx date of manufacture, and
2. Who made rubber donuts and when were they available.

I learnt that KEF used rubber donuts late in the production run and rubber replacement donuts were also available for a while. Then someone suggested that the woofers with rubber donuts may have been factory built replacements after the originals had failed, which would explain why one loudspeaker came to have woofers of a different material to the others. That lead me to checking the serial numbers of all four B200 woofers.

I found that they have the date of manufacture stamped on them! All four B200s were made on 22 May 1990. So my speakers would have been assembled sometime after that. It seems the rubber donuts were replacements, but it is a mystery why someone would only replace one set of woofer donuts and not the other!

Perhaps to save on repair costs if one speaker failed before the other. Sounds like rubber ones are going to last longer. Are the rubber donuts on your woofers a little stiff? As rubber stiffens a bit over time.

AMR-iFi R&D

ColinF, the rubber of the rubber donuts is nice and soft and very flexible. Am going to Auckland this Friday, 10th, and will drop woofers off for repairs. Decided on Axent for the practical reason Mark of Audiomark is away on holiday.

Axent have been doing this sort of work for years, and fingers crossed they have rubber donuts in stock! If not will have both pairs of woofers repaired using donuts made of the same material.

Big hifives to MikeA, who mounted all the replacement caps on the xover boards yesterday. Very professional workmanship. Shows the value of this forum. Great to meet a fellow Darklantern member face to face. Goodness me, what an awesome system this man has. Privileged to sit in the sweet spot and have a listen after the work was done.

Will.reinstall the xover boards today, but won’t be able to have a listen until woofer donuts replaced.

Looking good!
Is MikeA willing to tell of his awesome system?
😀

AMR-iFi R&D

Ha, not sure about awesome, just a bit of synergy that needed a good pre-amp to pull it all together.

Ah yes, the Audia Flight

AMR-iFi R&D

May 7, 2019 23:38:23 GMT 12 colinf said:

Ah yes, the Audia Flight

Ah yes, the Audia Flight  

Good to meet John on monday.

Strangely the polyswitch appeared to have been added at manufacture as the hole for one of the caps had never been soldered...

Oh! I wonder if any other 104/2 owners have a polyswitch as well, and also why weren’t the two crossovers the same...Where was the polyswitch soldered to?

AMR-iFi R&D

If you google for a schematic, the polyswitch was connected to the caps at C13&14 in the HF circuit. The schematic doesn’t show a polyswitch. There are many pics  on line of 104/2 xover boards. Most do not show a polyswitch, but I have seen two!
It seems that during the production life, several different xover layouts were used., just to make life confusing!
Descriptions of the xover says it has a protection circuit, but the schematic for mine doesn’t obviously show one.
On the second xover board - no polyswitch, but instead the original caps at C13&14 had been replaced - a standard looking 3.5uf with no name, and a 4.7uf. Made by Ripken Chemicals (since defunct I understand. A mystery?

Dropped the woofers at Axent Audio on way to Tauranga on the 10th, collected on way home on Friday 17th.  An excellent job done replacing  all four
donuts.  Thanks to me doing all the labour  taking things apart, only cost $193.50 - very reasonable.

put everything back together today, and have them playing away quietly breaking in the new capacitors.  Despite not being at their best, they are sounding very very good.

Soldering aided by a simple tool I made- screwed a wooden clothes peg to a board and used that to hold wires steady leaving me with two hands for the soldering!

thanks again MikeA

Happy to help and even happier that everything is working properly...

Glad it is all up and running and the donuts are all the same now.
Great idea with the wooden peg screwed to a board to hold cables in place while soldering. I’ve been using pliers with an elastic band to hold them shut onto the cable to do that, but a peg could work even better! A wooden one as well, so that soldering won’t melt a plastic one.

AMR-iFi R&D

This is the thing to use:-  www.conrad.com/p/third-hand-3-piece-toolcraft-zd-10d-588221

Most useful thing my ex ever gave me...

That looks the biz!

AMR-iFi R&D

Hmm, all not so well! Have been proceeding gently so far, playing non complex music to break the capacitors in.
Then put on an Ottmar Liebert cd with lots of bass and treble and turned volume up.
Noted L spkr (spkr B) louder than R (spkr A).

Found culprit, the upper midrange B110 on spkr A is not working. Swapped feed cables from the lower MR to upper, and upper started working, and the lower did not!
So, not a problem with the MR spkr, and logic tells me if the problem is with the xover, neither MR would work?

Pulled everything apart again. See photo. The white plastic box with an array of cables plugs into the crossover, and the cables connect (soldered on) to a small PCB with lugs that individual cables connect to to feed upper and lower MR, plus the tweeter.

The white cable from the plastic box connects to the +tive lug on the PCB, and the black to the negative.

I used.my multi meter and did a continuity test on all cables. All except the white and black cables tested ok.

So, perhaps a question for ColinF. It seems strange that both cables failed the continuity test. Could it be a dry solder problem? How are the cables connected to the while plastic box? Is it safe just to pull them out?

Edit:  Penny dropped and have done another continuity test.  My first test was done by inserting one probe into into the socket on the white box that a wire from the xover plugged into.  Second test I inserted one probe alongside the wire being tested, and this time both black and white cables tested ok for continuity.

That must mean there is a poor connection where those wires plug into the white plastic box.  Now have to work out how to fix that!

he

May 27, 2019 12:46:56 GMT 12 sub said:

Second test I inserted one probe alongside the wire being tested, and this time both black and white cables tested ok for continuity.

That seems unusual
When you say 'alongside', do you mean that you mean that you inserted the probe into the incoming wire-side of the pictured plastic box (connector block)?
And the other probe into the other side of the same plastic block?

DarkLantern blog - darklanternforowen.wordpress.com

May 27, 2019 16:28:59 GMT 12 Owen Y said:

May 27, 2019 12:46:56 GMT 12 sub said:

Second test I inserted one probe alongside the wire being tested, and this time both black and white cables tested ok for continuity.

That seems unusual
When you say 'alongside', do you mean that you mean that you inserted the probe into the incoming wire-side of the pictured plastic box (connector block)?
And the other probe into the other side of the same plastic block?

No, first test inserted first probe into the input side of the connector box, ie where the prongs on the xover go, with second probe touching the appropriate lug on the small PCB board, and all wires except the two feeding the upper MR tested ok.

Second test on the failed wires, inserted first probe alongside the offending wires, and again second probe onto the appropriate lug, and tested ok for continuity.  Made me think there was a bad connection inside the the connector box for the failed wires.  They seem to be a push in fit so gave them a firm push to seat them home, and repeated continuity test one with positive results.

Problem solved, thought I, but no after reassembly problem still exists!  Did another partial dismantle and swapped the wires - upper MR to lower and vice versa.  Now upper MR working fine, but no lower!

Not sure where to from here!

I think when Owen means that he thought you meant when you mean (🤔) that the wires into the connector block itself are open circuit on the upper midrange only.
Those white connectors can be unreliable over time in that the wire going in to the metal contact inside the white connector block can corrode, and make erratic contact. They are generally crimped on. Over time the copper corrodes and the wire loses electrical contact with the metal contact right next to it.
I always prefer to eliminate those connectors if convenient by soldering the wires directly. But otherwise they pop out of the connector block with a small screwdriver pressed into the side while coercing it out. Try soldering the wire onto the contact then re-insert it.
Sometimes the wire can be so badly corroded, even underneath the insulation, that it needs replacing.

AMR-iFi R&D

Yes, been thinking the problem was in the connector box, but unsure how to check. Will have a look at pulling the wires from the connector box in a few days - got other priorities in meantime! MaYbe some deoxit could help?

Perhaps should clarify.

The photo above is not very focussed - poor lighting. It shows the plastic connector box in foreground. The wires exiting from the box run to the small PCB in the background, which in the photo has been disconnected from its housing so I could see where each coloured wire from the connector box is connected, and are soldered on from under the PCB. The “empty” side of the connector box plugs onto a row of thin connector pins on the xover.

A photo below shows the PCB in situ with wires plugged onto it with spade connectors, each pair of wires goes to - upper MF driver, lower MF driver, and to tweeter.

Sorry for my layman’s descriptions!

I think we can grasp the gist of what is happening now.
As colinf suggests, you should try to get the 2 suspect wire end connectors out of the plastic block, to inspect them for corrosion, etc - ie lack of conductivity.

Sometimes you can prise / push them out, as colinf describes, either forward or backward.
Possiblty the plastic block can open up - eg can be prised open with a knife say. Check if it is a 2-part or 'hinged clamshell' type block.

DarkLantern blog - darklanternforowen.wordpress.com

Oh I see now, thanks for the updated description.
So from the crossover, the pins on it seat into the white connector block. Those wires from the connector then go to the small pcb on the rear wall of the midrange/tweeter enclosure. The spade connectors on the other side of the pcb go to each of the drivers. At the drivers the wires have been soldered on. Correct?
So when you swap the lower midrange wires with the upper midrange wires on the spades on the pcb, the sound swaps between the two midranges. That means, as you say, the problem open circuit wires could be on the white plastic connector, or its contact onto the crossover board.
I wonder if the contact itself has lost its spring pressure over time as well, which means it looks like it could be working but as soon as you plug it in to the crossover board, connection might not be made. In which case you need a new connector, or solder the wires to the crossover.
As Owen suggests as well, examine around there and clean/solder/do-whatever-it-takes to get the connection to have low resistance again. I suspect this could actually be the source of all your sound level problems after all, not just the polyswitch! Intermittent contact on wiring can be elusive but when it does go consistently faulty that makes it much easier to diagnose.

AMR-iFi R&D

Thanks ColinF and OwenY, will investigate further in a few days time, after I have attended to a couple of more pressing chores that have been lined up for me!